{
  "groups": {
    "batch": [
      {
        "label": "Batch 16",
        "id": "8ac87378bb5b"
      },
      {
        "label": "Batch 64",
        "id": "1150d4df4a82"
      },
      {
        "label": "Batch 256",
        "id": "50c1636b4f2c"
      },
      {
        "label": "Full batch",
        "id": "1fb401058891"
      }
    ],
    "update_budget": [
      {
        "label": "Batch 16",
        "id": "014fce8be51b"
      },
      {
        "label": "Batch 64",
        "id": "6a3fa29d2adc"
      },
      {
        "label": "Full batch",
        "id": "92630a9c55d2"
      }
    ],
    "learning_rate": [
      {
        "label": "LR 0.0001",
        "id": "449a997d9d64"
      },
      {
        "label": "LR 0.003",
        "id": "1150d4df4a82"
      },
      {
        "label": "LR 0.03",
        "id": "be8cbd994d03"
      }
    ],
    "schedule": [
      {
        "label": "Constant LR 0.003",
        "id": "1150d4df4a82"
      },
      {
        "label": "Step decay",
        "id": "32d8a0bb9dc3"
      },
      {
        "label": "Cosine decay",
        "id": "1298f0867b9f"
      },
      {
        "label": "Validation plateau",
        "id": "5c044ad193b4"
      }
    ],
    "architecture": [
      {
        "label": "Plain MLP (batch 64)",
        "id": "1150d4df4a82"
      },
      {
        "label": "LayerNorm",
        "id": "ed8dd0384733"
      },
      {
        "label": "BatchNorm",
        "id": "9967999fa0f8"
      },
      {
        "label": "Residual",
        "id": "c160e9485e77"
      },
      {
        "label": "Concatenated skip",
        "id": "d7d20cfbbc18"
      },
      {
        "label": "Dropout 0.1",
        "id": "a7bb18d2c173"
      }
    ],
    "activation": [
      {
        "label": "SiLU",
        "id": "1150d4df4a82"
      },
      {
        "label": "Tanh",
        "id": "5bc849f25b25"
      },
      {
        "label": "ReLU",
        "id": "9d4c0f0efefb"
      }
    ],
    "combined": [
      {
        "label": "LayerNorm + silu + constant LR",
        "id": "ed8dd0384733"
      },
      {
        "label": "LayerNorm + silu + plateau LR",
        "id": "14985df6b90f"
      },
      {
        "label": "LayerNorm + tanh + constant LR",
        "id": "be44f6a2021f"
      },
      {
        "label": "LayerNorm + tanh + plateau LR",
        "id": "13e54c648cc2"
      },
      {
        "label": "LayerNorm + relu + constant LR",
        "id": "ff9112ae3885"
      },
      {
        "label": "LayerNorm + relu + plateau LR",
        "id": "7a196ed2cdbd"
      }
    ],
    "loss": [
      {
        "label": "MSE + outliers",
        "id": "d5c218b8a06d"
      },
      {
        "label": "MAE + outliers",
        "id": "fc71c16e3acd"
      },
      {
        "label": "Huber + outliers",
        "id": "772fcca85495"
      }
    ]
  },
  "runs": {
    "014fce8be51b_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 64,
          "samples_seen": 1024,
          "online_loss": 0.2624188729969319,
          "train_loss": 0.1171494722366333,
          "validation_loss": 0.09625931084156036,
          "train_metric": 0.011798867024481297,
          "validation_metric": 0.009694884531199932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0342887900005735
        },
        {
          "epoch": 2,
          "updates": 128,
          "samples_seen": 2048,
          "online_loss": 0.08470732023124583,
          "train_loss": 0.07202138006687164,
          "validation_loss": 0.05905001237988472,
          "train_metric": 0.007253730203956366,
          "validation_metric": 0.005947301164269447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06424685999991198
        },
        {
          "epoch": 3,
          "updates": 192,
          "samples_seen": 3072,
          "online_loss": 0.06356297535239719,
          "train_loss": 0.054105907678604126,
          "validation_loss": 0.04593408852815628,
          "train_metric": 0.00544934906065464,
          "validation_metric": 0.004626313224434853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09389178399942466
        },
        {
          "epoch": 4,
          "updates": 256,
          "samples_seen": 4096,
          "online_loss": 0.054570174688706174,
          "train_loss": 0.04432539641857147,
          "validation_loss": 0.038256630301475525,
          "train_metric": 0.004464291501790285,
          "validation_metric": 0.003853068221360445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1242549560010957
        },
        {
          "epoch": 5,
          "updates": 320,
          "samples_seen": 5120,
          "online_loss": 0.04212082936282968,
          "train_loss": 0.03483256697654724,
          "validation_loss": 0.03159569203853607,
          "train_metric": 0.003508209250867367,
          "validation_metric": 0.0031822023447602987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15455862000089837
        },
        {
          "epoch": 6,
          "updates": 384,
          "samples_seen": 6144,
          "online_loss": 0.03481512506186846,
          "train_loss": 0.025710269808769226,
          "validation_loss": 0.021541653200984,
          "train_metric": 0.002589444164186716,
          "validation_metric": 0.002169596031308174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1850158930010366
        },
        {
          "epoch": 7,
          "updates": 448,
          "samples_seen": 7168,
          "online_loss": 0.02729321763399639,
          "train_loss": 0.020683294162154198,
          "validation_loss": 0.017299188300967216,
          "train_metric": 0.002083145547658205,
          "validation_metric": 0.0017423109384253621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21523390699985612
        },
        {
          "epoch": 8,
          "updates": 512,
          "samples_seen": 8192,
          "online_loss": 0.022401315309252823,
          "train_loss": 0.029240833595395088,
          "validation_loss": 0.024743523448705673,
          "train_metric": 0.002945029642432928,
          "validation_metric": 0.0024920771829783916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24482867200094915
        },
        {
          "epoch": 9,
          "updates": 576,
          "samples_seen": 9216,
          "online_loss": 0.024232984520494938,
          "train_loss": 0.019727662205696106,
          "validation_loss": 0.01657680608332157,
          "train_metric": 0.0019868980161845684,
          "validation_metric": 0.0016695550875738263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35505802599982417
        },
        {
          "epoch": 10,
          "updates": 640,
          "samples_seen": 10240,
          "online_loss": 0.02759063059784239,
          "train_loss": 0.022793717682361603,
          "validation_loss": 0.018758468329906464,
          "train_metric": 0.0022956999018788338,
          "validation_metric": 0.0018892838852480054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3855935330011562
        },
        {
          "epoch": 11,
          "updates": 704,
          "samples_seen": 11264,
          "online_loss": 0.019792282218986657,
          "train_loss": 0.018156468868255615,
          "validation_loss": 0.015147591009736061,
          "train_metric": 0.0018286528065800667,
          "validation_metric": 0.001525609870441258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4162633760006429
        },
        {
          "epoch": 12,
          "updates": 768,
          "samples_seen": 12288,
          "online_loss": 0.01675870963936177,
          "train_loss": 0.012159856967628002,
          "validation_loss": 0.01066514290869236,
          "train_metric": 0.0012246962869539857,
          "validation_metric": 0.0010741541627794504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4471983450002881
        },
        {
          "epoch": 13,
          "updates": 832,
          "samples_seen": 13312,
          "online_loss": 0.014479773462880985,
          "train_loss": 0.012895269319415092,
          "validation_loss": 0.010621171444654465,
          "train_metric": 0.0012987643713131547,
          "validation_metric": 0.0010697254911065102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.478025198000978
        },
        {
          "epoch": 14,
          "updates": 896,
          "samples_seen": 14336,
          "online_loss": 0.01538535028885235,
          "train_loss": 0.015002022497355938,
          "validation_loss": 0.010967210866510868,
          "train_metric": 0.0015109487576410174,
          "validation_metric": 0.0011045774444937706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5088478340003348
        },
        {
          "epoch": 15,
          "updates": 960,
          "samples_seen": 15360,
          "online_loss": 0.016265414022200275,
          "train_loss": 0.0220046304166317,
          "validation_loss": 0.017810842022299767,
          "train_metric": 0.0022162257228046656,
          "validation_metric": 0.0017938429955393076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5391119830001116
        },
        {
          "epoch": 16,
          "updates": 1024,
          "samples_seen": 16384,
          "online_loss": 0.012747942846544902,
          "train_loss": 0.0103490324690938,
          "validation_loss": 0.007960377261042595,
          "train_metric": 0.0010423166677355766,
          "validation_metric": 0.0008017400978133082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5698925430006057
        },
        {
          "epoch": 17,
          "updates": 1088,
          "samples_seen": 17408,
          "online_loss": 0.012613498466635065,
          "train_loss": 0.00944472849369049,
          "validation_loss": 0.008201558142900467,
          "train_metric": 0.000951238558627665,
          "validation_metric": 0.000826030969619751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6003217979996407
        },
        {
          "epoch": 18,
          "updates": 1152,
          "samples_seen": 18432,
          "online_loss": 0.009903730539008393,
          "train_loss": 0.012951320968568325,
          "validation_loss": 0.008948936127126217,
          "train_metric": 0.0013044095830991864,
          "validation_metric": 0.0009013040107674897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.630100100999698
        },
        {
          "epoch": 19,
          "updates": 1216,
          "samples_seen": 19456,
          "online_loss": 0.0183119301473198,
          "train_loss": 0.014333060011267662,
          "validation_loss": 0.012964108027517796,
          "train_metric": 0.0014435735065490007,
          "validation_metric": 0.0013056974858045578,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6596245960008673
        },
        {
          "epoch": 20,
          "updates": 1280,
          "samples_seen": 20480,
          "online_loss": 0.010544464745180449,
          "train_loss": 0.007037613540887833,
          "validation_loss": 0.005865545943379402,
          "train_metric": 0.0007088026613928378,
          "validation_metric": 0.0005907563026994467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6896938019999652
        },
        {
          "epoch": 21,
          "updates": 1344,
          "samples_seen": 21504,
          "online_loss": 0.007833955711248564,
          "train_loss": 0.006359841674566269,
          "validation_loss": 0.005162452347576618,
          "train_metric": 0.0006405399763025343,
          "validation_metric": 0.0005199433071538806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7200041560008685
        },
        {
          "epoch": 22,
          "updates": 1408,
          "samples_seen": 22528,
          "online_loss": 0.007362566893789335,
          "train_loss": 0.006041003856807947,
          "validation_loss": 0.004798504989594221,
          "train_metric": 0.0006084277993068099,
          "validation_metric": 0.00048328781849704683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7506199510007718
        },
        {
          "epoch": 23,
          "updates": 1472,
          "samples_seen": 23552,
          "online_loss": 0.007419797226248193,
          "train_loss": 0.006737899035215378,
          "validation_loss": 0.004314305726438761,
          "train_metric": 0.00067861657589674,
          "validation_metric": 0.00043452103273011744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8023271590009244
        },
        {
          "epoch": 24,
          "updates": 1536,
          "samples_seen": 24576,
          "online_loss": 0.008811210521344037,
          "train_loss": 0.007847617380321026,
          "validation_loss": 0.006709723733365536,
          "train_metric": 0.0007903833175078034,
          "validation_metric": 0.0006757788360118866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8323412419995293
        },
        {
          "epoch": 25,
          "updates": 1600,
          "samples_seen": 25600,
          "online_loss": 0.01035352962117031,
          "train_loss": 0.008859946392476559,
          "validation_loss": 0.005991469603031874,
          "train_metric": 0.0008923413115553558,
          "validation_metric": 0.0006034388206899166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8621125749996281
        }
      ],
      "best": {
        "epoch": 23,
        "updates": 1472
      },
      "best_validation_metric": 0.00043452103273011744,
      "config": {
        "label": "Batch 16",
        "batch": 16,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.02084516809071391
    },
    "014fce8be51b_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 64,
          "samples_seen": 1024,
          "online_loss": 0.30496128514641896,
          "train_loss": 0.09889104962348938,
          "validation_loss": 0.08465031534433365,
          "train_metric": 0.00995994359254837,
          "validation_metric": 0.008525669574737549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03426029100046435
        },
        {
          "epoch": 2,
          "updates": 128,
          "samples_seen": 2048,
          "online_loss": 0.08240926322469022,
          "train_loss": 0.060156285762786865,
          "validation_loss": 0.04975259304046631,
          "train_metric": 0.006058720406144857,
          "validation_metric": 0.005010898690670729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06466412500049046
        },
        {
          "epoch": 3,
          "updates": 192,
          "samples_seen": 3072,
          "online_loss": 0.055069187670596875,
          "train_loss": 0.05514586344361305,
          "validation_loss": 0.04595014825463295,
          "train_metric": 0.005554089788347483,
          "validation_metric": 0.00462793093174696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09510289299942087
        },
        {
          "epoch": 4,
          "updates": 256,
          "samples_seen": 4096,
          "online_loss": 0.042041635198984295,
          "train_loss": 0.03377453610301018,
          "validation_loss": 0.02893630415201187,
          "train_metric": 0.0034016475547105074,
          "validation_metric": 0.002914358163252473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12525590899895178
        },
        {
          "epoch": 5,
          "updates": 320,
          "samples_seen": 5120,
          "online_loss": 0.042440216333488934,
          "train_loss": 0.034161485731601715,
          "validation_loss": 0.0324280671775341,
          "train_metric": 0.003440620144829154,
          "validation_metric": 0.0032660362776368856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15688055899954634
        },
        {
          "epoch": 6,
          "updates": 384,
          "samples_seen": 6144,
          "online_loss": 0.03160349497193238,
          "train_loss": 0.024577395990490913,
          "validation_loss": 0.020045440644025803,
          "train_metric": 0.00247534504160285,
          "validation_metric": 0.0020189033821225166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2135267390003719
        },
        {
          "epoch": 7,
          "updates": 448,
          "samples_seen": 7168,
          "online_loss": 0.024795499339234084,
          "train_loss": 0.0337798036634922,
          "validation_loss": 0.025481587275862694,
          "train_metric": 0.003402178408578038,
          "validation_metric": 0.002566412091255188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25515793200065673
        },
        {
          "epoch": 8,
          "updates": 512,
          "samples_seen": 8192,
          "online_loss": 0.02799120203962957,
          "train_loss": 0.021791907027363777,
          "validation_loss": 0.015822535380721092,
          "train_metric": 0.002194801578298211,
          "validation_metric": 0.0015935878036543727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29832688699934806
        },
        {
          "epoch": 9,
          "updates": 576,
          "samples_seen": 9216,
          "online_loss": 0.02110055333650962,
          "train_loss": 0.016142243519425392,
          "validation_loss": 0.012869763188064098,
          "train_metric": 0.00162578746676445,
          "validation_metric": 0.0012961954344063997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4377293819998158
        },
        {
          "epoch": 10,
          "updates": 640,
          "samples_seen": 10240,
          "online_loss": 0.020033576111018192,
          "train_loss": 0.015082152560353279,
          "validation_loss": 0.012922706082463264,
          "train_metric": 0.0015190191334113479,
          "validation_metric": 0.0013015277218073606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4817478539989679
        },
        {
          "epoch": 11,
          "updates": 704,
          "samples_seen": 11264,
          "online_loss": 0.018872276048568892,
          "train_loss": 0.01877380535006523,
          "validation_loss": 0.016785724088549614,
          "train_metric": 0.0018908288329839706,
          "validation_metric": 0.0016905968077480793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5247709199993551
        },
        {
          "epoch": 12,
          "updates": 768,
          "samples_seen": 12288,
          "online_loss": 0.015564517714665271,
          "train_loss": 0.019443897530436516,
          "validation_loss": 0.016875337809324265,
          "train_metric": 0.001958318054676056,
          "validation_metric": 0.001699622254818678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5682341589999851
        },
        {
          "epoch": 13,
          "updates": 832,
          "samples_seen": 13312,
          "online_loss": 0.02062957177622593,
          "train_loss": 0.013845331966876984,
          "validation_loss": 0.011532033793628216,
          "train_metric": 0.001394451130181551,
          "validation_metric": 0.0011614641407504678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6119918759995926
        },
        {
          "epoch": 14,
          "updates": 896,
          "samples_seen": 14336,
          "online_loss": 0.014163261696012341,
          "train_loss": 0.013045270927250385,
          "validation_loss": 0.011753086000680923,
          "train_metric": 0.0013138719368726015,
          "validation_metric": 0.001183727872557938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6548827810001967
        },
        {
          "epoch": 15,
          "updates": 960,
          "samples_seen": 15360,
          "online_loss": 0.018057347393551026,
          "train_loss": 0.010781795717775822,
          "validation_loss": 0.00936091784387827,
          "train_metric": 0.0010859029134735465,
          "validation_metric": 0.0009427974000573158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6985740029995213
        },
        {
          "epoch": 16,
          "updates": 1024,
          "samples_seen": 16384,
          "online_loss": 0.01201313555611705,
          "train_loss": 0.010476985014975071,
          "validation_loss": 0.009610496461391449,
          "train_metric": 0.0010552034946158528,
          "validation_metric": 0.0009679339709691703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7411148440005491
        },
        {
          "epoch": 17,
          "updates": 1088,
          "samples_seen": 17408,
          "online_loss": 0.014973890263718204,
          "train_loss": 0.017886243760585785,
          "validation_loss": 0.013730121776461601,
          "train_metric": 0.0018014369998127222,
          "validation_metric": 0.001382847665809095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7841434509991814
        },
        {
          "epoch": 18,
          "updates": 1152,
          "samples_seen": 18432,
          "online_loss": 0.013085429758575629,
          "train_loss": 0.011683696880936623,
          "validation_loss": 0.009333208203315735,
          "train_metric": 0.0011767392279580235,
          "validation_metric": 0.0009400066337548196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8282827469993208
        },
        {
          "epoch": 19,
          "updates": 1216,
          "samples_seen": 19456,
          "online_loss": 0.017297398122536833,
          "train_loss": 0.02513301372528076,
          "validation_loss": 0.020134035497903824,
          "train_metric": 0.0025313054211437702,
          "validation_metric": 0.0020278263837099075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8719437639992975
        },
        {
          "epoch": 20,
          "updates": 1280,
          "samples_seen": 20480,
          "online_loss": 0.011894399105585762,
          "train_loss": 0.008214673027396202,
          "validation_loss": 0.007204909808933735,
          "train_metric": 0.0008273517014458776,
          "validation_metric": 0.000725652149412781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9402663099990605
        },
        {
          "epoch": 21,
          "updates": 1344,
          "samples_seen": 21504,
          "online_loss": 0.010655492063051497,
          "train_loss": 0.0107549037784338,
          "validation_loss": 0.010437926277518272,
          "train_metric": 0.0010831945110112429,
          "validation_metric": 0.001051269588060677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9988887649997196
        },
        {
          "epoch": 22,
          "updates": 1408,
          "samples_seen": 22528,
          "online_loss": 0.012774481894666678,
          "train_loss": 0.008350655436515808,
          "validation_loss": 0.006709737237542868,
          "train_metric": 0.0008410474401898682,
          "validation_metric": 0.0006757801165804267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0408161510003993
        },
        {
          "epoch": 23,
          "updates": 1472,
          "samples_seen": 23552,
          "online_loss": 0.010650439867276873,
          "train_loss": 0.007794754579663277,
          "validation_loss": 0.007431033067405224,
          "train_metric": 0.0007850591791793704,
          "validation_metric": 0.0007484264788217843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.072918486999697
        },
        {
          "epoch": 24,
          "updates": 1536,
          "samples_seen": 24576,
          "online_loss": 0.009276645722820831,
          "train_loss": 0.007059791125357151,
          "validation_loss": 0.006250465288758278,
          "train_metric": 0.0007110362639650702,
          "validation_metric": 0.000629523943644017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1036950809993868
        },
        {
          "epoch": 25,
          "updates": 1600,
          "samples_seen": 25600,
          "online_loss": 0.008429100917965116,
          "train_loss": 0.006332362070679665,
          "validation_loss": 0.005964370444417,
          "train_metric": 0.000637772202026099,
          "validation_metric": 0.000600709521677345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1351025250005478
        }
      ],
      "best": {
        "epoch": 25,
        "updates": 1600
      },
      "best_validation_metric": 0.000600709521677345,
      "config": {
        "label": "Batch 16",
        "batch": 16,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.024509376199270048
    },
    "50c1636b4f2c_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 4,
          "samples_seen": 1024,
          "online_loss": 0.9695521891117096,
          "train_loss": 0.9173716902732849,
          "validation_loss": 0.6648805141448975,
          "train_metric": 0.09239430725574493,
          "validation_metric": 0.06696432828903198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.009029807000842993
        },
        {
          "epoch": 2,
          "updates": 8,
          "samples_seen": 2048,
          "online_loss": 0.8723907023668289,
          "train_loss": 0.7863034009933472,
          "validation_loss": 0.5773224830627441,
          "train_metric": 0.0791936069726944,
          "validation_metric": 0.058145809918642044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.015735715000118944
        },
        {
          "epoch": 3,
          "updates": 12,
          "samples_seen": 3072,
          "online_loss": 0.7144555002450943,
          "train_loss": 0.5840524435043335,
          "validation_loss": 0.4395635724067688,
          "train_metric": 0.05882362648844719,
          "validation_metric": 0.04427122697234154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.022388667000996065
        },
        {
          "epoch": 4,
          "updates": 16,
          "samples_seen": 4096,
          "online_loss": 0.49066662043333054,
          "train_loss": 0.33232417702674866,
          "validation_loss": 0.2649034261703491,
          "train_metric": 0.03347047418355942,
          "validation_metric": 0.026680106297135353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02943988100014394
        },
        {
          "epoch": 5,
          "updates": 20,
          "samples_seen": 5120,
          "online_loss": 0.24944894015789032,
          "train_loss": 0.15672025084495544,
          "validation_loss": 0.1357194185256958,
          "train_metric": 0.015784289687871933,
          "validation_metric": 0.013669162057340145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03605274800065672
        },
        {
          "epoch": 6,
          "updates": 24,
          "samples_seen": 6144,
          "online_loss": 0.15909278020262718,
          "train_loss": 0.19022852182388306,
          "validation_loss": 0.15416836738586426,
          "train_metric": 0.019159115850925446,
          "validation_metric": 0.0155272725969553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1283208429995284
        },
        {
          "epoch": 7,
          "updates": 28,
          "samples_seen": 7168,
          "online_loss": 0.1935405544936657,
          "train_loss": 0.1695052683353424,
          "validation_loss": 0.14648866653442383,
          "train_metric": 0.017071951180696487,
          "validation_metric": 0.014753801748156548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1345089549995464
        },
        {
          "epoch": 8,
          "updates": 32,
          "samples_seen": 8192,
          "online_loss": 0.1470738984644413,
          "train_loss": 0.12593874335289001,
          "validation_loss": 0.10809101909399033,
          "train_metric": 0.012684090062975883,
          "validation_metric": 0.010886531323194504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14100107699960063
        },
        {
          "epoch": 9,
          "updates": 36,
          "samples_seen": 9216,
          "online_loss": 0.12766112200915813,
          "train_loss": 0.1277078092098236,
          "validation_loss": 0.1021735668182373,
          "train_metric": 0.01286226138472557,
          "validation_metric": 0.010290546342730522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14777703000072506
        },
        {
          "epoch": 10,
          "updates": 40,
          "samples_seen": 10240,
          "online_loss": 0.12615562416613102,
          "train_loss": 0.11690852791070938,
          "validation_loss": 0.09429637342691422,
          "train_metric": 0.011774598620831966,
          "validation_metric": 0.009497185237705708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1545122919997084
        },
        {
          "epoch": 11,
          "updates": 44,
          "samples_seen": 11264,
          "online_loss": 0.1099691316485405,
          "train_loss": 0.10117918998003006,
          "validation_loss": 0.08459017425775528,
          "train_metric": 0.010190397500991821,
          "validation_metric": 0.00851961225271225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1631495689998701
        },
        {
          "epoch": 12,
          "updates": 48,
          "samples_seen": 12288,
          "online_loss": 0.09843450225889683,
          "train_loss": 0.09595416486263275,
          "validation_loss": 0.0807890072464943,
          "train_metric": 0.009664151817560196,
          "validation_metric": 0.008136773481965065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16995889700046973
        },
        {
          "epoch": 13,
          "updates": 52,
          "samples_seen": 13312,
          "online_loss": 0.09469512104988098,
          "train_loss": 0.09012150764465332,
          "validation_loss": 0.07549820095300674,
          "train_metric": 0.009076708927750587,
          "validation_metric": 0.007603901904076338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1765341249993071
        },
        {
          "epoch": 14,
          "updates": 56,
          "samples_seen": 14336,
          "online_loss": 0.08635971695184708,
          "train_loss": 0.08211574703454971,
          "validation_loss": 0.06822966784238815,
          "train_metric": 0.008270397782325745,
          "validation_metric": 0.0068718427792191505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1832826270001533
        },
        {
          "epoch": 15,
          "updates": 60,
          "samples_seen": 15360,
          "online_loss": 0.08101723715662956,
          "train_loss": 0.0789894238114357,
          "validation_loss": 0.06436996161937714,
          "train_metric": 0.007955526001751423,
          "validation_metric": 0.006483107339590788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1899705569994694
        },
        {
          "epoch": 16,
          "updates": 64,
          "samples_seen": 16384,
          "online_loss": 0.07763338927179575,
          "train_loss": 0.0746479257941246,
          "validation_loss": 0.06035236641764641,
          "train_metric": 0.007518266327679157,
          "validation_metric": 0.006078468635678291,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30278368600011163
        },
        {
          "epoch": 17,
          "updates": 68,
          "samples_seen": 17408,
          "online_loss": 0.07265714928507805,
          "train_loss": 0.07003086805343628,
          "validation_loss": 0.05700680613517761,
          "train_metric": 0.007053252775222063,
          "validation_metric": 0.0057415165938436985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3098451229998318
        },
        {
          "epoch": 18,
          "updates": 72,
          "samples_seen": 18432,
          "online_loss": 0.06909829471260309,
          "train_loss": 0.06731940060853958,
          "validation_loss": 0.05514797940850258,
          "train_metric": 0.006780164781957865,
          "validation_metric": 0.005554302129894495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3166264139999839
        },
        {
          "epoch": 19,
          "updates": 76,
          "samples_seen": 19456,
          "online_loss": 0.06644832529127598,
          "train_loss": 0.06472302973270416,
          "validation_loss": 0.05297579616308212,
          "train_metric": 0.0065186661668121815,
          "validation_metric": 0.005335527937859297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32304359400040994
        },
        {
          "epoch": 20,
          "updates": 80,
          "samples_seen": 20480,
          "online_loss": 0.06388255581259727,
          "train_loss": 0.062328021973371506,
          "validation_loss": 0.05059530586004257,
          "train_metric": 0.006277450826019049,
          "validation_metric": 0.005095773842185736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32999448500049766
        },
        {
          "epoch": 21,
          "updates": 84,
          "samples_seen": 21504,
          "online_loss": 0.061680423095822334,
          "train_loss": 0.06072109192609787,
          "validation_loss": 0.04915696382522583,
          "train_metric": 0.006115606054663658,
          "validation_metric": 0.0049509089440107346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3367546890003723
        },
        {
          "epoch": 22,
          "updates": 88,
          "samples_seen": 22528,
          "online_loss": 0.06048145052045584,
          "train_loss": 0.05929984524846077,
          "validation_loss": 0.048177871853113174,
          "train_metric": 0.005972463171929121,
          "validation_metric": 0.0048522986471652985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3434503290009161
        },
        {
          "epoch": 23,
          "updates": 92,
          "samples_seen": 23552,
          "online_loss": 0.05906580202281475,
          "train_loss": 0.058192916214466095,
          "validation_loss": 0.047501735389232635,
          "train_metric": 0.005860976874828339,
          "validation_metric": 0.004784200806170702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35009064199948625
        },
        {
          "epoch": 24,
          "updates": 96,
          "samples_seen": 24576,
          "online_loss": 0.057800996117293835,
          "train_loss": 0.05694599077105522,
          "validation_loss": 0.046733271330595016,
          "train_metric": 0.00573539175093174,
          "validation_metric": 0.004706804174929857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35661931300091965
        },
        {
          "epoch": 25,
          "updates": 100,
          "samples_seen": 25600,
          "online_loss": 0.05661121942102909,
          "train_loss": 0.05580829083919525,
          "validation_loss": 0.04574165120720863,
          "train_metric": 0.005620806477963924,
          "validation_metric": 0.00460693147033453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36551907700049924
        },
        {
          "epoch": 26,
          "updates": 104,
          "samples_seen": 26624,
          "online_loss": 0.055486380122601986,
          "train_loss": 0.05477623641490936,
          "validation_loss": 0.04496937245130539,
          "train_metric": 0.005516860634088516,
          "validation_metric": 0.004529151134192944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3724071720007487
        },
        {
          "epoch": 27,
          "updates": 108,
          "samples_seen": 27648,
          "online_loss": 0.05456340778619051,
          "train_loss": 0.053731292486190796,
          "validation_loss": 0.04408695176243782,
          "train_metric": 0.005411618854850531,
          "validation_metric": 0.004440275486558676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.378993280000941
        },
        {
          "epoch": 28,
          "updates": 112,
          "samples_seen": 28672,
          "online_loss": 0.05335362907499075,
          "train_loss": 0.05266563594341278,
          "validation_loss": 0.04318744316697121,
          "train_metric": 0.005304289516061544,
          "validation_metric": 0.004349681083112955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38559297699976014
        },
        {
          "epoch": 29,
          "updates": 116,
          "samples_seen": 29696,
          "online_loss": 0.05235619191080332,
          "train_loss": 0.05151109769940376,
          "validation_loss": 0.04232766479253769,
          "train_metric": 0.005188009236007929,
          "validation_metric": 0.0042630876414477825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3920754249993479
        },
        {
          "epoch": 30,
          "updates": 120,
          "samples_seen": 30720,
          "online_loss": 0.05124880000948906,
          "train_loss": 0.050364576280117035,
          "validation_loss": 0.04151394963264465,
          "train_metric": 0.00507253548130393,
          "validation_metric": 0.00418113311752677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39867426499949943
        },
        {
          "epoch": 31,
          "updates": 124,
          "samples_seen": 31744,
          "online_loss": 0.04998783953487873,
          "train_loss": 0.04919150471687317,
          "validation_loss": 0.0404948964715004,
          "train_metric": 0.004954388365149498,
          "validation_metric": 0.0040784976445138454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40517432799970265
        },
        {
          "epoch": 32,
          "updates": 128,
          "samples_seen": 32768,
          "online_loss": 0.0492028808221221,
          "train_loss": 0.047964636236429214,
          "validation_loss": 0.03943555802106857,
          "train_metric": 0.0048308223485946655,
          "validation_metric": 0.003971804864704609,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4118433149997145
        },
        {
          "epoch": 33,
          "updates": 132,
          "samples_seen": 33792,
          "online_loss": 0.048264019191265106,
          "train_loss": 0.04685889184474945,
          "validation_loss": 0.03867345675826073,
          "train_metric": 0.004719455726444721,
          "validation_metric": 0.0038950492162257433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.418447443000332
        },
        {
          "epoch": 34,
          "updates": 136,
          "samples_seen": 34816,
          "online_loss": 0.04776485823094845,
          "train_loss": 0.04628922790288925,
          "validation_loss": 0.03795287013053894,
          "train_metric": 0.004662081599235535,
          "validation_metric": 0.003822474041953683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4250718820003385
        },
        {
          "epoch": 35,
          "updates": 140,
          "samples_seen": 35840,
          "online_loss": 0.045267921872437,
          "train_loss": 0.044987037777900696,
          "validation_loss": 0.037282414734363556,
          "train_metric": 0.004530929960310459,
          "validation_metric": 0.0037549484986811876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4316109739993408
        },
        {
          "epoch": 36,
          "updates": 144,
          "samples_seen": 36864,
          "online_loss": 0.04460034519433975,
          "train_loss": 0.043151095509529114,
          "validation_loss": 0.035498809069395065,
          "train_metric": 0.004346020054072142,
          "validation_metric": 0.0035753105767071247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4382433539994963
        },
        {
          "epoch": 37,
          "updates": 148,
          "samples_seen": 37888,
          "online_loss": 0.0428842231631279,
          "train_loss": 0.042120013386011124,
          "validation_loss": 0.03461400046944618,
          "train_metric": 0.00424217339605093,
          "validation_metric": 0.0034861951135098934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44503081500079134
        },
        {
          "epoch": 38,
          "updates": 152,
          "samples_seen": 38912,
          "online_loss": 0.04174661450088024,
          "train_loss": 0.040740713477134705,
          "validation_loss": 0.033530350774526596,
          "train_metric": 0.004103255458176136,
          "validation_metric": 0.0033770541194826365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4519444800007477
        },
        {
          "epoch": 39,
          "updates": 156,
          "samples_seen": 39936,
          "online_loss": 0.04087015148252249,
          "train_loss": 0.039552852511405945,
          "validation_loss": 0.03259068354964256,
          "train_metric": 0.003983618225902319,
          "validation_metric": 0.003282414050772786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45872571200015955
        },
        {
          "epoch": 40,
          "updates": 160,
          "samples_seen": 40960,
          "online_loss": 0.04108734894543886,
          "train_loss": 0.038540877401828766,
          "validation_loss": 0.03184613585472107,
          "train_metric": 0.00388169614598155,
          "validation_metric": 0.0032074262853711843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46538355499978934
        },
        {
          "epoch": 41,
          "updates": 164,
          "samples_seen": 41984,
          "online_loss": 0.03844865132123232,
          "train_loss": 0.03826240822672844,
          "validation_loss": 0.031497202813625336,
          "train_metric": 0.0038536495994776487,
          "validation_metric": 0.0031722825951874256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47201051199954236
        },
        {
          "epoch": 42,
          "updates": 168,
          "samples_seen": 43008,
          "online_loss": 0.03712467337027192,
          "train_loss": 0.03725095093250275,
          "validation_loss": 0.030861297622323036,
          "train_metric": 0.003751779440790415,
          "validation_metric": 0.0031082367058843374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47872224700040533
        },
        {
          "epoch": 43,
          "updates": 172,
          "samples_seen": 44032,
          "online_loss": 0.03658718150109053,
          "train_loss": 0.035190731287002563,
          "validation_loss": 0.02905161865055561,
          "train_metric": 0.0035442812368273735,
          "validation_metric": 0.002925972919911146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4854623459996219
        },
        {
          "epoch": 44,
          "updates": 176,
          "samples_seen": 45056,
          "online_loss": 0.03566525178030133,
          "train_loss": 0.034018006175756454,
          "validation_loss": 0.028033707290887833,
          "train_metric": 0.003426168579608202,
          "validation_metric": 0.0028234522324055433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4921435409996775
        },
        {
          "epoch": 45,
          "updates": 180,
          "samples_seen": 46080,
          "online_loss": 0.03420545067638159,
          "train_loss": 0.03323092311620712,
          "validation_loss": 0.02727651037275791,
          "train_metric": 0.003346896730363369,
          "validation_metric": 0.0027471897192299366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49888798800020595
        },
        {
          "epoch": 46,
          "updates": 184,
          "samples_seen": 47104,
          "online_loss": 0.03279589721933007,
          "train_loss": 0.032634325325489044,
          "validation_loss": 0.02701699361205101,
          "train_metric": 0.003286809893324971,
          "validation_metric": 0.002721052849665284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.505605172000287
        },
        {
          "epoch": 47,
          "updates": 188,
          "samples_seen": 48128,
          "online_loss": 0.03221900062635541,
          "train_loss": 0.03159918263554573,
          "validation_loss": 0.025981007143855095,
          "train_metric": 0.0031825541518628597,
          "validation_metric": 0.0026167118921875954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5123018770009367
        },
        {
          "epoch": 48,
          "updates": 192,
          "samples_seen": 49152,
          "online_loss": 0.03275986993685365,
          "train_loss": 0.030187062919139862,
          "validation_loss": 0.02475796267390251,
          "train_metric": 0.003040330484509468,
          "validation_metric": 0.0024935309775173664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5190623510006844
        },
        {
          "epoch": 49,
          "updates": 196,
          "samples_seen": 50176,
          "online_loss": 0.029504945036023855,
          "train_loss": 0.0307473074644804,
          "validation_loss": 0.025682035833597183,
          "train_metric": 0.0030967560596764088,
          "validation_metric": 0.0025866006035357714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5251825089999329
        },
        {
          "epoch": 50,
          "updates": 200,
          "samples_seen": 51200,
          "online_loss": 0.030057819094508886,
          "train_loss": 0.028973709791898727,
          "validation_loss": 0.023896589875221252,
          "train_metric": 0.0029181260615587234,
          "validation_metric": 0.0024067768827080727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5316946879993338
        },
        {
          "epoch": 51,
          "updates": 204,
          "samples_seen": 52224,
          "online_loss": 0.02851517451927066,
          "train_loss": 0.027704326435923576,
          "validation_loss": 0.02292887680232525,
          "train_metric": 0.00279027852229774,
          "validation_metric": 0.002309312578290701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.538293359999443
        },
        {
          "epoch": 52,
          "updates": 208,
          "samples_seen": 53248,
          "online_loss": 0.02883159089833498,
          "train_loss": 0.026960067451000214,
          "validation_loss": 0.02249200828373432,
          "train_metric": 0.0027153196278959513,
          "validation_metric": 0.00226531270891428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5447949299996253
        },
        {
          "epoch": 53,
          "updates": 212,
          "samples_seen": 54272,
          "online_loss": 0.027510047890245914,
          "train_loss": 0.026442058384418488,
          "validation_loss": 0.02169691026210785,
          "train_metric": 0.0026631474029272795,
          "validation_metric": 0.0021852334029972553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.551730311000938
        },
        {
          "epoch": 54,
          "updates": 216,
          "samples_seen": 55296,
          "online_loss": 0.02592663560062647,
          "train_loss": 0.02566816471517086,
          "validation_loss": 0.021415401250123978,
          "train_metric": 0.002585203852504492,
          "validation_metric": 0.002156880684196949,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5584312999999383
        },
        {
          "epoch": 55,
          "updates": 220,
          "samples_seen": 56320,
          "online_loss": 0.02558499900624156,
          "train_loss": 0.024860801175236702,
          "validation_loss": 0.020311016589403152,
          "train_metric": 0.002503888914361596,
          "validation_metric": 0.002045651199296117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5648817460005375
        },
        {
          "epoch": 56,
          "updates": 224,
          "samples_seen": 57344,
          "online_loss": 0.024958156514912844,
          "train_loss": 0.024092257022857666,
          "validation_loss": 0.020124902948737144,
          "train_metric": 0.00242648390121758,
          "validation_metric": 0.002026906469836831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5718394150007953
        },
        {
          "epoch": 57,
          "updates": 228,
          "samples_seen": 58368,
          "online_loss": 0.024232062976807356,
          "train_loss": 0.02345411665737629,
          "validation_loss": 0.01936262845993042,
          "train_metric": 0.0023622126318514347,
          "validation_metric": 0.001950133009813726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5785918990004575
        },
        {
          "epoch": 58,
          "updates": 232,
          "samples_seen": 59392,
          "online_loss": 0.023910372518002987,
          "train_loss": 0.022861018776893616,
          "validation_loss": 0.019169775769114494,
          "train_metric": 0.0023024778347462416,
          "validation_metric": 0.0019307095790281892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6835653270009061
        },
        {
          "epoch": 59,
          "updates": 236,
          "samples_seen": 60416,
          "online_loss": 0.023165713995695114,
          "train_loss": 0.022309254854917526,
          "validation_loss": 0.018551334738731384,
          "train_metric": 0.002246906515210867,
          "validation_metric": 0.0018684226088225842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6914396829997713
        },
        {
          "epoch": 60,
          "updates": 240,
          "samples_seen": 61440,
          "online_loss": 0.02281543775461614,
          "train_loss": 0.02167218178510666,
          "validation_loss": 0.017990531399846077,
          "train_metric": 0.0021827425807714462,
          "validation_metric": 0.0018119403393939137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6989106149994768
        },
        {
          "epoch": 61,
          "updates": 244,
          "samples_seen": 62464,
          "online_loss": 0.021899739280343056,
          "train_loss": 0.02113504521548748,
          "validation_loss": 0.017603978514671326,
          "train_metric": 0.002128644147887826,
          "validation_metric": 0.0017730083782225847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7069132150008954
        },
        {
          "epoch": 62,
          "updates": 248,
          "samples_seen": 63488,
          "online_loss": 0.022178709972649813,
          "train_loss": 0.020654551684856415,
          "validation_loss": 0.01726728491485119,
          "train_metric": 0.0020802507642656565,
          "validation_metric": 0.001739097642712295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7144282850003947
        },
        {
          "epoch": 63,
          "updates": 252,
          "samples_seen": 64512,
          "online_loss": 0.020945305470377207,
          "train_loss": 0.020133567973971367,
          "validation_loss": 0.017017504200339317,
          "train_metric": 0.002027778886258602,
          "validation_metric": 0.00171394064091146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7215505080002913
        },
        {
          "epoch": 64,
          "updates": 256,
          "samples_seen": 65536,
          "online_loss": 0.020474074874073267,
          "train_loss": 0.019680650904774666,
          "validation_loss": 0.016311468556523323,
          "train_metric": 0.0019821629393845797,
          "validation_metric": 0.0016428313683718443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7285384669994528
        },
        {
          "epoch": 65,
          "updates": 260,
          "samples_seen": 66560,
          "online_loss": 0.020504547050222754,
          "train_loss": 0.01933443918824196,
          "validation_loss": 0.016530554741621017,
          "train_metric": 0.0019472941057756543,
          "validation_metric": 0.001664897077716887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7349931560002005
        },
        {
          "epoch": 66,
          "updates": 264,
          "samples_seen": 67584,
          "online_loss": 0.02076398255303502,
          "train_loss": 0.01879744417965412,
          "validation_loss": 0.015985717996954918,
          "train_metric": 0.0018932097591459751,
          "validation_metric": 0.0016100232023745775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7419222959997569
        },
        {
          "epoch": 67,
          "updates": 268,
          "samples_seen": 68608,
          "online_loss": 0.01976860547438264,
          "train_loss": 0.018436191603541374,
          "validation_loss": 0.01529238186776638,
          "train_metric": 0.0018568260129541159,
          "validation_metric": 0.0015401925193145871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7492000029997143
        },
        {
          "epoch": 68,
          "updates": 272,
          "samples_seen": 69632,
          "online_loss": 0.02113993465900421,
          "train_loss": 0.018009228631854057,
          "validation_loss": 0.015355906449258327,
          "train_metric": 0.001813823590055108,
          "validation_metric": 0.0015465907054021955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.755731380000725
        },
        {
          "epoch": 69,
          "updates": 276,
          "samples_seen": 70656,
          "online_loss": 0.022903192322701216,
          "train_loss": 0.017667675390839577,
          "validation_loss": 0.014713720418512821,
          "train_metric": 0.0017794236773625016,
          "validation_metric": 0.0014819120988249779,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7631075860008423
        },
        {
          "epoch": 70,
          "updates": 280,
          "samples_seen": 71680,
          "online_loss": 0.019743677927181125,
          "train_loss": 0.017672378569841385,
          "validation_loss": 0.015312885865569115,
          "train_metric": 0.0017798971384763718,
          "validation_metric": 0.0015422576107084751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7699585549999028
        },
        {
          "epoch": 71,
          "updates": 284,
          "samples_seen": 72704,
          "online_loss": 0.018245853716507554,
          "train_loss": 0.01732063852250576,
          "validation_loss": 0.014257293194532394,
          "train_metric": 0.0017444712575525045,
          "validation_metric": 0.0014359424822032452,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7771159530002478
        },
        {
          "epoch": 72,
          "updates": 288,
          "samples_seen": 73728,
          "online_loss": 0.01722483243793249,
          "train_loss": 0.017125360667705536,
          "validation_loss": 0.01515265740454197,
          "train_metric": 0.0017248035874217749,
          "validation_metric": 0.001526120351627469,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7843157459992653
        },
        {
          "epoch": 73,
          "updates": 292,
          "samples_seen": 74752,
          "online_loss": 0.0176298045553267,
          "train_loss": 0.01640290953218937,
          "validation_loss": 0.013957659713923931,
          "train_metric": 0.001652040984481573,
          "validation_metric": 0.0014057643711566925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7914888190007332
        },
        {
          "epoch": 74,
          "updates": 296,
          "samples_seen": 75776,
          "online_loss": 0.016832913737744093,
          "train_loss": 0.01607702486217022,
          "validation_loss": 0.013766039162874222,
          "train_metric": 0.0016192190814763308,
          "validation_metric": 0.0013864650391042233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7988260130005074
        },
        {
          "epoch": 75,
          "updates": 300,
          "samples_seen": 76800,
          "online_loss": 0.016683853464201093,
          "train_loss": 0.0158058051019907,
          "validation_loss": 0.013238886371254921,
          "train_metric": 0.001591902575455606,
          "validation_metric": 0.001333372201770544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.806013666000581
        },
        {
          "epoch": 76,
          "updates": 304,
          "samples_seen": 77824,
          "online_loss": 0.016866830876097083,
          "train_loss": 0.015543053857982159,
          "validation_loss": 0.013281912542879581,
          "train_metric": 0.0015654393937438726,
          "validation_metric": 0.001337705529294908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8132698930003244
        },
        {
          "epoch": 77,
          "updates": 308,
          "samples_seen": 78848,
          "online_loss": 0.01660609431564808,
          "train_loss": 0.015267123468220234,
          "validation_loss": 0.012827575206756592,
          "train_metric": 0.0015376490773633122,
          "validation_metric": 0.0012919465079903603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.820893124999202
        },
        {
          "epoch": 78,
          "updates": 312,
          "samples_seen": 79872,
          "online_loss": 0.016127439448609948,
          "train_loss": 0.01502118818461895,
          "validation_loss": 0.012605500407516956,
          "train_metric": 0.0015128790400922298,
          "validation_metric": 0.001269579748623073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8285357160002604
        },
        {
          "epoch": 79,
          "updates": 316,
          "samples_seen": 80896,
          "online_loss": 0.015120953554287553,
          "train_loss": 0.014860540628433228,
          "validation_loss": 0.01254187896847725,
          "train_metric": 0.0014966991730034351,
          "validation_metric": 0.0012631722493097186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8355570290004835
        },
        {
          "epoch": 80,
          "updates": 320,
          "samples_seen": 81920,
          "online_loss": 0.015192954801023006,
          "train_loss": 0.014747895300388336,
          "validation_loss": 0.012671551667153835,
          "train_metric": 0.0014853539178147912,
          "validation_metric": 0.0012762323021888733,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8423442109997268
        },
        {
          "epoch": 81,
          "updates": 324,
          "samples_seen": 82944,
          "online_loss": 0.01466759410686791,
          "train_loss": 0.01434900239109993,
          "validation_loss": 0.012301871553063393,
          "train_metric": 0.0014451788738369942,
          "validation_metric": 0.0012389994226396084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8496456169996236
        },
        {
          "epoch": 82,
          "updates": 328,
          "samples_seen": 83968,
          "online_loss": 0.014473747927695513,
          "train_loss": 0.014180472120642662,
          "validation_loss": 0.012345113791525364,
          "train_metric": 0.0014282054034993052,
          "validation_metric": 0.0012433546362444758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8560467349998362
        },
        {
          "epoch": 83,
          "updates": 332,
          "samples_seen": 84992,
          "online_loss": 0.014322928618639708,
          "train_loss": 0.014054995961487293,
          "validation_loss": 0.0114727932959795,
          "train_metric": 0.0014155677054077387,
          "validation_metric": 0.0011554976226761937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8633800009993138
        },
        {
          "epoch": 84,
          "updates": 336,
          "samples_seen": 86016,
          "online_loss": 0.014517622534185648,
          "train_loss": 0.013718617148697376,
          "validation_loss": 0.011602189391851425,
          "train_metric": 0.0013816885184496641,
          "validation_metric": 0.0011685299687087536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8699736409998877
        },
        {
          "epoch": 85,
          "updates": 340,
          "samples_seen": 87040,
          "online_loss": 0.013837833423167467,
          "train_loss": 0.013487200252711773,
          "validation_loss": 0.011256178840994835,
          "train_metric": 0.0013583814725279808,
          "validation_metric": 0.0011336810421198606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8791587300001993
        },
        {
          "epoch": 86,
          "updates": 344,
          "samples_seen": 88064,
          "online_loss": 0.013746960321441293,
          "train_loss": 0.013683685101568699,
          "validation_loss": 0.01209786906838417,
          "train_metric": 0.0013781706802546978,
          "validation_metric": 0.001218453049659729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8868770600001881
        },
        {
          "epoch": 87,
          "updates": 348,
          "samples_seen": 89088,
          "online_loss": 0.013669044943526387,
          "train_loss": 0.013604950159788132,
          "validation_loss": 0.011093505658209324,
          "train_metric": 0.0013702407013624907,
          "validation_metric": 0.001117297331802547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8960743679999723
        },
        {
          "epoch": 88,
          "updates": 352,
          "samples_seen": 90112,
          "online_loss": 0.013428433099761605,
          "train_loss": 0.013393850065767765,
          "validation_loss": 0.011783355847001076,
          "train_metric": 0.001348979421891272,
          "validation_metric": 0.0011867765570059419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.903822565000155
        },
        {
          "epoch": 89,
          "updates": 356,
          "samples_seen": 91136,
          "online_loss": 0.013223708141595125,
          "train_loss": 0.012992444448173046,
          "validation_loss": 0.010563787072896957,
          "train_metric": 0.0013085512910038233,
          "validation_metric": 0.0010639459360390902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9128086990003794
        },
        {
          "epoch": 90,
          "updates": 360,
          "samples_seen": 92160,
          "online_loss": 0.012917972402647138,
          "train_loss": 0.01319979876279831,
          "validation_loss": 0.011923311278223991,
          "train_metric": 0.0013294352684170008,
          "validation_metric": 0.001200872240588069,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9196815819996118
        },
        {
          "epoch": 91,
          "updates": 364,
          "samples_seen": 93184,
          "online_loss": 0.013414895627647638,
          "train_loss": 0.012353373691439629,
          "validation_loss": 0.010234920307993889,
          "train_metric": 0.0012441865401342511,
          "validation_metric": 0.001030823914334178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9265620389996911
        },
        {
          "epoch": 92,
          "updates": 368,
          "samples_seen": 94208,
          "online_loss": 0.012525457888841629,
          "train_loss": 0.012142025865614414,
          "validation_loss": 0.01036704983562231,
          "train_metric": 0.0012229005806148052,
          "validation_metric": 0.0010441313497722149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9340399030006665
        },
        {
          "epoch": 93,
          "updates": 372,
          "samples_seen": 95232,
          "online_loss": 0.0124216980766505,
          "train_loss": 0.012619621120393276,
          "validation_loss": 0.010093805380165577,
          "train_metric": 0.0012710019946098328,
          "validation_metric": 0.001016611116938293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9411628550005844
        },
        {
          "epoch": 94,
          "updates": 376,
          "samples_seen": 96256,
          "online_loss": 0.012835508212447166,
          "train_loss": 0.012466749176383018,
          "validation_loss": 0.010997017845511436,
          "train_metric": 0.0012556052533909678,
          "validation_metric": 0.0011075793299824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9480103730002156
        },
        {
          "epoch": 95,
          "updates": 380,
          "samples_seen": 97280,
          "online_loss": 0.013087387196719646,
          "train_loss": 0.011874973773956299,
          "validation_loss": 0.009670274332165718,
          "train_metric": 0.0011960039846599102,
          "validation_metric": 0.000973954563960433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9550603269999556
        },
        {
          "epoch": 96,
          "updates": 384,
          "samples_seen": 98304,
          "online_loss": 0.012287742458283901,
          "train_loss": 0.011500914581120014,
          "validation_loss": 0.00986067857593298,
          "train_metric": 0.001158330007456243,
          "validation_metric": 0.0009931314270943403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9616832219999196
        },
        {
          "epoch": 97,
          "updates": 388,
          "samples_seen": 99328,
          "online_loss": 0.01203314051963389,
          "train_loss": 0.011352934874594212,
          "validation_loss": 0.0093900877982378,
          "train_metric": 0.0011434260522946715,
          "validation_metric": 0.0009457351407036185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9693535059996066
        },
        {
          "epoch": 98,
          "updates": 392,
          "samples_seen": 100352,
          "online_loss": 0.01201162685174495,
          "train_loss": 0.011825153604149818,
          "validation_loss": 0.010699083097279072,
          "train_metric": 0.0011909861350432038,
          "validation_metric": 0.0010775724658742547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9758596920000855
        },
        {
          "epoch": 99,
          "updates": 396,
          "samples_seen": 101376,
          "online_loss": 0.011701033916324377,
          "train_loss": 0.01132157351821661,
          "validation_loss": 0.00894322618842125,
          "train_metric": 0.0011402673553675413,
          "validation_metric": 0.0009007292101159692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9842764179993537
        },
        {
          "epoch": 100,
          "updates": 400,
          "samples_seen": 102400,
          "online_loss": 0.011247035348787904,
          "train_loss": 0.01138476375490427,
          "validation_loss": 0.01003410667181015,
          "train_metric": 0.001146631664596498,
          "validation_metric": 0.0010105984983965755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9913610420007899
        },
        {
          "epoch": 101,
          "updates": 404,
          "samples_seen": 103424,
          "online_loss": 0.01130581577308476,
          "train_loss": 0.010838943533599377,
          "validation_loss": 0.009051035158336163,
          "train_metric": 0.0010916587198153138,
          "validation_metric": 0.0009115870925597847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9977307579993067
        },
        {
          "epoch": 102,
          "updates": 408,
          "samples_seen": 104448,
          "online_loss": 0.01140603213571012,
          "train_loss": 0.010961528867483139,
          "validation_loss": 0.0085904560983181,
          "train_metric": 0.0011040051467716694,
          "validation_metric": 0.0008651993121020496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.005226795999988
        },
        {
          "epoch": 103,
          "updates": 412,
          "samples_seen": 105472,
          "online_loss": 0.011600629892200232,
          "train_loss": 0.010603735223412514,
          "validation_loss": 0.009279596619307995,
          "train_metric": 0.0010679694823920727,
          "validation_metric": 0.00093460688367486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0126863870009402
        },
        {
          "epoch": 104,
          "updates": 416,
          "samples_seen": 106496,
          "online_loss": 0.01060865749605,
          "train_loss": 0.01039145514369011,
          "validation_loss": 0.008259860798716545,
          "train_metric": 0.0010465893428772688,
          "validation_metric": 0.0008319029002450407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0196521430007124
        },
        {
          "epoch": 105,
          "updates": 420,
          "samples_seen": 107520,
          "online_loss": 0.010410238523036242,
          "train_loss": 0.010137380100786686,
          "validation_loss": 0.008514702320098877,
          "train_metric": 0.0010209998581558466,
          "validation_metric": 0.000857569626532495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0262481720001233
        },
        {
          "epoch": 106,
          "updates": 424,
          "samples_seen": 108544,
          "online_loss": 0.010464722523465753,
          "train_loss": 0.010248608887195587,
          "validation_loss": 0.008274774998426437,
          "train_metric": 0.0010322022717446089,
          "validation_metric": 0.0008334050653502345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0334149549998983
        },
        {
          "epoch": 107,
          "updates": 428,
          "samples_seen": 109568,
          "online_loss": 0.010266757803037763,
          "train_loss": 0.009881466627120972,
          "validation_loss": 0.008276029489934444,
          "train_metric": 0.0009952251566573977,
          "validation_metric": 0.0008335314341820776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0396217479992629
        },
        {
          "epoch": 108,
          "updates": 432,
          "samples_seen": 110592,
          "online_loss": 0.009944618213921785,
          "train_loss": 0.009757714346051216,
          "validation_loss": 0.007951238192617893,
          "train_metric": 0.0009827612666413188,
          "validation_metric": 0.0008008196018636227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0466153910001594
        },
        {
          "epoch": 109,
          "updates": 436,
          "samples_seen": 111616,
          "online_loss": 0.009699324844405055,
          "train_loss": 0.00955162476748228,
          "validation_loss": 0.007995803840458393,
          "train_metric": 0.00096200464759022,
          "validation_metric": 0.0008053079945966601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.053956900999765
        },
        {
          "epoch": 110,
          "updates": 440,
          "samples_seen": 112640,
          "online_loss": 0.00968865875620395,
          "train_loss": 0.009417249821126461,
          "validation_loss": 0.007677790243178606,
          "train_metric": 0.0009484710171818733,
          "validation_metric": 0.000773278996348381,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0613350750008976
        },
        {
          "epoch": 111,
          "updates": 444,
          "samples_seen": 113664,
          "online_loss": 0.009575624950230122,
          "train_loss": 0.009352347813546658,
          "validation_loss": 0.008003417402505875,
          "train_metric": 0.0009419342968612909,
          "validation_metric": 0.0008060747641138732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0680039170001692
        },
        {
          "epoch": 112,
          "updates": 448,
          "samples_seen": 114688,
          "online_loss": 0.009292406146414578,
          "train_loss": 0.009221491403877735,
          "validation_loss": 0.007491860073059797,
          "train_metric": 0.0009287547436542809,
          "validation_metric": 0.000754552660509944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0748714620003739
        },
        {
          "epoch": 113,
          "updates": 452,
          "samples_seen": 115712,
          "online_loss": 0.009847739362157881,
          "train_loss": 0.009966586716473103,
          "validation_loss": 0.008812771178781986,
          "train_metric": 0.001003798097372055,
          "validation_metric": 0.0008875900530256331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0815071369997895
        },
        {
          "epoch": 114,
          "updates": 456,
          "samples_seen": 116736,
          "online_loss": 0.009859348647296429,
          "train_loss": 0.009298722259700298,
          "validation_loss": 0.007390448823571205,
          "train_metric": 0.0009365331497974694,
          "validation_metric": 0.0007443387876264751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0886810080010036
        },
        {
          "epoch": 115,
          "updates": 460,
          "samples_seen": 117760,
          "online_loss": 0.009349630447104573,
          "train_loss": 0.008892081677913666,
          "validation_loss": 0.007032107561826706,
          "train_metric": 0.0008955778903327882,
          "validation_metric": 0.0007082481170073152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0958956520007632
        },
        {
          "epoch": 116,
          "updates": 464,
          "samples_seen": 118784,
          "online_loss": 0.009071881417185068,
          "train_loss": 0.008925252594053745,
          "validation_loss": 0.007705256342887878,
          "train_metric": 0.0008989186608232558,
          "validation_metric": 0.0007760451990179718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1024144090006303
        },
        {
          "epoch": 117,
          "updates": 468,
          "samples_seen": 119808,
          "online_loss": 0.009472119039855897,
          "train_loss": 0.00940267276018858,
          "validation_loss": 0.007396594155579805,
          "train_metric": 0.0009470026707276702,
          "validation_metric": 0.0007449578261002898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1084944150006777
        },
        {
          "epoch": 118,
          "updates": 472,
          "samples_seen": 120832,
          "online_loss": 0.009939232841134071,
          "train_loss": 0.009427398443222046,
          "validation_loss": 0.008199220523238182,
          "train_metric": 0.0009494931437075138,
          "validation_metric": 0.0008257954032160342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.115075412000806
        },
        {
          "epoch": 119,
          "updates": 476,
          "samples_seen": 121856,
          "online_loss": 0.009359046583995223,
          "train_loss": 0.0107400082051754,
          "validation_loss": 0.008174712769687176,
          "train_metric": 0.0010816942667588592,
          "validation_metric": 0.0008233272237703204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1211943460002658
        },
        {
          "epoch": 120,
          "updates": 480,
          "samples_seen": 122880,
          "online_loss": 0.009774971404112875,
          "train_loss": 0.011664396151900291,
          "validation_loss": 0.010700693354010582,
          "train_metric": 0.0011747952084988356,
          "validation_metric": 0.0010777345160022378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.12745998399987
        },
        {
          "epoch": 121,
          "updates": 484,
          "samples_seen": 123904,
          "online_loss": 0.01115534850396216,
          "train_loss": 0.012576436623930931,
          "validation_loss": 0.009786143898963928,
          "train_metric": 0.0012666527181863785,
          "validation_metric": 0.000985624617896974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1344996219995664
        },
        {
          "epoch": 122,
          "updates": 488,
          "samples_seen": 124928,
          "online_loss": 0.010408194968476892,
          "train_loss": 0.010936315171420574,
          "validation_loss": 0.010021894238889217,
          "train_metric": 0.0011014656629413366,
          "validation_metric": 0.0010093685705214739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1407502469992323
        },
        {
          "epoch": 123,
          "updates": 492,
          "samples_seen": 125952,
          "online_loss": 0.009280517231673002,
          "train_loss": 0.01158461719751358,
          "validation_loss": 0.008798210881650448,
          "train_metric": 0.0011667601065710187,
          "validation_metric": 0.0008861236274242401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1472102480001922
        },
        {
          "epoch": 124,
          "updates": 496,
          "samples_seen": 126976,
          "online_loss": 0.010858291178010404,
          "train_loss": 0.008831561543047428,
          "validation_loss": 0.007738890126347542,
          "train_metric": 0.0008894825004972517,
          "validation_metric": 0.0007794326520524919,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1537770150007418
        },
        {
          "epoch": 125,
          "updates": 500,
          "samples_seen": 128000,
          "online_loss": 0.011300774291157722,
          "train_loss": 0.007982473820447922,
          "validation_loss": 0.006062340456992388,
          "train_metric": 0.0008039655513130128,
          "validation_metric": 0.0006105767097324133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.160679237000295
        },
        {
          "epoch": 126,
          "updates": 504,
          "samples_seen": 129024,
          "online_loss": 0.008908587391488254,
          "train_loss": 0.00795592088252306,
          "validation_loss": 0.006716636475175619,
          "train_metric": 0.0008012912003323436,
          "validation_metric": 0.0006764749996364117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2417839280005865
        },
        {
          "epoch": 127,
          "updates": 508,
          "samples_seen": 130048,
          "online_loss": 0.008123657084070146,
          "train_loss": 0.007959316484630108,
          "validation_loss": 0.006222151219844818,
          "train_metric": 0.0008016332285478711,
          "validation_metric": 0.0006266723503358662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.248557812999934
        },
        {
          "epoch": 128,
          "updates": 512,
          "samples_seen": 131072,
          "online_loss": 0.008442080696113408,
          "train_loss": 0.00754531379789114,
          "validation_loss": 0.006155078299343586,
          "train_metric": 0.0007599363452754915,
          "validation_metric": 0.0006199169438332319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.255233068999587
        },
        {
          "epoch": 129,
          "updates": 516,
          "samples_seen": 132096,
          "online_loss": 0.008488746592774987,
          "train_loss": 0.007727129850536585,
          "validation_loss": 0.006506423465907574,
          "train_metric": 0.0007782481843605638,
          "validation_metric": 0.0006553031271323562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2616720980004175
        },
        {
          "epoch": 130,
          "updates": 520,
          "samples_seen": 133120,
          "online_loss": 0.008389452937990427,
          "train_loss": 0.007728844415396452,
          "validation_loss": 0.005690217483788729,
          "train_metric": 0.0007784210029058158,
          "validation_metric": 0.0005730978446081281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.268675280000025
        },
        {
          "epoch": 131,
          "updates": 524,
          "samples_seen": 134144,
          "online_loss": 0.0077995178289711475,
          "train_loss": 0.007659342605620623,
          "validation_loss": 0.006565903313457966,
          "train_metric": 0.0007714208913967013,
          "validation_metric": 0.0006612936849705875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2749825099999725
        },
        {
          "epoch": 132,
          "updates": 528,
          "samples_seen": 135168,
          "online_loss": 0.007510743453167379,
          "train_loss": 0.007357791997492313,
          "validation_loss": 0.005719173699617386,
          "train_metric": 0.0007410498801618814,
          "validation_metric": 0.0005760142230428755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2814527709997492
        },
        {
          "epoch": 133,
          "updates": 532,
          "samples_seen": 136192,
          "online_loss": 0.007291602320037782,
          "train_loss": 0.007175853941589594,
          "validation_loss": 0.00586494617164135,
          "train_metric": 0.0007227257592603564,
          "validation_metric": 0.0005906958249397576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2881371660005243
        },
        {
          "epoch": 134,
          "updates": 536,
          "samples_seen": 137216,
          "online_loss": 0.007291604182682931,
          "train_loss": 0.007205982226878405,
          "validation_loss": 0.005599671509116888,
          "train_metric": 0.0007257601828314364,
          "validation_metric": 0.0005639783921651542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.295411946999593
        },
        {
          "epoch": 135,
          "updates": 540,
          "samples_seen": 138240,
          "online_loss": 0.007426867028698325,
          "train_loss": 0.007119300309568644,
          "validation_loss": 0.005422936752438545,
          "train_metric": 0.0007170297903940082,
          "validation_metric": 0.0005461783148348331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3024450829998386
        },
        {
          "epoch": 136,
          "updates": 544,
          "samples_seen": 139264,
          "online_loss": 0.007183752371929586,
          "train_loss": 0.006937780417501926,
          "validation_loss": 0.0056188637390732765,
          "train_metric": 0.0006987478700466454,
          "validation_metric": 0.0005659113521687686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3089736519996222
        },
        {
          "epoch": 137,
          "updates": 548,
          "samples_seen": 140288,
          "online_loss": 0.007246837019920349,
          "train_loss": 0.006897151470184326,
          "validation_loss": 0.005479155108332634,
          "train_metric": 0.0006946558132767677,
          "validation_metric": 0.0005518404068425298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3157142689997272
        },
        {
          "epoch": 138,
          "updates": 552,
          "samples_seen": 141312,
          "online_loss": 0.007075306493788958,
          "train_loss": 0.007070711348205805,
          "validation_loss": 0.005468420684337616,
          "train_metric": 0.000712136214133352,
          "validation_metric": 0.0005507591995410621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3220388040008402
        },
        {
          "epoch": 139,
          "updates": 556,
          "samples_seen": 142336,
          "online_loss": 0.007378699374385178,
          "train_loss": 0.007405534386634827,
          "validation_loss": 0.006375310942530632,
          "train_metric": 0.0007458582986146212,
          "validation_metric": 0.0006420980207622051,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.328547721999712
        },
        {
          "epoch": 140,
          "updates": 560,
          "samples_seen": 143360,
          "online_loss": 0.009188592550344765,
          "train_loss": 0.010884840972721577,
          "validation_loss": 0.007882092148065567,
          "train_metric": 0.0010962813394144177,
          "validation_metric": 0.0007938554626889527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3349755439994624
        },
        {
          "epoch": 141,
          "updates": 564,
          "samples_seen": 144384,
          "online_loss": 0.010213974164798856,
          "train_loss": 0.009028350934386253,
          "validation_loss": 0.008175701834261417,
          "train_metric": 0.000909302500076592,
          "validation_metric": 0.0008234267588704824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3411878290007735
        },
        {
          "epoch": 142,
          "updates": 568,
          "samples_seen": 145408,
          "online_loss": 0.008107789093628526,
          "train_loss": 0.007513341493904591,
          "validation_loss": 0.005463245790451765,
          "train_metric": 0.0007567162392660975,
          "validation_metric": 0.0005502381245605648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3474808890005079
        },
        {
          "epoch": 143,
          "updates": 572,
          "samples_seen": 146432,
          "online_loss": 0.007420232868753374,
          "train_loss": 0.006564576644450426,
          "validation_loss": 0.0052946945652365685,
          "train_metric": 0.0006611600401811302,
          "validation_metric": 0.0005332622095011175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.354695174999506
        },
        {
          "epoch": 144,
          "updates": 576,
          "samples_seen": 147456,
          "online_loss": 0.006763080949895084,
          "train_loss": 0.006447571329772472,
          "validation_loss": 0.00517623545601964,
          "train_metric": 0.000649375724606216,
          "validation_metric": 0.000521331443451345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3620183450002514
        },
        {
          "epoch": 145,
          "updates": 580,
          "samples_seen": 148480,
          "online_loss": 0.006726964260451496,
          "train_loss": 0.00643564946949482,
          "validation_loss": 0.0050496067851781845,
          "train_metric": 0.000648175016976893,
          "validation_metric": 0.0005085778539068997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.369017311000789
        },
        {
          "epoch": 146,
          "updates": 584,
          "samples_seen": 149504,
          "online_loss": 0.006773887551389635,
          "train_loss": 0.006507103331387043,
          "validation_loss": 0.005337508860975504,
          "train_metric": 0.0006553716957569122,
          "validation_metric": 0.0005375744076445699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3758291949998238
        },
        {
          "epoch": 147,
          "updates": 588,
          "samples_seen": 150528,
          "online_loss": 0.0066409389255568385,
          "train_loss": 0.006286478601396084,
          "validation_loss": 0.0049365549348294735,
          "train_metric": 0.0006331510376185179,
          "validation_metric": 0.0004971917951479554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3831775540002127
        },
        {
          "epoch": 148,
          "updates": 592,
          "samples_seen": 151552,
          "online_loss": 0.006692886468954384,
          "train_loss": 0.006240714341402054,
          "validation_loss": 0.00488189747557044,
          "train_metric": 0.000628541863989085,
          "validation_metric": 0.0004916868056170642,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.390490650999709
        },
        {
          "epoch": 149,
          "updates": 596,
          "samples_seen": 152576,
          "online_loss": 0.0066522266715765,
          "train_loss": 0.006449826527386904,
          "validation_loss": 0.005378819536417723,
          "train_metric": 0.0006496029091067612,
          "validation_metric": 0.0005417350912466645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3970097370001895
        },
        {
          "epoch": 150,
          "updates": 600,
          "samples_seen": 153600,
          "online_loss": 0.00642472377512604,
          "train_loss": 0.0065321773290634155,
          "validation_loss": 0.00497270654886961,
          "train_metric": 0.0006578969769179821,
          "validation_metric": 0.0005008328007534146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4043074389992398
        },
        {
          "epoch": 151,
          "updates": 604,
          "samples_seen": 154624,
          "online_loss": 0.006438150769099593,
          "train_loss": 0.006243400741368532,
          "validation_loss": 0.0048819598741829395,
          "train_metric": 0.0006288124714046717,
          "validation_metric": 0.0004916930920444429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4107006169997476
        },
        {
          "epoch": 152,
          "updates": 608,
          "samples_seen": 155648,
          "online_loss": 0.006163563579320908,
          "train_loss": 0.006107897497713566,
          "validation_loss": 0.004709101282060146,
          "train_metric": 0.0006151651032269001,
          "validation_metric": 0.00047428341349586844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4182080320006207
        },
        {
          "epoch": 153,
          "updates": 612,
          "samples_seen": 156672,
          "online_loss": 0.006159049225971103,
          "train_loss": 0.006299720145761967,
          "validation_loss": 0.005293923895806074,
          "train_metric": 0.0006344847497530282,
          "validation_metric": 0.0005331846186891198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4245044649996998
        },
        {
          "epoch": 154,
          "updates": 616,
          "samples_seen": 157696,
          "online_loss": 0.006196186877787113,
          "train_loss": 0.006035080645233393,
          "validation_loss": 0.004599024076014757,
          "train_metric": 0.0006078312289901078,
          "validation_metric": 0.0004631968040484935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4324177250000503
        },
        {
          "epoch": 155,
          "updates": 620,
          "samples_seen": 158720,
          "online_loss": 0.0062309447675943375,
          "train_loss": 0.006111616734415293,
          "validation_loss": 0.00508724432438612,
          "train_metric": 0.0006155397277325392,
          "validation_metric": 0.0005123685696162283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4388648160002049
        },
        {
          "epoch": 156,
          "updates": 624,
          "samples_seen": 159744,
          "online_loss": 0.006087788264267147,
          "train_loss": 0.0058677056804299355,
          "validation_loss": 0.004460352472960949,
          "train_metric": 0.0005909738829359412,
          "validation_metric": 0.00044923037057742476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4458786850009346
        },
        {
          "epoch": 157,
          "updates": 628,
          "samples_seen": 160768,
          "online_loss": 0.005948200821876526,
          "train_loss": 0.00624513253569603,
          "validation_loss": 0.0053225234150886536,
          "train_metric": 0.0006289868033491075,
          "validation_metric": 0.0005360650247894228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.452805120999983
        },
        {
          "epoch": 158,
          "updates": 632,
          "samples_seen": 161792,
          "online_loss": 0.006033570505678654,
          "train_loss": 0.006047198083251715,
          "validation_loss": 0.004467954859137535,
          "train_metric": 0.0006090516690164804,
          "validation_metric": 0.0004499960341490805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4592086449993076
        },
        {
          "epoch": 159,
          "updates": 636,
          "samples_seen": 162816,
          "online_loss": 0.006026997929438949,
          "train_loss": 0.0058990721590816975,
          "validation_loss": 0.004872919525951147,
          "train_metric": 0.0005941329291090369,
          "validation_metric": 0.0004907825496047735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.465822707999905
        },
        {
          "epoch": 160,
          "updates": 640,
          "samples_seen": 163840,
          "online_loss": 0.005843803286552429,
          "train_loss": 0.005915867630392313,
          "validation_loss": 0.004323561675846577,
          "train_metric": 0.0005958245019428432,
          "validation_metric": 0.000435453315731138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4725619519995234
        },
        {
          "epoch": 161,
          "updates": 644,
          "samples_seen": 164864,
          "online_loss": 0.005885130958631635,
          "train_loss": 0.005636843387037516,
          "validation_loss": 0.004404799081385136,
          "train_metric": 0.0005677221924997866,
          "validation_metric": 0.00044363518827594817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4792369860006147
        },
        {
          "epoch": 162,
          "updates": 648,
          "samples_seen": 165888,
          "online_loss": 0.005719147273339331,
          "train_loss": 0.005603152327239513,
          "validation_loss": 0.0043688565492630005,
          "train_metric": 0.000564329035114497,
          "validation_metric": 0.00044001525384373963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4860009429994534
        },
        {
          "epoch": 163,
          "updates": 652,
          "samples_seen": 166912,
          "online_loss": 0.005620926269330084,
          "train_loss": 0.005593855865299702,
          "validation_loss": 0.004292977042496204,
          "train_metric": 0.000563392648473382,
          "validation_metric": 0.00043237299541942775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4928679229997215
        },
        {
          "epoch": 164,
          "updates": 656,
          "samples_seen": 167936,
          "online_loss": 0.005773326614871621,
          "train_loss": 0.005726131610572338,
          "validation_loss": 0.004252525512129068,
          "train_metric": 0.0005767149850726128,
          "validation_metric": 0.00042829883750528097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5005368780002755
        },
        {
          "epoch": 165,
          "updates": 660,
          "samples_seen": 168960,
          "online_loss": 0.00592881313059479,
          "train_loss": 0.006277122069150209,
          "validation_loss": 0.005479452665895224,
          "train_metric": 0.0006322088302113116,
          "validation_metric": 0.0005518704419955611,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5069605799999408
        },
        {
          "epoch": 166,
          "updates": 664,
          "samples_seen": 169984,
          "online_loss": 0.006270532729104161,
          "train_loss": 0.005576763302087784,
          "validation_loss": 0.004151294939219952,
          "train_metric": 0.0005616711569018662,
          "validation_metric": 0.00041810324182733893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.514078819000133
        },
        {
          "epoch": 167,
          "updates": 668,
          "samples_seen": 171008,
          "online_loss": 0.00547324592480436,
          "train_loss": 0.005422485992312431,
          "validation_loss": 0.004353723023086786,
          "train_metric": 0.0005461329710669816,
          "validation_metric": 0.0004384909989312291,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5205961749998096
        },
        {
          "epoch": 168,
          "updates": 672,
          "samples_seen": 172032,
          "online_loss": 0.005578214535489678,
          "train_loss": 0.0053920987993478775,
          "validation_loss": 0.004344979301095009,
          "train_metric": 0.0005430724122561514,
          "validation_metric": 0.00043761046254076064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.52717932700034
        },
        {
          "epoch": 169,
          "updates": 676,
          "samples_seen": 173056,
          "online_loss": 0.00581419572699815,
          "train_loss": 0.00602647103369236,
          "validation_loss": 0.004381829872727394,
          "train_metric": 0.0006069641676731408,
          "validation_metric": 0.0004413218703120947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5341095460007637
        },
        {
          "epoch": 170,
          "updates": 680,
          "samples_seen": 174080,
          "online_loss": 0.006110282731242478,
          "train_loss": 0.006504336837679148,
          "validation_loss": 0.005769707728177309,
          "train_metric": 0.0006550929974764585,
          "validation_metric": 0.0005811038427054882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5404835410008673
        },
        {
          "epoch": 171,
          "updates": 684,
          "samples_seen": 175104,
          "online_loss": 0.006658103666268289,
          "train_loss": 0.006820253096520901,
          "validation_loss": 0.004942486062645912,
          "train_metric": 0.0006869109929539263,
          "validation_metric": 0.0004977891221642494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5484853380003187
        },
        {
          "epoch": 172,
          "updates": 688,
          "samples_seen": 176128,
          "online_loss": 0.005944876582361758,
          "train_loss": 0.006272169295698404,
          "validation_loss": 0.005498883780092001,
          "train_metric": 0.0006317098741419613,
          "validation_metric": 0.0005538274417631328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5550422569995135
        },
        {
          "epoch": 173,
          "updates": 692,
          "samples_seen": 177152,
          "online_loss": 0.006097919656895101,
          "train_loss": 0.006779646500945091,
          "validation_loss": 0.004918672144412994,
          "train_metric": 0.0006828212062828243,
          "validation_metric": 0.00049539067549631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5613040570005978
        },
        {
          "epoch": 174,
          "updates": 696,
          "samples_seen": 178176,
          "online_loss": 0.006099157268181443,
          "train_loss": 0.006868700031191111,
          "validation_loss": 0.0062558613717556,
          "train_metric": 0.0006917903665453196,
          "validation_metric": 0.000630067428573966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5682846849995258
        },
        {
          "epoch": 175,
          "updates": 700,
          "samples_seen": 179200,
          "online_loss": 0.00620122451800853,
          "train_loss": 0.00585373817011714,
          "validation_loss": 0.004452620632946491,
          "train_metric": 0.0005895670619793236,
          "validation_metric": 0.00044845169759355485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.574639976999606
        },
        {
          "epoch": 176,
          "updates": 704,
          "samples_seen": 180224,
          "online_loss": 0.005916977766901255,
          "train_loss": 0.005310048349201679,
          "validation_loss": 0.003988733049482107,
          "train_metric": 0.0005348086124286056,
          "validation_metric": 0.0004017305909655988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5819130929994571
        },
        {
          "epoch": 177,
          "updates": 708,
          "samples_seen": 181248,
          "online_loss": 0.006168559135403484,
          "train_loss": 0.005656643304973841,
          "validation_loss": 0.00485645467415452,
          "train_metric": 0.0005697163287550211,
          "validation_metric": 0.0004891243297606707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6675638560009247
        },
        {
          "epoch": 178,
          "updates": 712,
          "samples_seen": 182272,
          "online_loss": 0.0056245679734274745,
          "train_loss": 0.00573745695874095,
          "validation_loss": 0.004067534115165472,
          "train_metric": 0.0005778555641882122,
          "validation_metric": 0.00040966711821965873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6743337989992142
        },
        {
          "epoch": 179,
          "updates": 716,
          "samples_seen": 183296,
          "online_loss": 0.0059484983794391155,
          "train_loss": 0.005876262206584215,
          "validation_loss": 0.0051709976978600025,
          "train_metric": 0.0005918355891481042,
          "validation_metric": 0.0005208039656281471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6808769599992957
        },
        {
          "epoch": 180,
          "updates": 720,
          "samples_seen": 184320,
          "online_loss": 0.0054383642273023725,
          "train_loss": 0.005765881389379501,
          "validation_loss": 0.004067170433700085,
          "train_metric": 0.000580718507990241,
          "validation_metric": 0.0004096305347047746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6873599739992642
        },
        {
          "epoch": 181,
          "updates": 724,
          "samples_seen": 185344,
          "online_loss": 0.005147158983163536,
          "train_loss": 0.005435269791632891,
          "validation_loss": 0.004647584166377783,
          "train_metric": 0.0005474204663187265,
          "validation_metric": 0.00046808773186057806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.693813534999208
        },
        {
          "epoch": 182,
          "updates": 728,
          "samples_seen": 186368,
          "online_loss": 0.005237859790213406,
          "train_loss": 0.005441064015030861,
          "validation_loss": 0.004048799630254507,
          "train_metric": 0.000548004056327045,
          "validation_metric": 0.0004077802295796573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7004791960007424
        },
        {
          "epoch": 183,
          "updates": 732,
          "samples_seen": 187392,
          "online_loss": 0.005174182238988578,
          "train_loss": 0.0051713380962610245,
          "validation_loss": 0.0043060933239758015,
          "train_metric": 0.0005208382499404252,
          "validation_metric": 0.0004336939309723675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7067479510005796
        },
        {
          "epoch": 184,
          "updates": 736,
          "samples_seen": 188416,
          "online_loss": 0.0050398277817294,
          "train_loss": 0.004932196345180273,
          "validation_loss": 0.003746675094589591,
          "train_metric": 0.000496752792969346,
          "validation_metric": 0.00037735135992988944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7137014679992717
        },
        {
          "epoch": 185,
          "updates": 740,
          "samples_seen": 189440,
          "online_loss": 0.005100349138956517,
          "train_loss": 0.00485992431640625,
          "validation_loss": 0.003687598742544651,
          "train_metric": 0.0004894738085567951,
          "validation_metric": 0.0003714013728313148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.720627216000139
        },
        {
          "epoch": 186,
          "updates": 744,
          "samples_seen": 190464,
          "online_loss": 0.00527846539625898,
          "train_loss": 0.004803177434951067,
          "validation_loss": 0.0038038224447518587,
          "train_metric": 0.00048375839833170176,
          "validation_metric": 0.0003831070498563349,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7270075819997146
        },
        {
          "epoch": 187,
          "updates": 748,
          "samples_seen": 191488,
          "online_loss": 0.005314036156050861,
          "train_loss": 0.0052468376234173775,
          "validation_loss": 0.004495000001043081,
          "train_metric": 0.0005284422659315169,
          "validation_metric": 0.00045271991984918714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7335805000002438
        },
        {
          "epoch": 188,
          "updates": 752,
          "samples_seen": 192512,
          "online_loss": 0.005319322575815022,
          "train_loss": 0.0052781980484724045,
          "validation_loss": 0.0037307802122086287,
          "train_metric": 0.0005316007882356644,
          "validation_metric": 0.00037575053283944726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7399029919997702
        },
        {
          "epoch": 189,
          "updates": 756,
          "samples_seen": 193536,
          "online_loss": 0.005059752555098385,
          "train_loss": 0.006521434988826513,
          "validation_loss": 0.006038823630660772,
          "train_metric": 0.0006568150129169226,
          "validation_metric": 0.0006082082400098443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7462718269998732
        },
        {
          "epoch": 190,
          "updates": 760,
          "samples_seen": 194560,
          "online_loss": 0.006297448766417801,
          "train_loss": 0.004986852873116732,
          "validation_loss": 0.0037775097880512476,
          "train_metric": 0.0005022575496695936,
          "validation_metric": 0.0003804569423664361,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.752620322000439
        },
        {
          "epoch": 191,
          "updates": 764,
          "samples_seen": 195584,
          "online_loss": 0.0052445061155594885,
          "train_loss": 0.004741066601127386,
          "validation_loss": 0.003689645556733012,
          "train_metric": 0.00047750285011716187,
          "validation_metric": 0.0003716076025739312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7588124230005633
        },
        {
          "epoch": 192,
          "updates": 768,
          "samples_seen": 196608,
          "online_loss": 0.005138739012181759,
          "train_loss": 0.004897554405033588,
          "validation_loss": 0.004064996261149645,
          "train_metric": 0.000493263709358871,
          "validation_metric": 0.0004094115283805877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7651027160009107
        },
        {
          "epoch": 193,
          "updates": 772,
          "samples_seen": 197632,
          "online_loss": 0.004820525762625039,
          "train_loss": 0.004927412141114473,
          "validation_loss": 0.0035210587084293365,
          "train_metric": 0.0004962708917446434,
          "validation_metric": 0.0003546281368471682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7722497070008103
        },
        {
          "epoch": 194,
          "updates": 776,
          "samples_seen": 198656,
          "online_loss": 0.004927562491502613,
          "train_loss": 0.004930955357849598,
          "validation_loss": 0.004180684220045805,
          "train_metric": 0.0004966277629137039,
          "validation_metric": 0.0004210632178001106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7796219869997003
        },
        {
          "epoch": 195,
          "updates": 780,
          "samples_seen": 199680,
          "online_loss": 0.00508161645848304,
          "train_loss": 0.005184325389564037,
          "validation_loss": 0.003777904435992241,
          "train_metric": 0.0005221462924964726,
          "validation_metric": 0.00038049669819884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7869071959994471
        },
        {
          "epoch": 196,
          "updates": 784,
          "samples_seen": 200704,
          "online_loss": 0.006852328253444284,
          "train_loss": 0.00959151890128851,
          "validation_loss": 0.009228128008544445,
          "train_metric": 0.0009660226060077548,
          "validation_metric": 0.000929423316847533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7932791389994236
        },
        {
          "epoch": 197,
          "updates": 788,
          "samples_seen": 201728,
          "online_loss": 0.008761042612604797,
          "train_loss": 0.008050591684877872,
          "validation_loss": 0.005521757528185844,
          "train_metric": 0.0008108262554742396,
          "validation_metric": 0.0005561312427744269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8000141419997817
        },
        {
          "epoch": 198,
          "updates": 792,
          "samples_seen": 202752,
          "online_loss": 0.007816763187292963,
          "train_loss": 0.0046900310553610325,
          "validation_loss": 0.0038949428126215935,
          "train_metric": 0.0004723627062048763,
          "validation_metric": 0.00039228436071425676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8063203130004695
        },
        {
          "epoch": 199,
          "updates": 796,
          "samples_seen": 203776,
          "online_loss": 0.0061726594576612115,
          "train_loss": 0.005147816147655249,
          "validation_loss": 0.004460660740733147,
          "train_metric": 0.000518469256348908,
          "validation_metric": 0.00044926145346835256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8136644320002233
        },
        {
          "epoch": 200,
          "updates": 800,
          "samples_seen": 204800,
          "online_loss": 0.005255774711258709,
          "train_loss": 0.004721092060208321,
          "validation_loss": 0.0034721598494797945,
          "train_metric": 0.00047549110604450107,
          "validation_metric": 0.00034970324486494064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.820746956000221
        },
        {
          "epoch": 201,
          "updates": 804,
          "samples_seen": 205824,
          "online_loss": 0.004587506526149809,
          "train_loss": 0.004755346104502678,
          "validation_loss": 0.00401246128603816,
          "train_metric": 0.00047894101589918137,
          "validation_metric": 0.0004041204520035535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8269263199999841
        },
        {
          "epoch": 202,
          "updates": 808,
          "samples_seen": 206848,
          "online_loss": 0.0048797979252412915,
          "train_loss": 0.00566001795232296,
          "validation_loss": 0.003972616046667099,
          "train_metric": 0.0005700562614947557,
          "validation_metric": 0.0004001073248218745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8335754330000782
        },
        {
          "epoch": 203,
          "updates": 812,
          "samples_seen": 207872,
          "online_loss": 0.005326757149305195,
          "train_loss": 0.005097195040434599,
          "validation_loss": 0.004448319785296917,
          "train_metric": 0.0005133708473294973,
          "validation_metric": 0.0004480184579733759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8399076580008114
        },
        {
          "epoch": 204,
          "updates": 816,
          "samples_seen": 208896,
          "online_loss": 0.004783886601217091,
          "train_loss": 0.004721126053482294,
          "validation_loss": 0.003458975115790963,
          "train_metric": 0.0004754944529850036,
          "validation_metric": 0.0003483752952888608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8479039350004314
        },
        {
          "epoch": 205,
          "updates": 820,
          "samples_seen": 209920,
          "online_loss": 0.004660422157030553,
          "train_loss": 0.004440572578459978,
          "validation_loss": 0.0036539456341415644,
          "train_metric": 0.00044723821338266134,
          "validation_metric": 0.0003680119989439845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.854423635000785
        },
        {
          "epoch": 206,
          "updates": 824,
          "samples_seen": 210944,
          "online_loss": 0.004735485359560698,
          "train_loss": 0.004506292287260294,
          "validation_loss": 0.0038703978061676025,
          "train_metric": 0.0004538572393357754,
          "validation_metric": 0.00038981231045909226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8606629999994766
        },
        {
          "epoch": 207,
          "updates": 828,
          "samples_seen": 211968,
          "online_loss": 0.004686791333369911,
          "train_loss": 0.004576239734888077,
          "validation_loss": 0.0032699063885957003,
          "train_metric": 0.00046090211253613234,
          "validation_metric": 0.0003293330373708159,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8681072649997077
        },
        {
          "epoch": 208,
          "updates": 832,
          "samples_seen": 212992,
          "online_loss": 0.004739412164781243,
          "train_loss": 0.00469451118260622,
          "validation_loss": 0.00410153903067112,
          "train_metric": 0.0004728139901999384,
          "validation_metric": 0.0004130919696763158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8745227100007469
        },
        {
          "epoch": 209,
          "updates": 836,
          "samples_seen": 214016,
          "online_loss": 0.00442184682469815,
          "train_loss": 0.004894668236374855,
          "validation_loss": 0.0035039603244513273,
          "train_metric": 0.00049297307850793,
          "validation_metric": 0.0003529060631990433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8812225100009528
        },
        {
          "epoch": 210,
          "updates": 840,
          "samples_seen": 215040,
          "online_loss": 0.004659323953092098,
          "train_loss": 0.005034927278757095,
          "validation_loss": 0.004545057658106089,
          "train_metric": 0.0005070993793196976,
          "validation_metric": 0.00045776154729537666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8875876729998708
        },
        {
          "epoch": 211,
          "updates": 844,
          "samples_seen": 216064,
          "online_loss": 0.00464281887980178,
          "train_loss": 0.005427856929600239,
          "validation_loss": 0.0037575243040919304,
          "train_metric": 0.0005466738366521895,
          "validation_metric": 0.00037844403414055705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8939602249993186
        },
        {
          "epoch": 212,
          "updates": 848,
          "samples_seen": 217088,
          "online_loss": 0.00480867363512516,
          "train_loss": 0.005467430222779512,
          "validation_loss": 0.005027767736464739,
          "train_metric": 0.0005506596062332392,
          "validation_metric": 0.0005063783610239625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9008672290001414
        },
        {
          "epoch": 213,
          "updates": 852,
          "samples_seen": 218112,
          "online_loss": 0.004953040450345725,
          "train_loss": 0.0046825106255710125,
          "validation_loss": 0.003325053723528981,
          "train_metric": 0.00047160530812107027,
          "validation_metric": 0.0003348872996866703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9071513149992825
        },
        {
          "epoch": 214,
          "updates": 856,
          "samples_seen": 219136,
          "online_loss": 0.004484567209146917,
          "train_loss": 0.0046012201346457005,
          "validation_loss": 0.0041000209748744965,
          "train_metric": 0.0004634179931599647,
          "validation_metric": 0.000412939116358757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9136132670009829
        },
        {
          "epoch": 215,
          "updates": 860,
          "samples_seen": 220160,
          "online_loss": 0.004388863744679838,
          "train_loss": 0.004953176248818636,
          "validation_loss": 0.003551032394170761,
          "train_metric": 0.0004988657892681658,
          "validation_metric": 0.0003576469898689538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9198357059995033
        },
        {
          "epoch": 216,
          "updates": 864,
          "samples_seen": 221184,
          "online_loss": 0.004554339102469385,
          "train_loss": 0.004855134058743715,
          "validation_loss": 0.004340296145528555,
          "train_metric": 0.0004889912670478225,
          "validation_metric": 0.00043713871855288744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.925794061000488
        },
        {
          "epoch": 217,
          "updates": 868,
          "samples_seen": 222208,
          "online_loss": 0.004486666759476066,
          "train_loss": 0.005335094407200813,
          "validation_loss": 0.0037313762586563826,
          "train_metric": 0.0005373311578296125,
          "validation_metric": 0.00037581054493784904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.932176824000635
        },
        {
          "epoch": 218,
          "updates": 872,
          "samples_seen": 223232,
          "online_loss": 0.004694714385550469,
          "train_loss": 0.005244765430688858,
          "validation_loss": 0.004788707010447979,
          "train_metric": 0.0005282335332594812,
          "validation_metric": 0.00048230105312541127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.938487052999335
        },
        {
          "epoch": 219,
          "updates": 876,
          "samples_seen": 224256,
          "online_loss": 0.005128097953274846,
          "train_loss": 0.005142275243997574,
          "validation_loss": 0.0036153479013592005,
          "train_metric": 0.0005179111612960696,
          "validation_metric": 0.0003641246003098786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9449013289995492
        },
        {
          "epoch": 220,
          "updates": 880,
          "samples_seen": 225280,
          "online_loss": 0.005097834509797394,
          "train_loss": 0.004150544758886099,
          "validation_loss": 0.0032629109919071198,
          "train_metric": 0.0004180277173873037,
          "validation_metric": 0.00032862843363545835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9522345170007611
        },
        {
          "epoch": 221,
          "updates": 884,
          "samples_seen": 226304,
          "online_loss": 0.004507303237915039,
          "train_loss": 0.004174862522631884,
          "validation_loss": 0.0034939427860081196,
          "train_metric": 0.00042047686292789876,
          "validation_metric": 0.00035189714981243014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9585357630003273
        },
        {
          "epoch": 222,
          "updates": 888,
          "samples_seen": 227328,
          "online_loss": 0.004981880134437233,
          "train_loss": 0.00597273837774992,
          "validation_loss": 0.004021317698061466,
          "train_metric": 0.0006015523103997111,
          "validation_metric": 0.0004050124261993915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9653015690000757
        },
        {
          "epoch": 223,
          "updates": 892,
          "samples_seen": 228352,
          "online_loss": 0.005149070580955595,
          "train_loss": 0.005294284783303738,
          "validation_loss": 0.005001806654036045,
          "train_metric": 0.0005332209402695298,
          "validation_metric": 0.0005037636728957295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.97159982499943
        },
        {
          "epoch": 224,
          "updates": 896,
          "samples_seen": 229376,
          "online_loss": 0.0051251358818262815,
          "train_loss": 0.00575705710798502,
          "validation_loss": 0.003921120893210173,
          "train_metric": 0.0005798296770080924,
          "validation_metric": 0.0003949209058191627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9779533660002926
        },
        {
          "epoch": 225,
          "updates": 900,
          "samples_seen": 230400,
          "online_loss": 0.005643898795824498,
          "train_loss": 0.004740901291370392,
          "validation_loss": 0.004359474405646324,
          "train_metric": 0.0004774862027261406,
          "validation_metric": 0.00043907028157263994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9848562330007553
        },
        {
          "epoch": 226,
          "updates": 904,
          "samples_seen": 231424,
          "online_loss": 0.005312400171533227,
          "train_loss": 0.004710732959210873,
          "validation_loss": 0.003253556089475751,
          "train_metric": 0.0004744477628264576,
          "validation_metric": 0.0003276862553320825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9921664939993207
        },
        {
          "epoch": 227,
          "updates": 908,
          "samples_seen": 232448,
          "online_loss": 0.005090569029562175,
          "train_loss": 0.004007077310234308,
          "validation_loss": 0.003029836108908057,
          "train_metric": 0.0004035781021229923,
          "validation_metric": 0.0003051539824809879,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0417923999993945
        },
        {
          "epoch": 228,
          "updates": 912,
          "samples_seen": 233472,
          "online_loss": 0.005621287098620087,
          "train_loss": 0.0047398717142641544,
          "validation_loss": 0.004227905068546534,
          "train_metric": 0.0004773825057782233,
          "validation_metric": 0.0004258191038388759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.089326751000044
        },
        {
          "epoch": 229,
          "updates": 916,
          "samples_seen": 234496,
          "online_loss": 0.005442690453492105,
          "train_loss": 0.005949027836322784,
          "validation_loss": 0.004119441378861666,
          "train_metric": 0.0005991642246954143,
          "validation_metric": 0.0004148950392846018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.121788956999808
        },
        {
          "epoch": 230,
          "updates": 920,
          "samples_seen": 235520,
          "online_loss": 0.005835753632709384,
          "train_loss": 0.005283336620777845,
          "validation_loss": 0.004755840636789799,
          "train_metric": 0.0005321182543411851,
          "validation_metric": 0.00047899087076075375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.142419867999706
        },
        {
          "epoch": 231,
          "updates": 924,
          "samples_seen": 236544,
          "online_loss": 0.00449197948910296,
          "train_loss": 0.004196472465991974,
          "validation_loss": 0.0029929163865745068,
          "train_metric": 0.0004226533346809447,
          "validation_metric": 0.00030143558979034424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1694650530007493
        },
        {
          "epoch": 232,
          "updates": 928,
          "samples_seen": 237568,
          "online_loss": 0.004350543720647693,
          "train_loss": 0.0039774696342647076,
          "validation_loss": 0.0033009392209351063,
          "train_metric": 0.0004005962109658867,
          "validation_metric": 0.00033245846861973405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.180910581999342
        },
        {
          "epoch": 233,
          "updates": 932,
          "samples_seen": 238592,
          "online_loss": 0.004667789209634066,
          "train_loss": 0.00384520529769361,
          "validation_loss": 0.0031571462750434875,
          "train_metric": 0.00038727495120838284,
          "validation_metric": 0.0003179761697538197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1973207450009795
        },
        {
          "epoch": 234,
          "updates": 936,
          "samples_seen": 239616,
          "online_loss": 0.004615488811396062,
          "train_loss": 0.004501578863710165,
          "validation_loss": 0.003222852945327759,
          "train_metric": 0.0004533825267571956,
          "validation_metric": 0.00032459391513839364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.209374095000385
        },
        {
          "epoch": 235,
          "updates": 940,
          "samples_seen": 240640,
          "online_loss": 0.00449463416589424,
          "train_loss": 0.0049210903234779835,
          "validation_loss": 0.0045022196136415005,
          "train_metric": 0.000495634158141911,
          "validation_metric": 0.00045344707905314863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.218157023000458
        },
        {
          "epoch": 236,
          "updates": 944,
          "samples_seen": 241664,
          "online_loss": 0.00459163764026016,
          "train_loss": 0.004945152439177036,
          "validation_loss": 0.0032366979867219925,
          "train_metric": 0.0004980575758963823,
          "validation_metric": 0.0003259883669670671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2252464880002663
        },
        {
          "epoch": 237,
          "updates": 948,
          "samples_seen": 242688,
          "online_loss": 0.004988239554222673,
          "train_loss": 0.0050831628032028675,
          "validation_loss": 0.004825833719223738,
          "train_metric": 0.0005119575071148574,
          "validation_metric": 0.00048604028415866196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2333408380000037
        },
        {
          "epoch": 238,
          "updates": 952,
          "samples_seen": 243712,
          "online_loss": 0.004285260220058262,
          "train_loss": 0.0045090713538229465,
          "validation_loss": 0.0031676641665399075,
          "train_metric": 0.0004541371308732778,
          "validation_metric": 0.00031903557828627527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2403082769997127
        },
        {
          "epoch": 239,
          "updates": 956,
          "samples_seen": 244736,
          "online_loss": 0.004265384573955089,
          "train_loss": 0.004078837111592293,
          "validation_loss": 0.0036378910299390554,
          "train_metric": 0.00041080551454797387,
          "validation_metric": 0.0003663950483314693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.248310905000835
        },
        {
          "epoch": 240,
          "updates": 960,
          "samples_seen": 245760,
          "online_loss": 0.0042668330715969205,
          "train_loss": 0.0037082515191286802,
          "validation_loss": 0.0028293279465287924,
          "train_metric": 0.00037348151090554893,
          "validation_metric": 0.0002849595621228218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2556710509998084
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 960
      },
      "best_validation_metric": 0.0002849595621228218,
      "config": {
        "label": "Batch 256",
        "batch": 256,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01688074530708943
    },
    "1150d4df4a82_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7750606071203947,
          "train_loss": 0.38536715507507324,
          "validation_loss": 0.30186760425567627,
          "train_metric": 0.03881276398897171,
          "validation_metric": 0.030403001233935356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01394530799916538
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.2060750015079975,
          "train_loss": 0.12688414752483368,
          "validation_loss": 0.1095275804400444,
          "train_metric": 0.01277930662035942,
          "validation_metric": 0.011031216941773891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.025526415000058478
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.1240247250534594,
          "train_loss": 0.09904789179563522,
          "validation_loss": 0.08297248184680939,
          "train_metric": 0.009975740686058998,
          "validation_metric": 0.008356684818863869,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.036793866000152775
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09276640950702131,
          "train_loss": 0.0849667638540268,
          "validation_loss": 0.06827860325574875,
          "train_metric": 0.00855754129588604,
          "validation_metric": 0.006876770872622728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04818809400057944
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07504639495164156,
          "train_loss": 0.06760113686323166,
          "validation_loss": 0.055644359439611435,
          "train_metric": 0.006808537989854813,
          "validation_metric": 0.005604295991361141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05970213999898988
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06591126136481762,
          "train_loss": 0.0652155801653862,
          "validation_loss": 0.052817028015851974,
          "train_metric": 0.0065682739950716496,
          "validation_metric": 0.005319537594914436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07081351499982702
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06305562076158822,
          "train_loss": 0.06514137238264084,
          "validation_loss": 0.05300118774175644,
          "train_metric": 0.00656080013141036,
          "validation_metric": 0.005338085815310478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08134332999907201
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06273406581021845,
          "train_loss": 0.05920339375734329,
          "validation_loss": 0.04777815192937851,
          "train_metric": 0.005962749011814594,
          "validation_metric": 0.0048120408318936825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09218506800061732
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05564656085334718,
          "train_loss": 0.052560701966285706,
          "validation_loss": 0.0434177927672863,
          "train_metric": 0.005293721333146095,
          "validation_metric": 0.004372881259769201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19722700099919166
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05131560959853232,
          "train_loss": 0.051526989787817,
          "validation_loss": 0.04277803748846054,
          "train_metric": 0.0051896097138524055,
          "validation_metric": 0.004308447707444429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20957076500053518
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04965876357164234,
          "train_loss": 0.0480007603764534,
          "validation_loss": 0.03882240504026413,
          "train_metric": 0.004834461025893688,
          "validation_metric": 0.003910050727427006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22167197899943858
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04525484796613455,
          "train_loss": 0.041656482964754105,
          "validation_loss": 0.034643612802028656,
          "train_metric": 0.004195488058030605,
          "validation_metric": 0.0034891776740550995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25080002899994724
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.0406425753608346,
          "train_loss": 0.03904619812965393,
          "validation_loss": 0.032580845057964325,
          "train_metric": 0.003932590130716562,
          "validation_metric": 0.003281423356384039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2800469840003643
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03730706369969994,
          "train_loss": 0.03512808308005333,
          "validation_loss": 0.029410352930426598,
          "train_metric": 0.003537971992045641,
          "validation_metric": 0.0029621028807014227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30235009799980617
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03614106320310384,
          "train_loss": 0.03985850140452385,
          "validation_loss": 0.033791474997997284,
          "train_metric": 0.004014401696622372,
          "validation_metric": 0.0034033532720059156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32629683499908424
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.034006054629571736,
          "train_loss": 0.03577851876616478,
          "validation_loss": 0.03025796450674534,
          "train_metric": 0.0036034814547747374,
          "validation_metric": 0.0030474714003503323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3499021530005848
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.030239578511100262,
          "train_loss": 0.02710607275366783,
          "validation_loss": 0.023179994896054268,
          "train_metric": 0.0027300245128571987,
          "validation_metric": 0.00233460427261889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37616774199887004
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029969822440762073,
          "train_loss": 0.03654274344444275,
          "validation_loss": 0.028627360239624977,
          "train_metric": 0.003680451540276408,
          "validation_metric": 0.0028832426760345697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39235470699895814
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.031163847947027534,
          "train_loss": 0.02538416162133217,
          "validation_loss": 0.019166603684425354,
          "train_metric": 0.002556599909439683,
          "validation_metric": 0.0019303900189697742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40843228099947737
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.023735167924314737,
          "train_loss": 0.02233070135116577,
          "validation_loss": 0.018863607197999954,
          "train_metric": 0.0022490667179226875,
          "validation_metric": 0.001899873255752027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42482974599988665
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023142108839238063,
          "train_loss": 0.01990034058690071,
          "validation_loss": 0.01686880923807621,
          "train_metric": 0.002004289533942938,
          "validation_metric": 0.0016989647410809994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44157424500008347
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02312045422149822,
          "train_loss": 0.020990870893001556,
          "validation_loss": 0.016307419165968895,
          "train_metric": 0.002114123897626996,
          "validation_metric": 0.0016424234490841627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4582701819999784
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.022139596985653043,
          "train_loss": 0.020363621413707733,
          "validation_loss": 0.01820952072739601,
          "train_metric": 0.0020509492605924606,
          "validation_metric": 0.0018339961534366012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47388019299978623
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.018391056975815445,
          "train_loss": 0.01689496450126171,
          "validation_loss": 0.01443193294107914,
          "train_metric": 0.001701599103398621,
          "validation_metric": 0.001453531440347433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4903377129994624
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019256686326116323,
          "train_loss": 0.019413551315665245,
          "validation_loss": 0.014614056795835495,
          "train_metric": 0.0019552616868168116,
          "validation_metric": 0.0014718743041157722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5145302839991928
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.018628374149557203,
          "train_loss": 0.021552782505750656,
          "validation_loss": 0.02029510959982872,
          "train_metric": 0.002170717343688011,
          "validation_metric": 0.0020440490916371346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5341339229998994
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.017534870916279033,
          "train_loss": 0.015243593603372574,
          "validation_loss": 0.013040601275861263,
          "train_metric": 0.001535278744995594,
          "validation_metric": 0.001313401386141777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5517825890001404
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015402383840410039,
          "train_loss": 0.014646788127720356,
          "validation_loss": 0.012762422673404217,
          "train_metric": 0.0014751707203686237,
          "validation_metric": 0.0012853844091296196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5690736340002331
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016872672975296155,
          "train_loss": 0.018696356564760208,
          "validation_loss": 0.015021326020359993,
          "train_metric": 0.0018830286571756005,
          "validation_metric": 0.001512893009930849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.58549527499963
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.016743629763368517,
          "train_loss": 0.01551545225083828,
          "validation_loss": 0.013348198495805264,
          "train_metric": 0.0015626593958586454,
          "validation_metric": 0.001344381715171039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6008882909991371
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.021813345549162477,
          "train_loss": 0.014776387251913548,
          "validation_loss": 0.011951805092394352,
          "train_metric": 0.001488223671913147,
          "validation_metric": 0.0012037419946864247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.617206470000383
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.018929659272544086,
          "train_loss": 0.014764338731765747,
          "validation_loss": 0.010697023011744022,
          "train_metric": 0.0014870099257677794,
          "validation_metric": 0.001077365013770759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6341964459988958
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.0153573855641298,
          "train_loss": 0.012026515789330006,
          "validation_loss": 0.009608716703951359,
          "train_metric": 0.0012112664990127087,
          "validation_metric": 0.0009677548077888787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6514862519998132
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013202551752328873,
          "train_loss": 0.01557985134422779,
          "validation_loss": 0.010827160440385342,
          "train_metric": 0.0015691454755142331,
          "validation_metric": 0.0010904718656092882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6678507759988861
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014358282889588736,
          "train_loss": 0.015606962144374847,
          "validation_loss": 0.014100464060902596,
          "train_metric": 0.0015718759968876839,
          "validation_metric": 0.001420147018507123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6842465459994855
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.016984775924356654,
          "train_loss": 0.02222522161900997,
          "validation_loss": 0.01624390110373497,
          "train_metric": 0.002238442888483405,
          "validation_metric": 0.001636026194319129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7007267890003277
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.012299630441702902,
          "train_loss": 0.011116322129964828,
          "validation_loss": 0.009583269245922565,
          "train_metric": 0.0011195952538400888,
          "validation_metric": 0.0009651916334405541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7176316069999302
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011669085710309446,
          "train_loss": 0.009766273200511932,
          "validation_loss": 0.007932213135063648,
          "train_metric": 0.0009836232056841254,
          "validation_metric": 0.0007989035220816731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7354386629995133
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01077001288649626,
          "train_loss": 0.009837810881435871,
          "validation_loss": 0.00763594638556242,
          "train_metric": 0.0009908282663673162,
          "validation_metric": 0.0007690644706599414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.753490167999189
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012661474363994785,
          "train_loss": 0.012773655354976654,
          "validation_loss": 0.009138710796833038,
          "train_metric": 0.0012865157332271338,
          "validation_metric": 0.0009204175439663231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7701550210003916
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010784950369270518,
          "train_loss": 0.009931527078151703,
          "validation_loss": 0.007522317115217447,
          "train_metric": 0.0010002671042457223,
          "validation_metric": 0.0007576202624477446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7876069109988748
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010854612351977266,
          "train_loss": 0.009563778527081013,
          "validation_loss": 0.007020589429885149,
          "train_metric": 0.0009632286964915693,
          "validation_metric": 0.0007070880965329707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8052136520000204
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011339242788380943,
          "train_loss": 0.0087605444714427,
          "validation_loss": 0.007057554554194212,
          "train_metric": 0.0008823300013318658,
          "validation_metric": 0.0007108110003173351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8224396419991535
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.00978471252892632,
          "train_loss": 0.00902597513049841,
          "validation_loss": 0.0073502580635249615,
          "train_metric": 0.0009090631501749158,
          "validation_metric": 0.0007402910850942135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8386514419999003
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013840032886946574,
          "train_loss": 0.010268403217196465,
          "validation_loss": 0.008520981296896935,
          "train_metric": 0.0010341958841308951,
          "validation_metric": 0.0008582019945606589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8549502639998536
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.012792111374437809,
          "train_loss": 0.010196718387305737,
          "validation_loss": 0.007601565681397915,
          "train_metric": 0.0010269759222865105,
          "validation_metric": 0.0007656019297428429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8715241690006224
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011586189008085057,
          "train_loss": 0.012726811692118645,
          "validation_loss": 0.011708073318004608,
          "train_metric": 0.0012817978858947754,
          "validation_metric": 0.00117919419426471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8881060740004614
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.009888891261653043,
          "train_loss": 0.014078420586884022,
          "validation_loss": 0.013561593368649483,
          "train_metric": 0.0014179269783198833,
          "validation_metric": 0.0013658740790560842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9046823600001517
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.01211502478690818,
          "train_loss": 0.010559041053056717,
          "validation_loss": 0.008403655141592026,
          "train_metric": 0.001063467818312347,
          "validation_metric": 0.0008463853155262768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9210181790003844
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.011117999012640212,
          "train_loss": 0.00816005002707243,
          "validation_loss": 0.006046057213097811,
          "train_metric": 0.0008218504954129457,
          "validation_metric": 0.0006089368253014982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9382332539989875
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.00928575212310534,
          "train_loss": 0.012129377573728561,
          "validation_loss": 0.01141818892210722,
          "train_metric": 0.0012216264149174094,
          "validation_metric": 0.0011499980464577675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9548690989995521
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008625807182397693,
          "train_loss": 0.00884421169757843,
          "validation_loss": 0.0063227578066289425,
          "train_metric": 0.0008907566079869866,
          "validation_metric": 0.0006368050235323608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9715625339995313
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008779473515460268,
          "train_loss": 0.010252740234136581,
          "validation_loss": 0.00750874076038599,
          "train_metric": 0.0010326184565201402,
          "validation_metric": 0.0007562528480775654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9878009809999639
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0091602051834343,
          "train_loss": 0.01068117842078209,
          "validation_loss": 0.010125876404345036,
          "train_metric": 0.0010757691925391555,
          "validation_metric": 0.0010198411764577031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0039477850004914
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.013468224395182915,
          "train_loss": 0.00999869592487812,
          "validation_loss": 0.007742003537714481,
          "train_metric": 0.0010070321150124073,
          "validation_metric": 0.0007797462749294937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.019932205999794
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.010853008774574846,
          "train_loss": 0.008161451667547226,
          "validation_loss": 0.0058862557634711266,
          "train_metric": 0.0008219915325753391,
          "validation_metric": 0.0005928422324359417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0369319139990694
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008787122336798348,
          "train_loss": 0.007462658453732729,
          "validation_loss": 0.006129884626716375,
          "train_metric": 0.0007516115438193083,
          "validation_metric": 0.0006173795554786921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.053408200999911
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007215693418402225,
          "train_loss": 0.006528432480990887,
          "validation_loss": 0.005115540232509375,
          "train_metric": 0.0006575197330676019,
          "validation_metric": 0.0005152184749022126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.070477502000358
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007845858577638865,
          "train_loss": 0.0069840820506215096,
          "validation_loss": 0.006143107544630766,
          "train_metric": 0.0007034111185930669,
          "validation_metric": 0.0006187112885527313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0867133600004308
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007135726671549492,
          "train_loss": 0.006284323520958424,
          "validation_loss": 0.005035600624978542,
          "train_metric": 0.0006329340394586325,
          "validation_metric": 0.0005071673076599836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1046867610002664
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.009110017359489575,
          "train_loss": 0.006696862168610096,
          "validation_loss": 0.005621404852718115,
          "train_metric": 0.0006744833663105965,
          "validation_metric": 0.0005661673494614661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1275177209990943
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007818591664545238,
          "train_loss": 0.006637724116444588,
          "validation_loss": 0.005256288219243288,
          "train_metric": 0.0006685271509923041,
          "validation_metric": 0.0005293941358104348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1664489020004112
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006970037764403969,
          "train_loss": 0.006174634676426649,
          "validation_loss": 0.004937969613820314,
          "train_metric": 0.0006218865746632218,
          "validation_metric": 0.0004973342875018716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2131031680000888
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007135511885280721,
          "train_loss": 0.009442456066608429,
          "validation_loss": 0.008755910210311413,
          "train_metric": 0.0009510095696896315,
          "validation_metric": 0.0008818632340990007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2557679589990585
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008260824164608493,
          "train_loss": 0.008309222757816315,
          "validation_loss": 0.0070204706862568855,
          "train_metric": 0.0008368745329789817,
          "validation_metric": 0.0007070761639624834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2836555289995886
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0069822732475586236,
          "train_loss": 0.0067450315691530704,
          "validation_loss": 0.004807147663086653,
          "train_metric": 0.000679334974847734,
          "validation_metric": 0.0004841582849621773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3006843109997135
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008227432139392477,
          "train_loss": 0.00827042106539011,
          "validation_loss": 0.005731015466153622,
          "train_metric": 0.0008329665288329124,
          "validation_metric": 0.0005772068398073316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3171404160002567
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00791827542707324,
          "train_loss": 0.006597149185836315,
          "validation_loss": 0.00453611183911562,
          "train_metric": 0.000664440740365535,
          "validation_metric": 0.00045686058001592755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3341098330001842
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0060867796710226685,
          "train_loss": 0.006203494966030121,
          "validation_loss": 0.005195828154683113,
          "train_metric": 0.0006247933488339186,
          "validation_metric": 0.0005233047413639724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3503126049999992
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006417378521291539,
          "train_loss": 0.005572378169745207,
          "validation_loss": 0.0043088579550385475,
          "train_metric": 0.0005612294771708548,
          "validation_metric": 0.000433972425526008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3675698099996225
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.007080948373186402,
          "train_loss": 0.006128726992756128,
          "validation_loss": 0.00446541141718626,
          "train_metric": 0.0006172629073262215,
          "validation_metric": 0.0004497398913372308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3840978480002377
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006899416650412604,
          "train_loss": 0.005498890299350023,
          "validation_loss": 0.0047007291577756405,
          "train_metric": 0.0005538280820474029,
          "validation_metric": 0.00047344021731987596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4000295309997455
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.010936821228824556,
          "train_loss": 0.01833980344235897,
          "validation_loss": 0.014932270161807537,
          "train_metric": 0.0018471179064363241,
          "validation_metric": 0.0015039234422147274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.416374698999789
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.010608561802655458,
          "train_loss": 0.006225930526852608,
          "validation_loss": 0.005035202018916607,
          "train_metric": 0.0006270530284382403,
          "validation_metric": 0.0005071270861662924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.432886184000381
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006480617972556502,
          "train_loss": 0.006028239615261555,
          "validation_loss": 0.004369248636066914,
          "train_metric": 0.0006071422249078751,
          "validation_metric": 0.0004400547768454999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4488358599992353
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.009216629012371413,
          "train_loss": 0.006511844228953123,
          "validation_loss": 0.005704014096409082,
          "train_metric": 0.0006558491731993854,
          "validation_metric": 0.0005744873778894544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.464543527999922
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007850948211853392,
          "train_loss": 0.009311886504292488,
          "validation_loss": 0.007155830971896648,
          "train_metric": 0.0009378591785207391,
          "validation_metric": 0.0007207090966403484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4801817719999235
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.008934477751608938,
          "train_loss": 0.012272932566702366,
          "validation_loss": 0.01282534096390009,
          "train_metric": 0.0012360847322270274,
          "validation_metric": 0.001291721360757947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4960346560001199
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00737155377282761,
          "train_loss": 0.005852990318089724,
          "validation_loss": 0.004597283899784088,
          "train_metric": 0.0005894917994737625,
          "validation_metric": 0.0004630216280929744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5121393190001982
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.006133655791927595,
          "train_loss": 0.005515925120562315,
          "validation_loss": 0.004237538203597069,
          "train_metric": 0.0005555438692681491,
          "validation_metric": 0.0004267893673386425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5290571400000772
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006042832348612137,
          "train_loss": 0.0049070012755692005,
          "validation_loss": 0.004042955581098795,
          "train_metric": 0.0004942152881994843,
          "validation_metric": 0.00040719169192016125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5459359029991901
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.005204449022130575,
          "train_loss": 0.005303549580276012,
          "validation_loss": 0.0037937252782285213,
          "train_metric": 0.0005341540672816336,
          "validation_metric": 0.0003820901329163462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5633182660003513
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006180046577355824,
          "train_loss": 0.006082515232264996,
          "validation_loss": 0.006012081168591976,
          "train_metric": 0.0006126086227595806,
          "validation_metric": 0.0006055147969163954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5794994179996138
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.007441859052050859,
          "train_loss": 0.008779780007898808,
          "validation_loss": 0.006213271990418434,
          "train_metric": 0.0008842672104947269,
          "validation_metric": 0.0006257780478335917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5958545499997854
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.006668586407613475,
          "train_loss": 0.005109636578708887,
          "validation_loss": 0.004553762264549732,
          "train_metric": 0.0005146238836459816,
          "validation_metric": 0.0004586382710840553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.611855827999534
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.009025622712215409,
          "train_loss": 0.006044340785592794,
          "validation_loss": 0.004663453903049231,
          "train_metric": 0.0006087638903409243,
          "validation_metric": 0.00046968599781394005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6276804259996425
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005890286578505766,
          "train_loss": 0.004747289698570967,
          "validation_loss": 0.003448423696681857,
          "train_metric": 0.00047812960110604763,
          "validation_metric": 0.00034731256891973317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6442683579989534
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.005285567880491726,
          "train_loss": 0.005177311599254608,
          "validation_loss": 0.003495278302580118,
          "train_metric": 0.0005214398261159658,
          "validation_metric": 0.00035203166771680117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.660286795000502
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0053627890010830015,
          "train_loss": 0.004558279644697905,
          "validation_loss": 0.0040615759789943695,
          "train_metric": 0.0004590932803694159,
          "validation_metric": 0.00040906708454713225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6765985029996955
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.005066284713393543,
          "train_loss": 0.004958731587976217,
          "validation_loss": 0.0037932542618364096,
          "train_metric": 0.000499425339512527,
          "validation_metric": 0.0003820426354650408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6927950699991925
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004962477920344099,
          "train_loss": 0.004719328135251999,
          "validation_loss": 0.0037969392724335194,
          "train_metric": 0.00047531345626339316,
          "validation_metric": 0.000382413825718686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7088990419997572
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004873314610449597,
          "train_loss": 0.004178574774414301,
          "validation_loss": 0.003482517320662737,
          "train_metric": 0.0004208507889416069,
          "validation_metric": 0.00035074641346000135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7247776509993855
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0052191740469424985,
          "train_loss": 0.004869947675615549,
          "validation_loss": 0.0042724465020000935,
          "train_metric": 0.0004904833040200174,
          "validation_metric": 0.00043030513916164637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7407105649999721
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004895404661510838,
          "train_loss": 0.004149910062551498,
          "validation_loss": 0.003166336566209793,
          "train_metric": 0.0004179638053756207,
          "validation_metric": 0.0003189018170814961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7570731679988967
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.005768019502284005,
          "train_loss": 0.00506983045488596,
          "validation_loss": 0.004650712013244629,
          "train_metric": 0.0005106147145852447,
          "validation_metric": 0.0004684026644099504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7732458409991523
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.005765228270320222,
          "train_loss": 0.004359415732324123,
          "validation_loss": 0.0034883166663348675,
          "train_metric": 0.00043906443170271814,
          "validation_metric": 0.0003513304691296071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7899568150005507
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005973441817332059,
          "train_loss": 0.004715489223599434,
          "validation_loss": 0.004207362420856953,
          "train_metric": 0.00047492681187577546,
          "validation_metric": 0.0004237501125317067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.806203808000646
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.005162642170034815,
          "train_loss": 0.004261848516762257,
          "validation_loss": 0.0031990520656108856,
          "train_metric": 0.0004292377270758152,
          "validation_metric": 0.0003221968072466552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8223062269989896
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.004465270045329817,
          "train_loss": 0.004232842940837145,
          "validation_loss": 0.003191899275407195,
          "train_metric": 0.00042631648830138147,
          "validation_metric": 0.00032147637102752924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.838768703999449
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004748770137666725,
          "train_loss": 0.006879557855427265,
          "validation_loss": 0.004877622704952955,
          "train_metric": 0.00069288385566324,
          "validation_metric": 0.0004912563017569482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.854920255000252
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005494646236911649,
          "train_loss": 0.004157326649874449,
          "validation_loss": 0.003637386718764901,
          "train_metric": 0.0004187106969766319,
          "validation_metric": 0.00036634429125115275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8716126250001253
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0061596137238666415,
          "train_loss": 0.01087939739227295,
          "validation_loss": 0.010176428593695164,
          "train_metric": 0.0010957331396639347,
          "validation_metric": 0.001024932600557804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8884867269989627
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0053262883448041975,
          "train_loss": 0.005102952476590872,
          "validation_loss": 0.003219985868781805,
          "train_metric": 0.0005139507120475173,
          "validation_metric": 0.00032430514693260193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9091439559997525
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004766024787386414,
          "train_loss": 0.003594722831621766,
          "validation_loss": 0.0028451078105717897,
          "train_metric": 0.0003620473144110292,
          "validation_metric": 0.00028654883499257267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.925720987999739
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004065916029503569,
          "train_loss": 0.006581991910934448,
          "validation_loss": 0.0065535674802958965,
          "train_metric": 0.0006629141862504184,
          "validation_metric": 0.0006600513006560504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9417855339997914
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004911221309157554,
          "train_loss": 0.00416746037080884,
          "validation_loss": 0.003098068991675973,
          "train_metric": 0.0004197313392069191,
          "validation_metric": 0.0003120261535514146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.958550538000054
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004501591378357261,
          "train_loss": 0.0038424625527113676,
          "validation_loss": 0.003060749266296625,
          "train_metric": 0.00038699875585734844,
          "validation_metric": 0.0003082674811594188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.974895485000161
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.004057249549077824,
          "train_loss": 0.004752623848617077,
          "validation_loss": 0.004476326052099466,
          "train_metric": 0.0004786668869201094,
          "validation_metric": 0.00045083920122124255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9912684079990868
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0054652090548188426,
          "train_loss": 0.0038708809297531843,
          "validation_loss": 0.002997540170326829,
          "train_metric": 0.000389860913855955,
          "validation_metric": 0.00030190128018148243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.007179128999269
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.007330891356104985,
          "train_loss": 0.005862218793481588,
          "validation_loss": 0.005248975940048695,
          "train_metric": 0.0005904212011955678,
          "validation_metric": 0.0005286575760692358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.023600185999385
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.013736682129092515,
          "train_loss": 0.015439323149621487,
          "validation_loss": 0.012253369204699993,
          "train_metric": 0.0015549920499324799,
          "validation_metric": 0.0012341145193204284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0400774810004805
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.011613299779128283,
          "train_loss": 0.006985759828239679,
          "validation_loss": 0.00433517899364233,
          "train_metric": 0.0007035801536403596,
          "validation_metric": 0.00043662331881932914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.056568715999674
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.00577308090578299,
          "train_loss": 0.004430924542248249,
          "validation_loss": 0.003574095666408539,
          "train_metric": 0.0004462665237952024,
          "validation_metric": 0.00035996982478536665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0728338030003215
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004578897365718149,
          "train_loss": 0.004321723245084286,
          "validation_loss": 0.003344193333759904,
          "train_metric": 0.00043526815716177225,
          "validation_metric": 0.0003368149045854807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.08947883700057
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.00459577061701566,
          "train_loss": 0.005147032905369997,
          "validation_loss": 0.004615332931280136,
          "train_metric": 0.0005183902685530484,
          "validation_metric": 0.000464839453343302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.105997276999915
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.006455756782088429,
          "train_loss": 0.008314036764204502,
          "validation_loss": 0.006992809474468231,
          "train_metric": 0.0008373594027943909,
          "validation_metric": 0.0007042901124805212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1228448599995318
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.004930981340294238,
          "train_loss": 0.0043556266464293,
          "validation_loss": 0.0029777733143419027,
          "train_metric": 0.00043868282227776945,
          "validation_metric": 0.00029991046176292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1398473720000766
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.005151931196451187,
          "train_loss": 0.005285498686134815,
          "validation_loss": 0.003771676681935787,
          "train_metric": 0.0005323360092006624,
          "validation_metric": 0.00037986948154866695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.157162616998903
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0067250564898131415,
          "train_loss": 0.004452633671462536,
          "validation_loss": 0.0039039028342813253,
          "train_metric": 0.0004484528908506036,
          "validation_metric": 0.0003931868413928896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1745157839995954
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.00488568190485239,
          "train_loss": 0.003696036757901311,
          "validation_loss": 0.0028655272908508778,
          "train_metric": 0.00037225126288831234,
          "validation_metric": 0.00028860539896413684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.192617706999954
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004163727862760425,
          "train_loss": 0.003524382831528783,
          "validation_loss": 0.0030515030957758427,
          "train_metric": 0.000354962918208912,
          "validation_metric": 0.0003073361876886338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2093268939988775
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0048245982397929765,
          "train_loss": 0.006045817397534847,
          "validation_loss": 0.004581610672175884,
          "train_metric": 0.0006089125527068973,
          "validation_metric": 0.00046144300722517073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.225678089000212
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0061344782006926835,
          "train_loss": 0.004396315198391676,
          "validation_loss": 0.0034335777163505554,
          "train_metric": 0.00044278078712522984,
          "validation_metric": 0.00034581738873384893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2423548070000834
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.005690893660357688,
          "train_loss": 0.004204600118100643,
          "validation_loss": 0.003060520626604557,
          "train_metric": 0.0004234719090163708,
          "validation_metric": 0.0003082444309256971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2587785969990364
        }
      ],
      "best": {
        "epoch": 104,
        "updates": 1664
      },
      "best_validation_metric": 0.00028654883499257267,
      "config": {
        "label": "Batch 64",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01692775339472349
    },
    "6a3fa29d2adc_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7750606071203947,
          "train_loss": 0.38536715507507324,
          "validation_loss": 0.30186760425567627,
          "train_metric": 0.03881276398897171,
          "validation_metric": 0.030403001233935356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01205971600029443
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.2060750015079975,
          "train_loss": 0.12688414752483368,
          "validation_loss": 0.1095275804400444,
          "train_metric": 0.01277930662035942,
          "validation_metric": 0.011031216941773891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.023776153999278904
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.1240247250534594,
          "train_loss": 0.09904789179563522,
          "validation_loss": 0.08297248184680939,
          "train_metric": 0.009975740686058998,
          "validation_metric": 0.008356684818863869,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03549776900035795
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09276640950702131,
          "train_loss": 0.0849667638540268,
          "validation_loss": 0.06827860325574875,
          "train_metric": 0.00855754129588604,
          "validation_metric": 0.006876770872622728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.047386379999807104
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07504639495164156,
          "train_loss": 0.06760113686323166,
          "validation_loss": 0.055644359439611435,
          "train_metric": 0.006808537989854813,
          "validation_metric": 0.005604295991361141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.059080363000248326
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06591126136481762,
          "train_loss": 0.0652155801653862,
          "validation_loss": 0.052817028015851974,
          "train_metric": 0.0065682739950716496,
          "validation_metric": 0.005319537594914436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07078410699978122
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06305562076158822,
          "train_loss": 0.06514137238264084,
          "validation_loss": 0.05300118774175644,
          "train_metric": 0.00656080013141036,
          "validation_metric": 0.005338085815310478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08212411900058214
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06273406581021845,
          "train_loss": 0.05920339375734329,
          "validation_loss": 0.04777815192937851,
          "train_metric": 0.005962749011814594,
          "validation_metric": 0.0048120408318936825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0937143159990228
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05564656085334718,
          "train_loss": 0.052560701966285706,
          "validation_loss": 0.0434177927672863,
          "train_metric": 0.005293721333146095,
          "validation_metric": 0.004372881259769201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10541270700014138
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05131560959853232,
          "train_loss": 0.051526989787817,
          "validation_loss": 0.04277803748846054,
          "train_metric": 0.0051896097138524055,
          "validation_metric": 0.004308447707444429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11732196800039674
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04965876357164234,
          "train_loss": 0.0480007603764534,
          "validation_loss": 0.03882240504026413,
          "train_metric": 0.004834461025893688,
          "validation_metric": 0.003910050727427006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1293284940002195
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04525484796613455,
          "train_loss": 0.041656482964754105,
          "validation_loss": 0.034643612802028656,
          "train_metric": 0.004195488058030605,
          "validation_metric": 0.0034891776740550995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14133813499938697
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.0406425753608346,
          "train_loss": 0.03904619812965393,
          "validation_loss": 0.032580845057964325,
          "train_metric": 0.003932590130716562,
          "validation_metric": 0.003281423356384039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15303046699955303
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03730706369969994,
          "train_loss": 0.03512808308005333,
          "validation_loss": 0.029410352930426598,
          "train_metric": 0.003537971992045641,
          "validation_metric": 0.0029621028807014227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16481036900040635
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03614106320310384,
          "train_loss": 0.03985850140452385,
          "validation_loss": 0.033791474997997284,
          "train_metric": 0.004014401696622372,
          "validation_metric": 0.0034033532720059156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1759062309993169
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.034006054629571736,
          "train_loss": 0.03577851876616478,
          "validation_loss": 0.03025796450674534,
          "train_metric": 0.0036034814547747374,
          "validation_metric": 0.0030474714003503323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18698433999998088
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.030239578511100262,
          "train_loss": 0.02710607275366783,
          "validation_loss": 0.023179994896054268,
          "train_metric": 0.0027300245128571987,
          "validation_metric": 0.00233460427261889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19896924199929344
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029969822440762073,
          "train_loss": 0.03654274344444275,
          "validation_loss": 0.028627360239624977,
          "train_metric": 0.003680451540276408,
          "validation_metric": 0.0028832426760345697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2101544750003086
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.031163847947027534,
          "train_loss": 0.02538416162133217,
          "validation_loss": 0.019166603684425354,
          "train_metric": 0.002556599909439683,
          "validation_metric": 0.0019303900189697742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22164486099973146
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.023735167924314737,
          "train_loss": 0.02233070135116577,
          "validation_loss": 0.018863607197999954,
          "train_metric": 0.0022490667179226875,
          "validation_metric": 0.001899873255752027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2338336220000201
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023142108839238063,
          "train_loss": 0.01990034058690071,
          "validation_loss": 0.01686880923807621,
          "train_metric": 0.002004289533942938,
          "validation_metric": 0.0016989647410809994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24565626799994789
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02312045422149822,
          "train_loss": 0.020990870893001556,
          "validation_loss": 0.016307419165968895,
          "train_metric": 0.002114123897626996,
          "validation_metric": 0.0016424234490841627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2575671600006899
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.022139596985653043,
          "train_loss": 0.020363621413707733,
          "validation_loss": 0.01820952072739601,
          "train_metric": 0.0020509492605924606,
          "validation_metric": 0.0018339961534366012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2688493349996861
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.018391056975815445,
          "train_loss": 0.01689496450126171,
          "validation_loss": 0.01443193294107914,
          "train_metric": 0.001701599103398621,
          "validation_metric": 0.001453531440347433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.280361241000719
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019256686326116323,
          "train_loss": 0.019413551315665245,
          "validation_loss": 0.014614056795835495,
          "train_metric": 0.0019552616868168116,
          "validation_metric": 0.0014718743041157722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2914370960006636
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.018628374149557203,
          "train_loss": 0.021552782505750656,
          "validation_loss": 0.02029510959982872,
          "train_metric": 0.002170717343688011,
          "validation_metric": 0.0020440490916371346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30245343100068567
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.017534870916279033,
          "train_loss": 0.015243593603372574,
          "validation_loss": 0.013040601275861263,
          "train_metric": 0.001535278744995594,
          "validation_metric": 0.001313401386141777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3154404559991235
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015402383840410039,
          "train_loss": 0.014646788127720356,
          "validation_loss": 0.012762422673404217,
          "train_metric": 0.0014751707203686237,
          "validation_metric": 0.0012853844091296196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32721743900037836
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016872672975296155,
          "train_loss": 0.018696356564760208,
          "validation_loss": 0.015021326020359993,
          "train_metric": 0.0018830286571756005,
          "validation_metric": 0.001512893009930849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3385624629991071
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.016743629763368517,
          "train_loss": 0.01551545225083828,
          "validation_loss": 0.013348198495805264,
          "train_metric": 0.0015626593958586454,
          "validation_metric": 0.001344381715171039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.349595424000654
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.021813345549162477,
          "train_loss": 0.014776387251913548,
          "validation_loss": 0.011951805092394352,
          "train_metric": 0.001488223671913147,
          "validation_metric": 0.0012037419946864247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3610829200006265
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.018929659272544086,
          "train_loss": 0.014764338731765747,
          "validation_loss": 0.010697023011744022,
          "train_metric": 0.0014870099257677794,
          "validation_metric": 0.001077365013770759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37333930599925225
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.0153573855641298,
          "train_loss": 0.012026515789330006,
          "validation_loss": 0.009608716703951359,
          "train_metric": 0.0012112664990127087,
          "validation_metric": 0.0009677548077888787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38502799500020046
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013202551752328873,
          "train_loss": 0.01557985134422779,
          "validation_loss": 0.010827160440385342,
          "train_metric": 0.0015691454755142331,
          "validation_metric": 0.0010904718656092882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3961450139995577
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014358282889588736,
          "train_loss": 0.015606962144374847,
          "validation_loss": 0.014100464060902596,
          "train_metric": 0.0015718759968876839,
          "validation_metric": 0.001420147018507123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4072805010000593
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.016984775924356654,
          "train_loss": 0.02222522161900997,
          "validation_loss": 0.01624390110373497,
          "train_metric": 0.002238442888483405,
          "validation_metric": 0.001636026194319129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4182149410007696
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.012299630441702902,
          "train_loss": 0.011116322129964828,
          "validation_loss": 0.009583269245922565,
          "train_metric": 0.0011195952538400888,
          "validation_metric": 0.0009651916334405541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4296825790006551
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011669085710309446,
          "train_loss": 0.009766273200511932,
          "validation_loss": 0.007932213135063648,
          "train_metric": 0.0009836232056841254,
          "validation_metric": 0.0007989035220816731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44127167300030123
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01077001288649626,
          "train_loss": 0.009837810881435871,
          "validation_loss": 0.00763594638556242,
          "train_metric": 0.0009908282663673162,
          "validation_metric": 0.0007690644706599414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45293590700021014
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012661474363994785,
          "train_loss": 0.012773655354976654,
          "validation_loss": 0.009138710796833038,
          "train_metric": 0.0012865157332271338,
          "validation_metric": 0.0009204175439663231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46410955999999715
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010784950369270518,
          "train_loss": 0.009931527078151703,
          "validation_loss": 0.007522317115217447,
          "train_metric": 0.0010002671042457223,
          "validation_metric": 0.0007576202624477446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4758265130003565
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010854612351977266,
          "train_loss": 0.009563778527081013,
          "validation_loss": 0.007020589429885149,
          "train_metric": 0.0009632286964915693,
          "validation_metric": 0.0007070880965329707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4876189179994981
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011339242788380943,
          "train_loss": 0.0087605444714427,
          "validation_loss": 0.007057554554194212,
          "train_metric": 0.0008823300013318658,
          "validation_metric": 0.0007108110003173351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4987366599998495
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.00978471252892632,
          "train_loss": 0.00902597513049841,
          "validation_loss": 0.0073502580635249615,
          "train_metric": 0.0009090631501749158,
          "validation_metric": 0.0007402910850942135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5098059569991165
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013840032886946574,
          "train_loss": 0.010268403217196465,
          "validation_loss": 0.008520981296896935,
          "train_metric": 0.0010341958841308951,
          "validation_metric": 0.0008582019945606589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5209284680004203
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.012792111374437809,
          "train_loss": 0.010196718387305737,
          "validation_loss": 0.007601565681397915,
          "train_metric": 0.0010269759222865105,
          "validation_metric": 0.0007656019297428429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5318509709995851
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011586189008085057,
          "train_loss": 0.012726811692118645,
          "validation_loss": 0.011708073318004608,
          "train_metric": 0.0012817978858947754,
          "validation_metric": 0.00117919419426471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5427864560006128
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.009888891261653043,
          "train_loss": 0.014078420586884022,
          "validation_loss": 0.013561593368649483,
          "train_metric": 0.0014179269783198833,
          "validation_metric": 0.0013658740790560842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5538756759997341
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.01211502478690818,
          "train_loss": 0.010559041053056717,
          "validation_loss": 0.008403655141592026,
          "train_metric": 0.001063467818312347,
          "validation_metric": 0.0008463853155262768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5646860840006411
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.011117999012640212,
          "train_loss": 0.00816005002707243,
          "validation_loss": 0.006046057213097811,
          "train_metric": 0.0008218504954129457,
          "validation_metric": 0.0006089368253014982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5761580069993215
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.00928575212310534,
          "train_loss": 0.012129377573728561,
          "validation_loss": 0.01141818892210722,
          "train_metric": 0.0012216264149174094,
          "validation_metric": 0.0011499980464577675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5875406760005717
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008625807182397693,
          "train_loss": 0.00884421169757843,
          "validation_loss": 0.0063227578066289425,
          "train_metric": 0.0008907566079869866,
          "validation_metric": 0.0006368050235323608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5988813749991095
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008779473515460268,
          "train_loss": 0.010252740234136581,
          "validation_loss": 0.00750874076038599,
          "train_metric": 0.0010326184565201402,
          "validation_metric": 0.0007562528480775654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6100532210002712
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0091602051834343,
          "train_loss": 0.01068117842078209,
          "validation_loss": 0.010125876404345036,
          "train_metric": 0.0010757691925391555,
          "validation_metric": 0.0010198411764577031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6210301089995482
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.013468224395182915,
          "train_loss": 0.00999869592487812,
          "validation_loss": 0.007742003537714481,
          "train_metric": 0.0010070321150124073,
          "validation_metric": 0.0007797462749294937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6321296290007012
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.010853008774574846,
          "train_loss": 0.008161451667547226,
          "validation_loss": 0.0058862557634711266,
          "train_metric": 0.0008219915325753391,
          "validation_metric": 0.0005928422324359417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.64391735999925
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008787122336798348,
          "train_loss": 0.007462658453732729,
          "validation_loss": 0.006129884626716375,
          "train_metric": 0.0007516115438193083,
          "validation_metric": 0.0006173795554786921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6551671519991942
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007215693418402225,
          "train_loss": 0.006528432480990887,
          "validation_loss": 0.005115540232509375,
          "train_metric": 0.0006575197330676019,
          "validation_metric": 0.0005152184749022126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6666349829993123
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007845858577638865,
          "train_loss": 0.0069840820506215096,
          "validation_loss": 0.006143107544630766,
          "train_metric": 0.0007034111185930669,
          "validation_metric": 0.0006187112885527313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6778855249995104
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007135726671549492,
          "train_loss": 0.006284323520958424,
          "validation_loss": 0.005035600624978542,
          "train_metric": 0.0006329340394586325,
          "validation_metric": 0.0005071673076599836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6897345390007104
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.009110017359489575,
          "train_loss": 0.006696862168610096,
          "validation_loss": 0.005621404852718115,
          "train_metric": 0.0006744833663105965,
          "validation_metric": 0.0005661673494614661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7008925240006647
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007818591664545238,
          "train_loss": 0.006637724116444588,
          "validation_loss": 0.005256288219243288,
          "train_metric": 0.0006685271509923041,
          "validation_metric": 0.0005293941358104348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7119962760007184
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006970037764403969,
          "train_loss": 0.006174634676426649,
          "validation_loss": 0.004937969613820314,
          "train_metric": 0.0006218865746632218,
          "validation_metric": 0.0004973342875018716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7235149000007368
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007135511885280721,
          "train_loss": 0.009442456066608429,
          "validation_loss": 0.008755910210311413,
          "train_metric": 0.0009510095696896315,
          "validation_metric": 0.0008818632340990007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7345061949999945
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008260824164608493,
          "train_loss": 0.008309222757816315,
          "validation_loss": 0.0070204706862568855,
          "train_metric": 0.0008368745329789817,
          "validation_metric": 0.0007070761639624834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7453087979993143
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0069822732475586236,
          "train_loss": 0.0067450315691530704,
          "validation_loss": 0.004807147663086653,
          "train_metric": 0.000679334974847734,
          "validation_metric": 0.0004841582849621773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7570796189993416
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008227432139392477,
          "train_loss": 0.00827042106539011,
          "validation_loss": 0.005731015466153622,
          "train_metric": 0.0008329665288329124,
          "validation_metric": 0.0005772068398073316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7681979209992278
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00791827542707324,
          "train_loss": 0.006597149185836315,
          "validation_loss": 0.00453611183911562,
          "train_metric": 0.000664440740365535,
          "validation_metric": 0.00045686058001592755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7802309620001324
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0060867796710226685,
          "train_loss": 0.006203494966030121,
          "validation_loss": 0.005195828154683113,
          "train_metric": 0.0006247933488339186,
          "validation_metric": 0.0005233047413639724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7920725149997452
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006417378521291539,
          "train_loss": 0.005572378169745207,
          "validation_loss": 0.0043088579550385475,
          "train_metric": 0.0005612294771708548,
          "validation_metric": 0.000433972425526008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8043815670007461
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.007080948373186402,
          "train_loss": 0.006128726992756128,
          "validation_loss": 0.00446541141718626,
          "train_metric": 0.0006172629073262215,
          "validation_metric": 0.0004497398913372308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.815526016000149
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006899416650412604,
          "train_loss": 0.005498890299350023,
          "validation_loss": 0.0047007291577756405,
          "train_metric": 0.0005538280820474029,
          "validation_metric": 0.00047344021731987596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8268972519999807
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.010936821228824556,
          "train_loss": 0.01833980344235897,
          "validation_loss": 0.014932270161807537,
          "train_metric": 0.0018471179064363241,
          "validation_metric": 0.0015039234422147274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8383149559995218
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.010608561802655458,
          "train_loss": 0.006225930526852608,
          "validation_loss": 0.005035202018916607,
          "train_metric": 0.0006270530284382403,
          "validation_metric": 0.0005071270861662924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8501039709990437
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006480617972556502,
          "train_loss": 0.006028239615261555,
          "validation_loss": 0.004369248636066914,
          "train_metric": 0.0006071422249078751,
          "validation_metric": 0.0004400547768454999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8614587280007981
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.009216629012371413,
          "train_loss": 0.006511844228953123,
          "validation_loss": 0.005704014096409082,
          "train_metric": 0.0006558491731993854,
          "validation_metric": 0.0005744873778894544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8727113099994313
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007850948211853392,
          "train_loss": 0.009311886504292488,
          "validation_loss": 0.007155830971896648,
          "train_metric": 0.0009378591785207391,
          "validation_metric": 0.0007207090966403484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8837972929995885
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.008934477751608938,
          "train_loss": 0.012272932566702366,
          "validation_loss": 0.01282534096390009,
          "train_metric": 0.0012360847322270274,
          "validation_metric": 0.001291721360757947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8951808189995063
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00737155377282761,
          "train_loss": 0.005852990318089724,
          "validation_loss": 0.004597283899784088,
          "train_metric": 0.0005894917994737625,
          "validation_metric": 0.0004630216280929744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9067963830002554
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.006133655791927595,
          "train_loss": 0.005515925120562315,
          "validation_loss": 0.004237538203597069,
          "train_metric": 0.0005555438692681491,
          "validation_metric": 0.0004267893673386425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9188077389990212
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006042832348612137,
          "train_loss": 0.0049070012755692005,
          "validation_loss": 0.004042955581098795,
          "train_metric": 0.0004942152881994843,
          "validation_metric": 0.00040719169192016125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9311244509990502
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.005204449022130575,
          "train_loss": 0.005303549580276012,
          "validation_loss": 0.0037937252782285213,
          "train_metric": 0.0005341540672816336,
          "validation_metric": 0.0003820901329163462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9432462839995424
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006180046577355824,
          "train_loss": 0.006082515232264996,
          "validation_loss": 0.006012081168591976,
          "train_metric": 0.0006126086227595806,
          "validation_metric": 0.0006055147969163954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9551863500000763
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.007441859052050859,
          "train_loss": 0.008779780007898808,
          "validation_loss": 0.006213271990418434,
          "train_metric": 0.0008842672104947269,
          "validation_metric": 0.0006257780478335917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9667473520003114
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.006668586407613475,
          "train_loss": 0.005109636578708887,
          "validation_loss": 0.004553762264549732,
          "train_metric": 0.0005146238836459816,
          "validation_metric": 0.0004586382710840553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9782986619993608
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.009025622712215409,
          "train_loss": 0.006044340785592794,
          "validation_loss": 0.004663453903049231,
          "train_metric": 0.0006087638903409243,
          "validation_metric": 0.00046968599781394005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9902161120007804
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005890286578505766,
          "train_loss": 0.004747289698570967,
          "validation_loss": 0.003448423696681857,
          "train_metric": 0.00047812960110604763,
          "validation_metric": 0.00034731256891973317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0022315650003293
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.005285567880491726,
          "train_loss": 0.005177311599254608,
          "validation_loss": 0.003495278302580118,
          "train_metric": 0.0005214398261159658,
          "validation_metric": 0.00035203166771680117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0135428869998577
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0053627890010830015,
          "train_loss": 0.004558279644697905,
          "validation_loss": 0.0040615759789943695,
          "train_metric": 0.0004590932803694159,
          "validation_metric": 0.00040906708454713225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.025240880999263
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.005066284713393543,
          "train_loss": 0.004958731587976217,
          "validation_loss": 0.0037932542618364096,
          "train_metric": 0.000499425339512527,
          "validation_metric": 0.0003820426354650408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0366485659997124
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004962477920344099,
          "train_loss": 0.004719328135251999,
          "validation_loss": 0.0037969392724335194,
          "train_metric": 0.00047531345626339316,
          "validation_metric": 0.000382413825718686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.048399191999124
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004873314610449597,
          "train_loss": 0.004178574774414301,
          "validation_loss": 0.003482517320662737,
          "train_metric": 0.0004208507889416069,
          "validation_metric": 0.00035074641346000135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0596983619998355
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0052191740469424985,
          "train_loss": 0.004869947675615549,
          "validation_loss": 0.0042724465020000935,
          "train_metric": 0.0004904833040200174,
          "validation_metric": 0.00043030513916164637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0712497149997944
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004895404661510838,
          "train_loss": 0.004149910062551498,
          "validation_loss": 0.003166336566209793,
          "train_metric": 0.0004179638053756207,
          "validation_metric": 0.0003189018170814961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0833640260007087
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.005768019502284005,
          "train_loss": 0.00506983045488596,
          "validation_loss": 0.004650712013244629,
          "train_metric": 0.0005106147145852447,
          "validation_metric": 0.0004684026644099504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0949218820005626
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.005765228270320222,
          "train_loss": 0.004359415732324123,
          "validation_loss": 0.0034883166663348675,
          "train_metric": 0.00043906443170271814,
          "validation_metric": 0.0003513304691296071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1064119279999431
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005973441817332059,
          "train_loss": 0.004715489223599434,
          "validation_loss": 0.004207362420856953,
          "train_metric": 0.00047492681187577546,
          "validation_metric": 0.0004237501125317067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1180634669999563
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.005162642170034815,
          "train_loss": 0.004261848516762257,
          "validation_loss": 0.0031990520656108856,
          "train_metric": 0.0004292377270758152,
          "validation_metric": 0.0003221968072466552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1295160120007495
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.004465270045329817,
          "train_loss": 0.004232842940837145,
          "validation_loss": 0.003191899275407195,
          "train_metric": 0.00042631648830138147,
          "validation_metric": 0.00032147637102752924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1408185119998961
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004748770137666725,
          "train_loss": 0.006879557855427265,
          "validation_loss": 0.004877622704952955,
          "train_metric": 0.00069288385566324,
          "validation_metric": 0.0004912563017569482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.152049435999288
        }
      ],
      "best": {
        "epoch": 94,
        "updates": 1504
      },
      "best_validation_metric": 0.0003189018170814961,
      "config": {
        "label": "Batch 64",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01785782229392756
    },
    "50c1636b4f2c_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 4,
          "samples_seen": 1024,
          "online_loss": 0.8927699774503708,
          "train_loss": 0.7838672399520874,
          "validation_loss": 0.5631930232048035,
          "train_metric": 0.0789482444524765,
          "validation_metric": 0.0567227378487587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.012258438999197097
        },
        {
          "epoch": 2,
          "updates": 8,
          "samples_seen": 2048,
          "online_loss": 0.7028519213199615,
          "train_loss": 0.5566926598548889,
          "validation_loss": 0.3958049714565277,
          "train_metric": 0.056068047881126404,
          "validation_metric": 0.03986402601003647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02252566300012404
        },
        {
          "epoch": 3,
          "updates": 12,
          "samples_seen": 3072,
          "online_loss": 0.46128150075674057,
          "train_loss": 0.31972208619117737,
          "validation_loss": 0.21971434354782104,
          "train_metric": 0.032201237976551056,
          "validation_metric": 0.022128824144601822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03276968699901772
        },
        {
          "epoch": 4,
          "updates": 16,
          "samples_seen": 4096,
          "online_loss": 0.27318407595157623,
          "train_loss": 0.23402482271194458,
          "validation_loss": 0.1558881551027298,
          "train_metric": 0.023570124059915543,
          "validation_metric": 0.015700483694672585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.042568454999127425
        },
        {
          "epoch": 5,
          "updates": 20,
          "samples_seen": 5120,
          "online_loss": 0.23123947530984879,
          "train_loss": 0.22038634121418,
          "validation_loss": 0.1731768697500229,
          "train_metric": 0.02219650335609913,
          "validation_metric": 0.01744174212217331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05323017400041863
        },
        {
          "epoch": 6,
          "updates": 24,
          "samples_seen": 6144,
          "online_loss": 0.21517503634095192,
          "train_loss": 0.2001335173845291,
          "validation_loss": 0.1783643513917923,
          "train_metric": 0.02015671506524086,
          "validation_metric": 0.017964206635951996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16279101799955242
        },
        {
          "epoch": 7,
          "updates": 28,
          "samples_seen": 7168,
          "online_loss": 0.18522999808192253,
          "train_loss": 0.1599196195602417,
          "validation_loss": 0.13869014382362366,
          "train_metric": 0.016106517985463142,
          "validation_metric": 0.013968363404273987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18332111000017903
        },
        {
          "epoch": 8,
          "updates": 32,
          "samples_seen": 8192,
          "online_loss": 0.15052815526723862,
          "train_loss": 0.14103472232818604,
          "validation_loss": 0.11476744711399078,
          "train_metric": 0.014204500243067741,
          "validation_metric": 0.0115589564666152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2074929139998858
        },
        {
          "epoch": 9,
          "updates": 36,
          "samples_seen": 9216,
          "online_loss": 0.13794264197349548,
          "train_loss": 0.13377687335014343,
          "validation_loss": 0.1090945452451706,
          "train_metric": 0.013473517261445522,
          "validation_metric": 0.010987603105604649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23131889599972055
        },
        {
          "epoch": 10,
          "updates": 40,
          "samples_seen": 10240,
          "online_loss": 0.1300338003784418,
          "train_loss": 0.12288288027048111,
          "validation_loss": 0.10423184931278229,
          "train_metric": 0.012376313097774982,
          "validation_metric": 0.010497849434614182,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25504024399924674
        },
        {
          "epoch": 11,
          "updates": 44,
          "samples_seen": 11264,
          "online_loss": 0.11971715651452541,
          "train_loss": 0.11575634777545929,
          "validation_loss": 0.10013993829488754,
          "train_metric": 0.011658554896712303,
          "validation_metric": 0.010085728019475937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2766418179999164
        },
        {
          "epoch": 12,
          "updates": 48,
          "samples_seen": 12288,
          "online_loss": 0.11496703699231148,
          "train_loss": 0.11185318976640701,
          "validation_loss": 0.09616605937480927,
          "train_metric": 0.011265442706644535,
          "validation_metric": 0.00968549307435751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3004765829991811
        },
        {
          "epoch": 13,
          "updates": 52,
          "samples_seen": 13312,
          "online_loss": 0.10953292436897755,
          "train_loss": 0.10469567775726318,
          "validation_loss": 0.08836554735898972,
          "train_metric": 0.010544564574956894,
          "validation_metric": 0.008899855427443981,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3240976029992453
        },
        {
          "epoch": 14,
          "updates": 56,
          "samples_seen": 14336,
          "online_loss": 0.10282902792096138,
          "train_loss": 0.0995367169380188,
          "validation_loss": 0.0829203873872757,
          "train_metric": 0.010024972259998322,
          "validation_metric": 0.00835143867880106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3488508329992328
        },
        {
          "epoch": 15,
          "updates": 60,
          "samples_seen": 15360,
          "online_loss": 0.0973251722753048,
          "train_loss": 0.09327718615531921,
          "validation_loss": 0.07762794941663742,
          "train_metric": 0.009394535794854164,
          "validation_metric": 0.007818402722477913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3726027709999471
        },
        {
          "epoch": 16,
          "updates": 64,
          "samples_seen": 16384,
          "online_loss": 0.09105087257921696,
          "train_loss": 0.08781687915325165,
          "validation_loss": 0.07299760729074478,
          "train_metric": 0.008844595402479172,
          "validation_metric": 0.007352051325142384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3960897489996569
        },
        {
          "epoch": 17,
          "updates": 68,
          "samples_seen": 17408,
          "online_loss": 0.08612624928355217,
          "train_loss": 0.08305052667856216,
          "validation_loss": 0.0685667097568512,
          "train_metric": 0.008364544250071049,
          "validation_metric": 0.006905788090080023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41808085099910386
        },
        {
          "epoch": 18,
          "updates": 72,
          "samples_seen": 18432,
          "online_loss": 0.0811981875449419,
          "train_loss": 0.07795611023902893,
          "validation_loss": 0.0636921375989914,
          "train_metric": 0.007851453498005867,
          "validation_metric": 0.006414839066565037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44241936599973997
        },
        {
          "epoch": 19,
          "updates": 76,
          "samples_seen": 19456,
          "online_loss": 0.07647538371384144,
          "train_loss": 0.07393479347229004,
          "validation_loss": 0.05989176779985428,
          "train_metric": 0.007446441799402237,
          "validation_metric": 0.006032078992575407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4669429539990233
        },
        {
          "epoch": 20,
          "updates": 80,
          "samples_seen": 20480,
          "online_loss": 0.07253853883594275,
          "train_loss": 0.07020120322704315,
          "validation_loss": 0.05688726156949997,
          "train_metric": 0.007070409134030342,
          "validation_metric": 0.005729476921260357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49097546600023634
        },
        {
          "epoch": 21,
          "updates": 84,
          "samples_seen": 21504,
          "online_loss": 0.06900657713413239,
          "train_loss": 0.06720419973134995,
          "validation_loss": 0.054863519966602325,
          "train_metric": 0.006768560968339443,
          "validation_metric": 0.005525652784854174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5184421419999126
        },
        {
          "epoch": 22,
          "updates": 88,
          "samples_seen": 22528,
          "online_loss": 0.06631501391530037,
          "train_loss": 0.06486304849386215,
          "validation_loss": 0.05314850062131882,
          "train_metric": 0.006532769650220871,
          "validation_metric": 0.005352921783924103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.53243857400048
        },
        {
          "epoch": 23,
          "updates": 92,
          "samples_seen": 23552,
          "online_loss": 0.0643466031178832,
          "train_loss": 0.06311765313148499,
          "validation_loss": 0.05190986767411232,
          "train_metric": 0.006356978788971901,
          "validation_metric": 0.005228172056376934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5428908319991024
        },
        {
          "epoch": 24,
          "updates": 96,
          "samples_seen": 24576,
          "online_loss": 0.06278229411691427,
          "train_loss": 0.061847228556871414,
          "validation_loss": 0.0508357472717762,
          "train_metric": 0.006229025777429342,
          "validation_metric": 0.005119990557432175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5527645499987557
        },
        {
          "epoch": 25,
          "updates": 100,
          "samples_seen": 25600,
          "online_loss": 0.061596945859491825,
          "train_loss": 0.06071282923221588,
          "validation_loss": 0.050245556980371475,
          "train_metric": 0.0061147743836045265,
          "validation_metric": 0.0050605484284460545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5626102349997382
        },
        {
          "epoch": 26,
          "updates": 104,
          "samples_seen": 26624,
          "online_loss": 0.0606892304494977,
          "train_loss": 0.0598691925406456,
          "validation_loss": 0.049954816699028015,
          "train_metric": 0.006029805634170771,
          "validation_metric": 0.005031266715377569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5730075489991577
        },
        {
          "epoch": 27,
          "updates": 108,
          "samples_seen": 27648,
          "online_loss": 0.05940595269203186,
          "train_loss": 0.05859355628490448,
          "validation_loss": 0.04851940646767616,
          "train_metric": 0.005901328753679991,
          "validation_metric": 0.004886697046458721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5833691999996518
        },
        {
          "epoch": 28,
          "updates": 112,
          "samples_seen": 28672,
          "online_loss": 0.058253493160009384,
          "train_loss": 0.05754619836807251,
          "validation_loss": 0.04745740443468094,
          "train_metric": 0.005795842967927456,
          "validation_metric": 0.004779735580086708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.593608465998841
        },
        {
          "epoch": 29,
          "updates": 116,
          "samples_seen": 29696,
          "online_loss": 0.05721741821616888,
          "train_loss": 0.0562891885638237,
          "validation_loss": 0.04667796194553375,
          "train_metric": 0.005669239908456802,
          "validation_metric": 0.004701233468949795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6035992130000523
        },
        {
          "epoch": 30,
          "updates": 120,
          "samples_seen": 30720,
          "online_loss": 0.056246569380164146,
          "train_loss": 0.05515860393643379,
          "validation_loss": 0.04598258063197136,
          "train_metric": 0.0055553726851940155,
          "validation_metric": 0.004631197080016136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6124358830002166
        },
        {
          "epoch": 31,
          "updates": 124,
          "samples_seen": 31744,
          "online_loss": 0.05458447989076376,
          "train_loss": 0.05390247330069542,
          "validation_loss": 0.04469452425837517,
          "train_metric": 0.0054288599640131,
          "validation_metric": 0.00450146896764636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6197655529995245
        },
        {
          "epoch": 32,
          "updates": 128,
          "samples_seen": 32768,
          "online_loss": 0.05342755187302828,
          "train_loss": 0.05251681059598923,
          "validation_loss": 0.04360233619809151,
          "train_metric": 0.00528930127620697,
          "validation_metric": 0.004391467198729515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6268064209998556
        },
        {
          "epoch": 33,
          "updates": 132,
          "samples_seen": 33792,
          "online_loss": 0.05207055900245905,
          "train_loss": 0.051131974905729294,
          "validation_loss": 0.042406003922224045,
          "train_metric": 0.0051498254761099815,
          "validation_metric": 0.0042709773406386375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6337615359989286
        },
        {
          "epoch": 34,
          "updates": 136,
          "samples_seen": 34816,
          "online_loss": 0.05105511471629143,
          "train_loss": 0.04972994327545166,
          "validation_loss": 0.04166335612535477,
          "train_metric": 0.005008617881685495,
          "validation_metric": 0.004196180496364832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6463960839992069
        },
        {
          "epoch": 35,
          "updates": 140,
          "samples_seen": 35840,
          "online_loss": 0.0498435040935874,
          "train_loss": 0.0484432578086853,
          "validation_loss": 0.040269363671541214,
          "train_metric": 0.004879027605056763,
          "validation_metric": 0.004055782686918974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6596227640002326
        },
        {
          "epoch": 36,
          "updates": 144,
          "samples_seen": 36864,
          "online_loss": 0.04783166106790304,
          "train_loss": 0.046853359788656235,
          "validation_loss": 0.039341624826192856,
          "train_metric": 0.004718898795545101,
          "validation_metric": 0.003962344489991665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6674904760002391
        },
        {
          "epoch": 37,
          "updates": 148,
          "samples_seen": 37888,
          "online_loss": 0.04643342411145568,
          "train_loss": 0.045086611062288284,
          "validation_loss": 0.03785566985607147,
          "train_metric": 0.004540957976132631,
          "validation_metric": 0.003812684677541256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6759430689999135
        },
        {
          "epoch": 38,
          "updates": 152,
          "samples_seen": 38912,
          "online_loss": 0.044761464931070805,
          "train_loss": 0.04367851838469505,
          "validation_loss": 0.03652653098106384,
          "train_metric": 0.004399139899760485,
          "validation_metric": 0.0036788186989724636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6833259529994393
        },
        {
          "epoch": 39,
          "updates": 156,
          "samples_seen": 39936,
          "online_loss": 0.04315029922872782,
          "train_loss": 0.04187144711613655,
          "validation_loss": 0.03517572581768036,
          "train_metric": 0.004217138513922691,
          "validation_metric": 0.00354277016595006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.690289711999867
        },
        {
          "epoch": 40,
          "updates": 160,
          "samples_seen": 40960,
          "online_loss": 0.041457753628492355,
          "train_loss": 0.040091127157211304,
          "validation_loss": 0.03354387357831001,
          "train_metric": 0.004037830978631973,
          "validation_metric": 0.0033784157130867243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6984388379987649
        },
        {
          "epoch": 41,
          "updates": 164,
          "samples_seen": 41984,
          "online_loss": 0.03959817672148347,
          "train_loss": 0.0383581817150116,
          "validation_loss": 0.03263454884290695,
          "train_metric": 0.003863295540213585,
          "validation_metric": 0.0032868324778974056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7060889059994224
        },
        {
          "epoch": 42,
          "updates": 168,
          "samples_seen": 43008,
          "online_loss": 0.03794359741732478,
          "train_loss": 0.03656475991010666,
          "validation_loss": 0.03107885830104351,
          "train_metric": 0.003682668786495924,
          "validation_metric": 0.003130148397758603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7134715419997519
        },
        {
          "epoch": 43,
          "updates": 172,
          "samples_seen": 44032,
          "online_loss": 0.03702738834545016,
          "train_loss": 0.03505612537264824,
          "validation_loss": 0.029719127342104912,
          "train_metric": 0.0035307249054312706,
          "validation_metric": 0.00299320206977427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7205674289998569
        },
        {
          "epoch": 44,
          "updates": 176,
          "samples_seen": 45056,
          "online_loss": 0.0352835925295949,
          "train_loss": 0.034563034772872925,
          "validation_loss": 0.028098957613110542,
          "train_metric": 0.003481062361970544,
          "validation_metric": 0.0028300245758146048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7290376929995546
        },
        {
          "epoch": 45,
          "updates": 180,
          "samples_seen": 46080,
          "online_loss": 0.03477455209940672,
          "train_loss": 0.03313404321670532,
          "validation_loss": 0.028732705861330032,
          "train_metric": 0.00333713972941041,
          "validation_metric": 0.0028938530012965202,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7357397849991685
        },
        {
          "epoch": 46,
          "updates": 184,
          "samples_seen": 47104,
          "online_loss": 0.03174919635057449,
          "train_loss": 0.031500764191150665,
          "validation_loss": 0.02603911980986595,
          "train_metric": 0.00317264162003994,
          "validation_metric": 0.00262256502173841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7498889260004944
        },
        {
          "epoch": 47,
          "updates": 188,
          "samples_seen": 48128,
          "online_loss": 0.03065170720219612,
          "train_loss": 0.029573338106274605,
          "validation_loss": 0.024837631732225418,
          "train_metric": 0.002978518372401595,
          "validation_metric": 0.002501555485650897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7591088469998795
        },
        {
          "epoch": 48,
          "updates": 192,
          "samples_seen": 49152,
          "online_loss": 0.02898234874010086,
          "train_loss": 0.028064358979463577,
          "validation_loss": 0.0240057110786438,
          "train_metric": 0.0028265395667403936,
          "validation_metric": 0.0024177671875804663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7662029650000477
        },
        {
          "epoch": 49,
          "updates": 196,
          "samples_seen": 50176,
          "online_loss": 0.027960652485489845,
          "train_loss": 0.027301957830786705,
          "validation_loss": 0.022566648200154305,
          "train_metric": 0.0027497531846165657,
          "validation_metric": 0.002272830344736576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7743541559993901
        },
        {
          "epoch": 50,
          "updates": 200,
          "samples_seen": 51200,
          "online_loss": 0.027206380385905504,
          "train_loss": 0.02606351301074028,
          "validation_loss": 0.021608058363199234,
          "train_metric": 0.002625021617859602,
          "validation_metric": 0.002176284324377775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7817412860003969
        },
        {
          "epoch": 51,
          "updates": 204,
          "samples_seen": 52224,
          "online_loss": 0.02741432934999466,
          "train_loss": 0.025300130248069763,
          "validation_loss": 0.021350732073187828,
          "train_metric": 0.002548136515542865,
          "validation_metric": 0.0021503674797713757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.788991931000055
        },
        {
          "epoch": 52,
          "updates": 208,
          "samples_seen": 53248,
          "online_loss": 0.02509255288168788,
          "train_loss": 0.024278726428747177,
          "validation_loss": 0.019641421735286713,
          "train_metric": 0.0024452644865959883,
          "validation_metric": 0.001978212036192417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7962625729996944
        },
        {
          "epoch": 53,
          "updates": 212,
          "samples_seen": 54272,
          "online_loss": 0.024482179898768663,
          "train_loss": 0.023111987859010696,
          "validation_loss": 0.019425369799137115,
          "train_metric": 0.0023277546279132366,
          "validation_metric": 0.0019564516842365265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8034208770004625
        },
        {
          "epoch": 54,
          "updates": 216,
          "samples_seen": 55296,
          "online_loss": 0.023405313026160002,
          "train_loss": 0.02251766249537468,
          "validation_loss": 0.01879245415329933,
          "train_metric": 0.002267896430566907,
          "validation_metric": 0.00189270731061697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8102392809996672
        },
        {
          "epoch": 55,
          "updates": 220,
          "samples_seen": 56320,
          "online_loss": 0.022968745324760675,
          "train_loss": 0.022083811461925507,
          "validation_loss": 0.018141770735383034,
          "train_metric": 0.002224200638011098,
          "validation_metric": 0.0018271728185936809,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8172767819996807
        },
        {
          "epoch": 56,
          "updates": 224,
          "samples_seen": 57344,
          "online_loss": 0.021559132728725672,
          "train_loss": 0.02161281555891037,
          "validation_loss": 0.0184086412191391,
          "train_metric": 0.0021767637226730585,
          "validation_metric": 0.0018540510209277272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8236346149988094
        },
        {
          "epoch": 57,
          "updates": 228,
          "samples_seen": 58368,
          "online_loss": 0.021912138909101486,
          "train_loss": 0.020717740058898926,
          "validation_loss": 0.017061706632375717,
          "train_metric": 0.0020866149570792913,
          "validation_metric": 0.0017183927120640874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8308412709993718
        },
        {
          "epoch": 58,
          "updates": 232,
          "samples_seen": 59392,
          "online_loss": 0.02081365929916501,
          "train_loss": 0.020041897892951965,
          "validation_loss": 0.016603678464889526,
          "train_metric": 0.0020185464527457952,
          "validation_metric": 0.0016722618602216244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9407744740001363
        },
        {
          "epoch": 59,
          "updates": 236,
          "samples_seen": 60416,
          "online_loss": 0.019945974927395582,
          "train_loss": 0.01977899856865406,
          "validation_loss": 0.016615934669971466,
          "train_metric": 0.0019920682534575462,
          "validation_metric": 0.0016734960954636335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9472362439992139
        },
        {
          "epoch": 60,
          "updates": 240,
          "samples_seen": 61440,
          "online_loss": 0.01988102961331606,
          "train_loss": 0.01908220537006855,
          "validation_loss": 0.0157000832259655,
          "train_metric": 0.0019218900706619024,
          "validation_metric": 0.0015812551137059927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9560831769995275
        },
        {
          "epoch": 61,
          "updates": 244,
          "samples_seen": 62464,
          "online_loss": 0.01917440816760063,
          "train_loss": 0.01876659132540226,
          "validation_loss": 0.015651728957891464,
          "train_metric": 0.0018901024013757706,
          "validation_metric": 0.001576384762302041,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9638684279998415
        },
        {
          "epoch": 62,
          "updates": 248,
          "samples_seen": 63488,
          "online_loss": 0.018794892355799675,
          "train_loss": 0.01835780404508114,
          "validation_loss": 0.014976565726101398,
          "train_metric": 0.0018489309586584568,
          "validation_metric": 0.001508384826593101,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9718829510002251
        },
        {
          "epoch": 63,
          "updates": 252,
          "samples_seen": 64512,
          "online_loss": 0.01824756618589163,
          "train_loss": 0.01789248175919056,
          "validation_loss": 0.014856128953397274,
          "train_metric": 0.001802065409719944,
          "validation_metric": 0.0014962549321353436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9790248360004625
        },
        {
          "epoch": 64,
          "updates": 256,
          "samples_seen": 65536,
          "online_loss": 0.018214376643300056,
          "train_loss": 0.018124133348464966,
          "validation_loss": 0.014819689095020294,
          "train_metric": 0.0018253965536132455,
          "validation_metric": 0.0014925848226994276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9863225429999147
        },
        {
          "epoch": 65,
          "updates": 260,
          "samples_seen": 66560,
          "online_loss": 0.01864674175158143,
          "train_loss": 0.01761454902589321,
          "validation_loss": 0.014659319072961807,
          "train_metric": 0.0017740728799253702,
          "validation_metric": 0.0014764328952878714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9929909969996515
        },
        {
          "epoch": 66,
          "updates": 264,
          "samples_seen": 67584,
          "online_loss": 0.018221605569124222,
          "train_loss": 0.01709895208477974,
          "validation_loss": 0.014254579320549965,
          "train_metric": 0.0017221438465639949,
          "validation_metric": 0.0014356691390275955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0007785849993525
        },
        {
          "epoch": 67,
          "updates": 268,
          "samples_seen": 68608,
          "online_loss": 0.018767610657960176,
          "train_loss": 0.019480634480714798,
          "validation_loss": 0.016175614669919014,
          "train_metric": 0.0019620179664343596,
          "validation_metric": 0.0016291487263515592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0070959249987936
        },
        {
          "epoch": 68,
          "updates": 272,
          "samples_seen": 69632,
          "online_loss": 0.019670295994728804,
          "train_loss": 0.019674241542816162,
          "validation_loss": 0.016917770728468895,
          "train_metric": 0.0019815173000097275,
          "validation_metric": 0.0017038958612829447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0147059430000809
        },
        {
          "epoch": 69,
          "updates": 276,
          "samples_seen": 70656,
          "online_loss": 0.018515140982344747,
          "train_loss": 0.01733163744211197,
          "validation_loss": 0.014565298333764076,
          "train_metric": 0.001745578832924366,
          "validation_metric": 0.0014669636730104685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0213242680001713
        },
        {
          "epoch": 70,
          "updates": 280,
          "samples_seen": 71680,
          "online_loss": 0.016841652570292354,
          "train_loss": 0.016561243683099747,
          "validation_loss": 0.013918273150920868,
          "train_metric": 0.0016679879045113921,
          "validation_metric": 0.0014017975190654397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0287675159997889
        },
        {
          "epoch": 71,
          "updates": 284,
          "samples_seen": 72704,
          "online_loss": 0.016381912399083376,
          "train_loss": 0.01593470759689808,
          "validation_loss": 0.01309140957891941,
          "train_metric": 0.0016048852121457458,
          "validation_metric": 0.0013185188872739673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0360718419997283
        },
        {
          "epoch": 72,
          "updates": 288,
          "samples_seen": 73728,
          "online_loss": 0.016020000213757157,
          "train_loss": 0.015757184475660324,
          "validation_loss": 0.013135469518601894,
          "train_metric": 0.0015870059141889215,
          "validation_metric": 0.0013229562900960445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0423912899987045
        },
        {
          "epoch": 73,
          "updates": 292,
          "samples_seen": 74752,
          "online_loss": 0.01585944276303053,
          "train_loss": 0.015520074404776096,
          "validation_loss": 0.012820014730095863,
          "train_metric": 0.0015631250571459532,
          "validation_metric": 0.0012911849189549685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0505114190000313
        },
        {
          "epoch": 74,
          "updates": 296,
          "samples_seen": 75776,
          "online_loss": 0.015788345132023096,
          "train_loss": 0.01522061787545681,
          "validation_loss": 0.012727867811918259,
          "train_metric": 0.00153296475764364,
          "validation_metric": 0.0012819041730836034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0590211959988665
        },
        {
          "epoch": 75,
          "updates": 300,
          "samples_seen": 76800,
          "online_loss": 0.015731432242318988,
          "train_loss": 0.014986981637775898,
          "validation_loss": 0.012499343603849411,
          "train_metric": 0.0015094337286427617,
          "validation_metric": 0.0012588881654664874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0683776129990292
        },
        {
          "epoch": 76,
          "updates": 304,
          "samples_seen": 77824,
          "online_loss": 0.01556265028193593,
          "train_loss": 0.014907960779964924,
          "validation_loss": 0.012407660484313965,
          "train_metric": 0.0015014751115813851,
          "validation_metric": 0.001249654102139175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0756358639991959
        },
        {
          "epoch": 77,
          "updates": 308,
          "samples_seen": 78848,
          "online_loss": 0.015611872542649508,
          "train_loss": 0.01492665708065033,
          "validation_loss": 0.01245479378849268,
          "train_metric": 0.0015033581294119358,
          "validation_metric": 0.0012544012861326337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.083120945999326
        },
        {
          "epoch": 78,
          "updates": 312,
          "samples_seen": 79872,
          "online_loss": 0.015160626033321023,
          "train_loss": 0.014865630306303501,
          "validation_loss": 0.012511109001934528,
          "train_metric": 0.0014972119824960828,
          "validation_metric": 0.0012600731570273638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0898291409994272
        },
        {
          "epoch": 79,
          "updates": 316,
          "samples_seen": 80896,
          "online_loss": 0.01511056674644351,
          "train_loss": 0.014543719589710236,
          "validation_loss": 0.012223483994603157,
          "train_metric": 0.0014647901989519596,
          "validation_metric": 0.001231104601174593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.09744937799951
        },
        {
          "epoch": 80,
          "updates": 320,
          "samples_seen": 81920,
          "online_loss": 0.014877725392580032,
          "train_loss": 0.014267274178564548,
          "validation_loss": 0.011918426491320133,
          "train_metric": 0.0014369479613378644,
          "validation_metric": 0.0012003802694380283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1047476690000622
        },
        {
          "epoch": 81,
          "updates": 324,
          "samples_seen": 82944,
          "online_loss": 0.014705633046105504,
          "train_loss": 0.014231361448764801,
          "validation_loss": 0.011843073181807995,
          "train_metric": 0.0014333307044580579,
          "validation_metric": 0.0011927909217774868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1119529079987842
        },
        {
          "epoch": 82,
          "updates": 328,
          "samples_seen": 83968,
          "online_loss": 0.014560605399310589,
          "train_loss": 0.01388575043529272,
          "validation_loss": 0.011672848835587502,
          "train_metric": 0.0013985221739858389,
          "validation_metric": 0.001175646553747356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1208368949992291
        },
        {
          "epoch": 83,
          "updates": 332,
          "samples_seen": 84992,
          "online_loss": 0.013987923506647348,
          "train_loss": 0.013848618604242802,
          "validation_loss": 0.01165960356593132,
          "train_metric": 0.001394782098941505,
          "validation_metric": 0.001174312666989863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.128023272000064
        },
        {
          "epoch": 84,
          "updates": 336,
          "samples_seen": 86016,
          "online_loss": 0.013945605605840683,
          "train_loss": 0.013536591082811356,
          "validation_loss": 0.01135789044201374,
          "train_metric": 0.001363355782814324,
          "validation_metric": 0.0011439248919487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1358525380001083
        },
        {
          "epoch": 85,
          "updates": 340,
          "samples_seen": 87040,
          "online_loss": 0.013730731094256043,
          "train_loss": 0.013765407726168633,
          "validation_loss": 0.011543959379196167,
          "train_metric": 0.001386401243507862,
          "validation_metric": 0.0011626651976257563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1424184669995157
        },
        {
          "epoch": 86,
          "updates": 344,
          "samples_seen": 88064,
          "online_loss": 0.014001176925376058,
          "train_loss": 0.01650753803551197,
          "validation_loss": 0.014369630254805088,
          "train_metric": 0.0016625786665827036,
          "validation_metric": 0.0014472564216703176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1507018140000582
        },
        {
          "epoch": 87,
          "updates": 348,
          "samples_seen": 89088,
          "online_loss": 0.014920398127287626,
          "train_loss": 0.015504194423556328,
          "validation_loss": 0.013304232619702816,
          "train_metric": 0.0015615257434546947,
          "validation_metric": 0.001339953625574708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.157339919000151
        },
        {
          "epoch": 88,
          "updates": 352,
          "samples_seen": 90112,
          "online_loss": 0.015166940167546272,
          "train_loss": 0.013136584311723709,
          "validation_loss": 0.01102067343890667,
          "train_metric": 0.0013230685144662857,
          "validation_metric": 0.0011099617695435882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.166931439000109
        },
        {
          "epoch": 89,
          "updates": 356,
          "samples_seen": 91136,
          "online_loss": 0.013407392194494605,
          "train_loss": 0.013018171302974224,
          "validation_loss": 0.01098991185426712,
          "train_metric": 0.0013111424632370472,
          "validation_metric": 0.001106863608583808,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1741739320004854
        },
        {
          "epoch": 90,
          "updates": 360,
          "samples_seen": 92160,
          "online_loss": 0.013228385709226131,
          "train_loss": 0.013831810094416142,
          "validation_loss": 0.011642353609204292,
          "train_metric": 0.0013930891873314977,
          "validation_metric": 0.0011725752847269177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1814544219996606
        },
        {
          "epoch": 91,
          "updates": 364,
          "samples_seen": 93184,
          "online_loss": 0.01387221785262227,
          "train_loss": 0.013109811581671238,
          "validation_loss": 0.011199072934687138,
          "train_metric": 0.0013203721027821302,
          "validation_metric": 0.001127929543145001,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1880195369994908
        },
        {
          "epoch": 92,
          "updates": 368,
          "samples_seen": 94208,
          "online_loss": 0.012916152831166983,
          "train_loss": 0.013218840584158897,
          "validation_loss": 0.011166906915605068,
          "train_metric": 0.0013313532108440995,
          "validation_metric": 0.0011246900539845228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1947633820000192
        },
        {
          "epoch": 93,
          "updates": 372,
          "samples_seen": 95232,
          "online_loss": 0.01299629919230938,
          "train_loss": 0.01281115971505642,
          "validation_loss": 0.010896896943449974,
          "train_metric": 0.0012902930611744523,
          "validation_metric": 0.0010974955512210727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.203759687999991
        },
        {
          "epoch": 94,
          "updates": 376,
          "samples_seen": 96256,
          "online_loss": 0.013007995905354619,
          "train_loss": 0.012368918396532536,
          "validation_loss": 0.010432178154587746,
          "train_metric": 0.0012457522097975016,
          "validation_metric": 0.0010506907710805535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2108736999998655
        },
        {
          "epoch": 95,
          "updates": 380,
          "samples_seen": 97280,
          "online_loss": 0.012988792499527335,
          "train_loss": 0.013112707994878292,
          "validation_loss": 0.01117703691124916,
          "train_metric": 0.0013206637231633067,
          "validation_metric": 0.001125710434280336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2183889119987725
        },
        {
          "epoch": 96,
          "updates": 384,
          "samples_seen": 98304,
          "online_loss": 0.012812299421057105,
          "train_loss": 0.012695219367742538,
          "validation_loss": 0.010938855819404125,
          "train_metric": 0.0012786160223186016,
          "validation_metric": 0.001101721660234034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2247813329995552
        },
        {
          "epoch": 97,
          "updates": 388,
          "samples_seen": 99328,
          "online_loss": 0.012952247867360711,
          "train_loss": 0.012822939082980156,
          "validation_loss": 0.010814803652465343,
          "train_metric": 0.0012914793333038688,
          "validation_metric": 0.0010892273858189583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2319891989991447
        },
        {
          "epoch": 98,
          "updates": 392,
          "samples_seen": 100352,
          "online_loss": 0.013082674937322736,
          "train_loss": 0.012401094660162926,
          "validation_loss": 0.010702277533710003,
          "train_metric": 0.0012489927466958761,
          "validation_metric": 0.0010778942378237844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2385900789995503
        },
        {
          "epoch": 99,
          "updates": 396,
          "samples_seen": 101376,
          "online_loss": 0.013572125928476453,
          "train_loss": 0.012177780270576477,
          "validation_loss": 0.010390298441052437,
          "train_metric": 0.0012265014229342341,
          "validation_metric": 0.0010464726947247982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2463568999992276
        },
        {
          "epoch": 100,
          "updates": 400,
          "samples_seen": 102400,
          "online_loss": 0.013589591486379504,
          "train_loss": 0.012679776176810265,
          "validation_loss": 0.010946561582386494,
          "train_metric": 0.001277060597203672,
          "validation_metric": 0.0011024978011846542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2535389860004216
        },
        {
          "epoch": 101,
          "updates": 404,
          "samples_seen": 103424,
          "online_loss": 0.013513091020286083,
          "train_loss": 0.013637401163578033,
          "validation_loss": 0.011773942038416862,
          "train_metric": 0.0013735090615227818,
          "validation_metric": 0.0011858283542096615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2598399250000512
        },
        {
          "epoch": 102,
          "updates": 408,
          "samples_seen": 104448,
          "online_loss": 0.01318713091313839,
          "train_loss": 0.013715009205043316,
          "validation_loss": 0.011901432648301125,
          "train_metric": 0.0013813255354762077,
          "validation_metric": 0.0011986687313765287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2671387859991228
        },
        {
          "epoch": 103,
          "updates": 412,
          "samples_seen": 105472,
          "online_loss": 0.01269127638079226,
          "train_loss": 0.013021263293921947,
          "validation_loss": 0.011426027864217758,
          "train_metric": 0.001311453990638256,
          "validation_metric": 0.0011507878080010414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2739730710000003
        },
        {
          "epoch": 104,
          "updates": 416,
          "samples_seen": 106496,
          "online_loss": 0.012749484041705728,
          "train_loss": 0.011981407180428505,
          "validation_loss": 0.010255225002765656,
          "train_metric": 0.0012067235074937344,
          "validation_metric": 0.0010328686330467463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.287220617999992
        },
        {
          "epoch": 105,
          "updates": 420,
          "samples_seen": 107520,
          "online_loss": 0.012644917936995625,
          "train_loss": 0.01175631396472454,
          "validation_loss": 0.009833182208240032,
          "train_metric": 0.001184052904136479,
          "validation_metric": 0.0009903621394187212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2944068710003194
        },
        {
          "epoch": 106,
          "updates": 424,
          "samples_seen": 108544,
          "online_loss": 0.013076340779662132,
          "train_loss": 0.012411773204803467,
          "validation_loss": 0.010819581337273121,
          "train_metric": 0.0012500683078542352,
          "validation_metric": 0.0010897086467593908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3038817759988888
        },
        {
          "epoch": 107,
          "updates": 428,
          "samples_seen": 109568,
          "online_loss": 0.012161561520770192,
          "train_loss": 0.011304476298391819,
          "validation_loss": 0.009558042511343956,
          "train_metric": 0.0011385454563423991,
          "validation_metric": 0.0009626509854570031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3106595419994846
        },
        {
          "epoch": 108,
          "updates": 432,
          "samples_seen": 110592,
          "online_loss": 0.011366640450432897,
          "train_loss": 0.011169984936714172,
          "validation_loss": 0.009580687619745731,
          "train_metric": 0.0011249998351559043,
          "validation_metric": 0.0009649317362345755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3190232239994657
        },
        {
          "epoch": 109,
          "updates": 436,
          "samples_seen": 111616,
          "online_loss": 0.011283340398222208,
          "train_loss": 0.011128545738756657,
          "validation_loss": 0.009544639848172665,
          "train_metric": 0.0011208263458684087,
          "validation_metric": 0.0009613011498004198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3261186929994437
        },
        {
          "epoch": 110,
          "updates": 440,
          "samples_seen": 112640,
          "online_loss": 0.012227755505591631,
          "train_loss": 0.011057651601731777,
          "validation_loss": 0.009550319984555244,
          "train_metric": 0.0011136861285194755,
          "validation_metric": 0.0009618733311071992,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3345653159994981
        },
        {
          "epoch": 111,
          "updates": 444,
          "samples_seen": 113664,
          "online_loss": 0.010979894897900522,
          "train_loss": 0.011171825230121613,
          "validation_loss": 0.00935143418610096,
          "train_metric": 0.0011251854011788964,
          "validation_metric": 0.0009418421541340649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3453862619990105
        },
        {
          "epoch": 112,
          "updates": 448,
          "samples_seen": 114688,
          "online_loss": 0.011469229240901768,
          "train_loss": 0.011732554994523525,
          "validation_loss": 0.010139798745512962,
          "train_metric": 0.0011816598707810044,
          "validation_metric": 0.0010212435154244304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.352178859999185
        },
        {
          "epoch": 113,
          "updates": 452,
          "samples_seen": 115712,
          "online_loss": 0.011609440203756094,
          "train_loss": 0.011318930424749851,
          "validation_loss": 0.009769273921847343,
          "train_metric": 0.0011400013463571668,
          "validation_metric": 0.00098392553627491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.358686513000066
        },
        {
          "epoch": 114,
          "updates": 456,
          "samples_seen": 116736,
          "online_loss": 0.011219935724511743,
          "train_loss": 0.011771002784371376,
          "validation_loss": 0.010124029591679573,
          "train_metric": 0.0011855323100462556,
          "validation_metric": 0.0010196551447734237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.365729551000186
        },
        {
          "epoch": 115,
          "updates": 460,
          "samples_seen": 117760,
          "online_loss": 0.011433018371462822,
          "train_loss": 0.010786794126033783,
          "validation_loss": 0.009327390231192112,
          "train_metric": 0.001086406409740448,
          "validation_metric": 0.0009394205990247428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3728734819997044
        },
        {
          "epoch": 116,
          "updates": 464,
          "samples_seen": 118784,
          "online_loss": 0.010983258485794067,
          "train_loss": 0.010562482289969921,
          "validation_loss": 0.008990852162241936,
          "train_metric": 0.0010638146195560694,
          "validation_metric": 0.0009055256377905607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3806472460000805
        },
        {
          "epoch": 117,
          "updates": 468,
          "samples_seen": 119808,
          "online_loss": 0.011208148906007409,
          "train_loss": 0.010545728728175163,
          "validation_loss": 0.009080763906240463,
          "train_metric": 0.0010621272958815098,
          "validation_metric": 0.0009145813528448343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.387840375999076
        },
        {
          "epoch": 118,
          "updates": 472,
          "samples_seen": 120832,
          "online_loss": 0.01135411229915917,
          "train_loss": 0.011078987270593643,
          "validation_loss": 0.00969491433352232,
          "train_metric": 0.001115834922529757,
          "validation_metric": 0.0009764361893758178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3942679840001801
        },
        {
          "epoch": 119,
          "updates": 476,
          "samples_seen": 121856,
          "online_loss": 0.011718705762177706,
          "train_loss": 0.011586069129407406,
          "validation_loss": 0.010045596398413181,
          "train_metric": 0.0011669064406305552,
          "validation_metric": 0.0010117556666955352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4042754900001455
        },
        {
          "epoch": 120,
          "updates": 480,
          "samples_seen": 122880,
          "online_loss": 0.011677124071866274,
          "train_loss": 0.011187659576535225,
          "validation_loss": 0.009641106240451336,
          "train_metric": 0.0011267800582572818,
          "validation_metric": 0.0009710169397294521,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.410954621998826
        },
        {
          "epoch": 121,
          "updates": 484,
          "samples_seen": 123904,
          "online_loss": 0.010964553453959525,
          "train_loss": 0.01072279829531908,
          "validation_loss": 0.009265118278563023,
          "train_metric": 0.0010799610754474998,
          "validation_metric": 0.0009331486653536558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4188971780004067
        },
        {
          "epoch": 122,
          "updates": 488,
          "samples_seen": 124928,
          "online_loss": 0.010688987327739596,
          "train_loss": 0.01040389109402895,
          "validation_loss": 0.008776698261499405,
          "train_metric": 0.001047841738909483,
          "validation_metric": 0.0008839569054543972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4261281509989203
        },
        {
          "epoch": 123,
          "updates": 492,
          "samples_seen": 125952,
          "online_loss": 0.010362839326262474,
          "train_loss": 0.010243715718388557,
          "validation_loss": 0.008923083543777466,
          "train_metric": 0.0010317096021026373,
          "validation_metric": 0.0008987003238871694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4332437779994507
        },
        {
          "epoch": 124,
          "updates": 496,
          "samples_seen": 126976,
          "online_loss": 0.01028748624958098,
          "train_loss": 0.01005138736218214,
          "validation_loss": 0.008531589061021805,
          "train_metric": 0.0010123387910425663,
          "validation_metric": 0.0008592703379690647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4406196019990603
        },
        {
          "epoch": 125,
          "updates": 500,
          "samples_seen": 128000,
          "online_loss": 0.010166870197281241,
          "train_loss": 0.010160899721086025,
          "validation_loss": 0.008633250370621681,
          "train_metric": 0.0010233685607090592,
          "validation_metric": 0.00086950947297737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4480183929990744
        },
        {
          "epoch": 126,
          "updates": 504,
          "samples_seen": 129024,
          "online_loss": 0.010716716293245554,
          "train_loss": 0.010650202631950378,
          "validation_loss": 0.00946366973221302,
          "train_metric": 0.0010726493783295155,
          "validation_metric": 0.000953146256506443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.548254998999255
        },
        {
          "epoch": 127,
          "updates": 508,
          "samples_seen": 130048,
          "online_loss": 0.010738982819020748,
          "train_loss": 0.010680163279175758,
          "validation_loss": 0.009050249122083187,
          "train_metric": 0.0010756668634712696,
          "validation_metric": 0.0009115081047639251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5590300709991425
        },
        {
          "epoch": 128,
          "updates": 512,
          "samples_seen": 131072,
          "online_loss": 0.01122938678599894,
          "train_loss": 0.010537359863519669,
          "validation_loss": 0.009300739504396915,
          "train_metric": 0.001061284332536161,
          "validation_metric": 0.0009367364691570401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5691336749987386
        },
        {
          "epoch": 129,
          "updates": 516,
          "samples_seen": 132096,
          "online_loss": 0.01078926061745733,
          "train_loss": 0.009761260822415352,
          "validation_loss": 0.008406694047152996,
          "train_metric": 0.0009831184288486838,
          "validation_metric": 0.0008466913714073598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5799604109997745
        },
        {
          "epoch": 130,
          "updates": 520,
          "samples_seen": 133120,
          "online_loss": 0.010300866677425802,
          "train_loss": 0.009722896851599216,
          "validation_loss": 0.008283067494630814,
          "train_metric": 0.0009792546043172479,
          "validation_metric": 0.0008342403452843428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5912689249998948
        },
        {
          "epoch": 131,
          "updates": 524,
          "samples_seen": 134144,
          "online_loss": 0.009812177391722798,
          "train_loss": 0.010107164271175861,
          "validation_loss": 0.00902146939188242,
          "train_metric": 0.0010179565288126469,
          "validation_metric": 0.0009086093632504344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.601292461000412
        },
        {
          "epoch": 132,
          "updates": 528,
          "samples_seen": 135168,
          "online_loss": 0.010883114067837596,
          "train_loss": 0.009710311889648438,
          "validation_loss": 0.00819335039705038,
          "train_metric": 0.000977987190708518,
          "validation_metric": 0.000825204246211797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6117699129990797
        },
        {
          "epoch": 133,
          "updates": 532,
          "samples_seen": 136192,
          "online_loss": 0.01057828776538372,
          "train_loss": 0.009545951150357723,
          "validation_loss": 0.008311483077704906,
          "train_metric": 0.0009614332229830325,
          "validation_metric": 0.0008371021249331534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6212373219987057
        },
        {
          "epoch": 134,
          "updates": 536,
          "samples_seen": 137216,
          "online_loss": 0.010409342590719461,
          "train_loss": 0.010280601680278778,
          "validation_loss": 0.008997753262519836,
          "train_metric": 0.0010354246478527784,
          "validation_metric": 0.0009062207536771894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6314318860004278
        },
        {
          "epoch": 135,
          "updates": 540,
          "samples_seen": 138240,
          "online_loss": 0.009995768778026104,
          "train_loss": 0.010470544919371605,
          "validation_loss": 0.009112165309488773,
          "train_metric": 0.0010545548284426332,
          "validation_metric": 0.0009177440078929067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.641351536000002
        },
        {
          "epoch": 136,
          "updates": 544,
          "samples_seen": 139264,
          "online_loss": 0.010396041325293481,
          "train_loss": 0.01031721755862236,
          "validation_loss": 0.009081299416720867,
          "train_metric": 0.0010391122195869684,
          "validation_metric": 0.0009146352531388402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6503087940000114
        },
        {
          "epoch": 137,
          "updates": 548,
          "samples_seen": 140288,
          "online_loss": 0.010272214189171791,
          "train_loss": 0.009422391653060913,
          "validation_loss": 0.008196561597287655,
          "train_metric": 0.0009489887743256986,
          "validation_metric": 0.0008255276479758322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6602257829999871
        },
        {
          "epoch": 138,
          "updates": 552,
          "samples_seen": 141312,
          "online_loss": 0.009399155504070222,
          "train_loss": 0.009135555475950241,
          "validation_loss": 0.0078118047676980495,
          "train_metric": 0.0009200996719300747,
          "validation_metric": 0.0007867764215916395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6711900239988609
        },
        {
          "epoch": 139,
          "updates": 556,
          "samples_seen": 142336,
          "online_loss": 0.01024860970210284,
          "train_loss": 0.009425101801753044,
          "validation_loss": 0.00827015284448862,
          "train_metric": 0.0009492617100477219,
          "validation_metric": 0.0008329394040629268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6817676299997402
        },
        {
          "epoch": 140,
          "updates": 560,
          "samples_seen": 143360,
          "online_loss": 0.011157535482198,
          "train_loss": 0.010195941664278507,
          "validation_loss": 0.008841649629175663,
          "train_metric": 0.0010268979240208864,
          "validation_metric": 0.0008904984570108354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6923347629999625
        },
        {
          "epoch": 141,
          "updates": 564,
          "samples_seen": 144384,
          "online_loss": 0.011377985356375575,
          "train_loss": 0.010039441287517548,
          "validation_loss": 0.008835389278829098,
          "train_metric": 0.0010111358715221286,
          "validation_metric": 0.0008898680098354816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7025244670003303
        },
        {
          "epoch": 142,
          "updates": 568,
          "samples_seen": 145408,
          "online_loss": 0.009806387592107058,
          "train_loss": 0.009467842057347298,
          "validation_loss": 0.008165869861841202,
          "train_metric": 0.0009535663994029164,
          "validation_metric": 0.0008224365301430225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.71236305399907
        },
        {
          "epoch": 143,
          "updates": 572,
          "samples_seen": 146432,
          "online_loss": 0.010087788105010986,
          "train_loss": 0.009172376245260239,
          "validation_loss": 0.007679418195039034,
          "train_metric": 0.0009238080237992108,
          "validation_metric": 0.0007734429673291743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7226192179987265
        },
        {
          "epoch": 144,
          "updates": 576,
          "samples_seen": 147456,
          "online_loss": 0.013163595693185925,
          "train_loss": 0.01090830285102129,
          "validation_loss": 0.009923497214913368,
          "train_metric": 0.0010986443376168609,
          "validation_metric": 0.0009994582505896688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7332179879995238
        },
        {
          "epoch": 145,
          "updates": 580,
          "samples_seen": 148480,
          "online_loss": 0.012256405781954527,
          "train_loss": 0.011521747335791588,
          "validation_loss": 0.009865480475127697,
          "train_metric": 0.0011604281608015299,
          "validation_metric": 0.0009936150163412094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7432574559989007
        },
        {
          "epoch": 146,
          "updates": 584,
          "samples_seen": 149504,
          "online_loss": 0.014443362946622074,
          "train_loss": 0.013063505291938782,
          "validation_loss": 0.012058527208864689,
          "train_metric": 0.0013157083885744214,
          "validation_metric": 0.0012144907377660275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.752631207999002
        },
        {
          "epoch": 147,
          "updates": 588,
          "samples_seen": 150528,
          "online_loss": 0.014559586066752672,
          "train_loss": 0.01591595634818077,
          "validation_loss": 0.014141879975795746,
          "train_metric": 0.0016029967227950692,
          "validation_metric": 0.0014243184123188257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7628866609993565
        },
        {
          "epoch": 148,
          "updates": 592,
          "samples_seen": 151552,
          "online_loss": 0.014214057009667158,
          "train_loss": 0.014312944374978542,
          "validation_loss": 0.013317218981683254,
          "train_metric": 0.001441547297872603,
          "validation_metric": 0.0013412615517154336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7735798800003977
        },
        {
          "epoch": 149,
          "updates": 596,
          "samples_seen": 152576,
          "online_loss": 0.01216944307088852,
          "train_loss": 0.012541702017188072,
          "validation_loss": 0.010947536677122116,
          "train_metric": 0.0012631540885195136,
          "validation_metric": 0.0011025958228856325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7829767759994866
        },
        {
          "epoch": 150,
          "updates": 600,
          "samples_seen": 153600,
          "online_loss": 0.010184445069171488,
          "train_loss": 0.010840014554560184,
          "validation_loss": 0.009492816403508186,
          "train_metric": 0.0010917666368186474,
          "validation_metric": 0.0009560817270539701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.793247142999462
        },
        {
          "epoch": 151,
          "updates": 604,
          "samples_seen": 154624,
          "online_loss": 0.010164747713133693,
          "train_loss": 0.009294478222727776,
          "validation_loss": 0.007902438752353191,
          "train_metric": 0.0009361058473587036,
          "validation_metric": 0.0007959046633914113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8031393729997944
        },
        {
          "epoch": 152,
          "updates": 608,
          "samples_seen": 155648,
          "online_loss": 0.009376043686643243,
          "train_loss": 0.008603010326623917,
          "validation_loss": 0.007454227656126022,
          "train_metric": 0.0008664637571200728,
          "validation_metric": 0.0007507625268772244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8128902429998561
        },
        {
          "epoch": 153,
          "updates": 612,
          "samples_seen": 156672,
          "online_loss": 0.008974196971394122,
          "train_loss": 0.008457507006824017,
          "validation_loss": 0.007235311437398195,
          "train_metric": 0.0008518091635778546,
          "validation_metric": 0.0007287140469998121,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.823131621000357
        },
        {
          "epoch": 154,
          "updates": 616,
          "samples_seen": 157696,
          "online_loss": 0.008588458062149584,
          "train_loss": 0.008580794557929039,
          "validation_loss": 0.007564314641058445,
          "train_metric": 0.0008642262546345592,
          "validation_metric": 0.0007618499803356826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8324324450004497
        },
        {
          "epoch": 155,
          "updates": 620,
          "samples_seen": 158720,
          "online_loss": 0.008304791175760329,
          "train_loss": 0.008956914767622948,
          "validation_loss": 0.007384310010820627,
          "train_metric": 0.000902107625734061,
          "validation_metric": 0.0007437206222675741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8417495220001001
        },
        {
          "epoch": 156,
          "updates": 624,
          "samples_seen": 159744,
          "online_loss": 0.00883965112734586,
          "train_loss": 0.0083331149071455,
          "validation_loss": 0.007353876251727343,
          "train_metric": 0.0008392808376811445,
          "validation_metric": 0.0007406554650515318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8506843189989013
        },
        {
          "epoch": 157,
          "updates": 628,
          "samples_seen": 160768,
          "online_loss": 0.008455623523332179,
          "train_loss": 0.008696702308952808,
          "validation_loss": 0.007263078819960356,
          "train_metric": 0.0008759000338613987,
          "validation_metric": 0.0007315105758607388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.859834609000245
        },
        {
          "epoch": 158,
          "updates": 632,
          "samples_seen": 161792,
          "online_loss": 0.00860157422721386,
          "train_loss": 0.008131071925163269,
          "validation_loss": 0.007074220571666956,
          "train_metric": 0.0008189317304641008,
          "validation_metric": 0.0007124895346350968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.870179374000145
        },
        {
          "epoch": 159,
          "updates": 636,
          "samples_seen": 162816,
          "online_loss": 0.008375897770747542,
          "train_loss": 0.008102167397737503,
          "validation_loss": 0.00705597223713994,
          "train_metric": 0.0008160206489264965,
          "validation_metric": 0.000710651685949415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8804217479992076
        },
        {
          "epoch": 160,
          "updates": 640,
          "samples_seen": 163840,
          "online_loss": 0.009229527087882161,
          "train_loss": 0.00832963827997446,
          "validation_loss": 0.006867293268442154,
          "train_metric": 0.0008389306603930891,
          "validation_metric": 0.0006916485726833344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8906886000004306
        },
        {
          "epoch": 161,
          "updates": 644,
          "samples_seen": 164864,
          "online_loss": 0.009537733276374638,
          "train_loss": 0.008118660189211369,
          "validation_loss": 0.007089728955179453,
          "train_metric": 0.000817681779153645,
          "validation_metric": 0.0007140514790080488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9001894220000395
        },
        {
          "epoch": 162,
          "updates": 648,
          "samples_seen": 165888,
          "online_loss": 0.008581783389672637,
          "train_loss": 0.008028890006244183,
          "validation_loss": 0.0069169034250080585,
          "train_metric": 0.0008086403831839561,
          "validation_metric": 0.0006966451765038073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9095021899993299
        },
        {
          "epoch": 163,
          "updates": 652,
          "samples_seen": 166912,
          "online_loss": 0.00817348004784435,
          "train_loss": 0.007959261536598206,
          "validation_loss": 0.006871366873383522,
          "train_metric": 0.0008016275824047625,
          "validation_metric": 0.0006920589949004352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9187830060000124
        },
        {
          "epoch": 164,
          "updates": 656,
          "samples_seen": 167936,
          "online_loss": 0.008292731596156955,
          "train_loss": 0.007946291007101536,
          "validation_loss": 0.006968763191252947,
          "train_metric": 0.0008003212860785425,
          "validation_metric": 0.0007018683245405555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.928322551999372
        },
        {
          "epoch": 165,
          "updates": 660,
          "samples_seen": 168960,
          "online_loss": 0.008426027954556048,
          "train_loss": 0.00801036786288023,
          "validation_loss": 0.006885179318487644,
          "train_metric": 0.0008067748276516795,
          "validation_metric": 0.0006934500415809453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.937974624999697
        },
        {
          "epoch": 166,
          "updates": 664,
          "samples_seen": 169984,
          "online_loss": 0.008133150171488523,
          "train_loss": 0.007797403261065483,
          "validation_loss": 0.006789105013012886,
          "train_metric": 0.0007853257120586932,
          "validation_metric": 0.0006837737746536732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9485347209993051
        },
        {
          "epoch": 167,
          "updates": 668,
          "samples_seen": 171008,
          "online_loss": 0.00788860023021698,
          "train_loss": 0.007705662399530411,
          "validation_loss": 0.006632314063608646,
          "train_metric": 0.000776086060795933,
          "validation_metric": 0.0006679823854938149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9587142479995236
        },
        {
          "epoch": 168,
          "updates": 672,
          "samples_seen": 172032,
          "online_loss": 0.008051420794799924,
          "train_loss": 0.007845576852560043,
          "validation_loss": 0.006952325347810984,
          "train_metric": 0.0007901777280494571,
          "validation_metric": 0.0007002127240411937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.968178443999932
        },
        {
          "epoch": 169,
          "updates": 676,
          "samples_seen": 173056,
          "online_loss": 0.007967867888510227,
          "train_loss": 0.007638285867869854,
          "validation_loss": 0.006644437089562416,
          "train_metric": 0.000769300211686641,
          "validation_metric": 0.0006692032911814749,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9776015139996161
        },
        {
          "epoch": 170,
          "updates": 680,
          "samples_seen": 174080,
          "online_loss": 0.007996629807166755,
          "train_loss": 0.007580652367323637,
          "validation_loss": 0.006659618578851223,
          "train_metric": 0.0007634956855326891,
          "validation_metric": 0.0006707324064336717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9871345829997153
        },
        {
          "epoch": 171,
          "updates": 684,
          "samples_seen": 175104,
          "online_loss": 0.008495549787767231,
          "train_loss": 0.007961499504745007,
          "validation_loss": 0.006616098806262016,
          "train_metric": 0.0008018531370908022,
          "validation_metric": 0.0006663493113592267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9973122039991722
        },
        {
          "epoch": 172,
          "updates": 688,
          "samples_seen": 176128,
          "online_loss": 0.007881498895585537,
          "train_loss": 0.007473092060536146,
          "validation_loss": 0.006389337591826916,
          "train_metric": 0.0007526624831371009,
          "validation_metric": 0.000643510720692575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.008056360999035
        },
        {
          "epoch": 173,
          "updates": 692,
          "samples_seen": 177152,
          "online_loss": 0.007704538060352206,
          "train_loss": 0.007442928850650787,
          "validation_loss": 0.006444028113037348,
          "train_metric": 0.0007496245088987052,
          "validation_metric": 0.0006490189698524773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.017639452998992
        },
        {
          "epoch": 174,
          "updates": 696,
          "samples_seen": 178176,
          "online_loss": 0.007984658353962004,
          "train_loss": 0.007663033902645111,
          "validation_loss": 0.006807868834584951,
          "train_metric": 0.0007717926055192947,
          "validation_metric": 0.0006856635445728898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0270924100004777
        },
        {
          "epoch": 175,
          "updates": 700,
          "samples_seen": 179200,
          "online_loss": 0.008410930284298956,
          "train_loss": 0.008395363576710224,
          "validation_loss": 0.0070778378285467625,
          "train_metric": 0.0008455502684228122,
          "validation_metric": 0.0007128539145924151,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0365563449995534
        },
        {
          "epoch": 176,
          "updates": 704,
          "samples_seen": 180224,
          "online_loss": 0.009604620980098844,
          "train_loss": 0.008203781209886074,
          "validation_loss": 0.00728667713701725,
          "train_metric": 0.0008262547780759633,
          "validation_metric": 0.0007338874274864793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.046019395998883
        },
        {
          "epoch": 177,
          "updates": 708,
          "samples_seen": 181248,
          "online_loss": 0.009730860125273466,
          "train_loss": 0.008297295309603214,
          "validation_loss": 0.007122292183339596,
          "train_metric": 0.0008356732432730496,
          "validation_metric": 0.0007173311314545572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.158940586999961
        },
        {
          "epoch": 178,
          "updates": 712,
          "samples_seen": 182272,
          "online_loss": 0.00875178852584213,
          "train_loss": 0.008335231803357601,
          "validation_loss": 0.00742054870352149,
          "train_metric": 0.0008394941105507314,
          "validation_metric": 0.0007473704754374921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1684366520003096
        },
        {
          "epoch": 179,
          "updates": 716,
          "samples_seen": 183296,
          "online_loss": 0.008688100147992373,
          "train_loss": 0.008084921166300774,
          "validation_loss": 0.006934524513781071,
          "train_metric": 0.0008142837323248386,
          "validation_metric": 0.0006984199280850589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1778258649992495
        },
        {
          "epoch": 180,
          "updates": 720,
          "samples_seen": 184320,
          "online_loss": 0.008144194260239601,
          "train_loss": 0.008003349415957928,
          "validation_loss": 0.0069488538429141045,
          "train_metric": 0.0008060680702328682,
          "validation_metric": 0.0006998630706220865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1874919359997875
        },
        {
          "epoch": 181,
          "updates": 724,
          "samples_seen": 185344,
          "online_loss": 0.008285569958388805,
          "train_loss": 0.007926593534648418,
          "validation_loss": 0.006684586405754089,
          "train_metric": 0.0007983375107869506,
          "validation_metric": 0.0006732470938004553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.19769308099967
        },
        {
          "epoch": 182,
          "updates": 728,
          "samples_seen": 186368,
          "online_loss": 0.008521553594619036,
          "train_loss": 0.007907642982900143,
          "validation_loss": 0.007244608830660582,
          "train_metric": 0.0007964288233779371,
          "validation_metric": 0.0007296505500562489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2073098740002024
        },
        {
          "epoch": 183,
          "updates": 732,
          "samples_seen": 187392,
          "online_loss": 0.008713381481356919,
          "train_loss": 0.007822273299098015,
          "validation_loss": 0.006351042538881302,
          "train_metric": 0.0007878307369537652,
          "validation_metric": 0.000639653648249805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2175558429989906
        },
        {
          "epoch": 184,
          "updates": 736,
          "samples_seen": 188416,
          "online_loss": 0.007631850545294583,
          "train_loss": 0.007765674963593483,
          "validation_loss": 0.006812789011746645,
          "train_metric": 0.0007821303443051875,
          "validation_metric": 0.0006861591245979071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2270827409993217
        },
        {
          "epoch": 185,
          "updates": 740,
          "samples_seen": 189440,
          "online_loss": 0.007401791401207447,
          "train_loss": 0.007605060935020447,
          "validation_loss": 0.006197862792760134,
          "train_metric": 0.0007659539114683867,
          "validation_metric": 0.0006242260569706559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2371243409997987
        },
        {
          "epoch": 186,
          "updates": 744,
          "samples_seen": 190464,
          "online_loss": 0.007099099340848625,
          "train_loss": 0.00745462067425251,
          "validation_loss": 0.0066587841138243675,
          "train_metric": 0.0007508021080866456,
          "validation_metric": 0.0006706483545713127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2467446939990623
        },
        {
          "epoch": 187,
          "updates": 748,
          "samples_seen": 191488,
          "online_loss": 0.007725291070528328,
          "train_loss": 0.0071050068363547325,
          "validation_loss": 0.005890827625989914,
          "train_metric": 0.0007155902567319572,
          "validation_metric": 0.0005933025968261063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2572775709995767
        },
        {
          "epoch": 188,
          "updates": 752,
          "samples_seen": 192512,
          "online_loss": 0.006990526104345918,
          "train_loss": 0.006923337001353502,
          "validation_loss": 0.006183223333209753,
          "train_metric": 0.0006972932023927569,
          "validation_metric": 0.0006227516569197178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.266982086999633
        },
        {
          "epoch": 189,
          "updates": 756,
          "samples_seen": 193536,
          "online_loss": 0.006957501405850053,
          "train_loss": 0.006932463496923447,
          "validation_loss": 0.005818644072860479,
          "train_metric": 0.0006982123595662415,
          "validation_metric": 0.0005860325181856751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2773874449994764
        },
        {
          "epoch": 190,
          "updates": 760,
          "samples_seen": 194560,
          "online_loss": 0.007026332546956837,
          "train_loss": 0.006973690818995237,
          "validation_loss": 0.006222044117748737,
          "train_metric": 0.0007023644866421819,
          "validation_metric": 0.0006266614655032754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.287497698998777
        },
        {
          "epoch": 191,
          "updates": 764,
          "samples_seen": 195584,
          "online_loss": 0.00692484644241631,
          "train_loss": 0.00693591358140111,
          "validation_loss": 0.005908163730055094,
          "train_metric": 0.0006985596846789122,
          "validation_metric": 0.0005950485938228667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2974383359996864
        },
        {
          "epoch": 192,
          "updates": 768,
          "samples_seen": 196608,
          "online_loss": 0.006903484463691711,
          "train_loss": 0.0068733664229512215,
          "validation_loss": 0.006070878356695175,
          "train_metric": 0.0006922603351995349,
          "validation_metric": 0.0006114366697147489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3070987800001603
        },
        {
          "epoch": 193,
          "updates": 772,
          "samples_seen": 197632,
          "online_loss": 0.007533212890848517,
          "train_loss": 0.007093310356140137,
          "validation_loss": 0.0060461764223873615,
          "train_metric": 0.0007144121336750686,
          "validation_metric": 0.0006089487578719854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3167288320000807
        },
        {
          "epoch": 194,
          "updates": 776,
          "samples_seen": 198656,
          "online_loss": 0.007464523077942431,
          "train_loss": 0.007146721705794334,
          "validation_loss": 0.006340811960399151,
          "train_metric": 0.0007197915692813694,
          "validation_metric": 0.0006386233726516366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3264681059990835
        },
        {
          "epoch": 195,
          "updates": 780,
          "samples_seen": 199680,
          "online_loss": 0.007982956944033504,
          "train_loss": 0.007437867112457752,
          "validation_loss": 0.0063516986556351185,
          "train_metric": 0.0007491147262044251,
          "validation_metric": 0.0006397197721526027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3359885919999215
        },
        {
          "epoch": 196,
          "updates": 784,
          "samples_seen": 200704,
          "online_loss": 0.008304919465444982,
          "train_loss": 0.007036933209747076,
          "validation_loss": 0.006206633523106575,
          "train_metric": 0.0007087340927682817,
          "validation_metric": 0.0006251094746403396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3452479240004322
        },
        {
          "epoch": 197,
          "updates": 788,
          "samples_seen": 201728,
          "online_loss": 0.008046927745454013,
          "train_loss": 0.007120000664144754,
          "validation_loss": 0.005923486314713955,
          "train_metric": 0.0007171003962866962,
          "validation_metric": 0.0005965917953290045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.354825267999331
        },
        {
          "epoch": 198,
          "updates": 792,
          "samples_seen": 202752,
          "online_loss": 0.008439777535386384,
          "train_loss": 0.006965816020965576,
          "validation_loss": 0.006323610432446003,
          "train_metric": 0.0007015714654698968,
          "validation_metric": 0.0006368908798322082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3647420529996452
        },
        {
          "epoch": 199,
          "updates": 796,
          "samples_seen": 203776,
          "online_loss": 0.007201079279184341,
          "train_loss": 0.006878797430545092,
          "validation_loss": 0.00577886588871479,
          "train_metric": 0.0006928072543814778,
          "validation_metric": 0.000582026201300323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3750431569987995
        },
        {
          "epoch": 200,
          "updates": 800,
          "samples_seen": 204800,
          "online_loss": 0.0068270828342065215,
          "train_loss": 0.006546244025230408,
          "validation_loss": 0.005829211324453354,
          "train_metric": 0.0006593137513846159,
          "validation_metric": 0.000587096787057817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.384581053000147
        },
        {
          "epoch": 201,
          "updates": 804,
          "samples_seen": 205824,
          "online_loss": 0.006435095798224211,
          "train_loss": 0.0063955700024962425,
          "validation_loss": 0.0053709507919847965,
          "train_metric": 0.000644138315692544,
          "validation_metric": 0.0005409424193203449,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.394773272999373
        },
        {
          "epoch": 202,
          "updates": 808,
          "samples_seen": 206848,
          "online_loss": 0.006791444844566286,
          "train_loss": 0.0065199038945138454,
          "validation_loss": 0.005826747976243496,
          "train_metric": 0.0006566608790308237,
          "validation_metric": 0.0005868487060070038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4039158769992355
        },
        {
          "epoch": 203,
          "updates": 812,
          "samples_seen": 207872,
          "online_loss": 0.0063974070362746716,
          "train_loss": 0.006337268278002739,
          "validation_loss": 0.00532032735645771,
          "train_metric": 0.0006382664432749152,
          "validation_metric": 0.0005358439520932734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4139369240001542
        },
        {
          "epoch": 204,
          "updates": 816,
          "samples_seen": 208896,
          "online_loss": 0.006228009006008506,
          "train_loss": 0.006172243505716324,
          "validation_loss": 0.005420459434390068,
          "train_metric": 0.0006216458277776837,
          "validation_metric": 0.0005459287203848362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.423531598000409
        },
        {
          "epoch": 205,
          "updates": 820,
          "samples_seen": 209920,
          "online_loss": 0.006259533343836665,
          "train_loss": 0.006134003400802612,
          "validation_loss": 0.005230304319411516,
          "train_metric": 0.0006177944596856833,
          "validation_metric": 0.0005267770029604435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4338417629987816
        },
        {
          "epoch": 206,
          "updates": 824,
          "samples_seen": 210944,
          "online_loss": 0.006355962308589369,
          "train_loss": 0.006037548184394836,
          "validation_loss": 0.0052249133586883545,
          "train_metric": 0.0006080797174945474,
          "validation_metric": 0.0005262341583147645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.444215389999954
        },
        {
          "epoch": 207,
          "updates": 828,
          "samples_seen": 211968,
          "online_loss": 0.006168331659864634,
          "train_loss": 0.006031825207173824,
          "validation_loss": 0.005261190701276064,
          "train_metric": 0.0006075034034438431,
          "validation_metric": 0.0005298879113979638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4542197390001093
        },
        {
          "epoch": 208,
          "updates": 832,
          "samples_seen": 212992,
          "online_loss": 0.006120747071690857,
          "train_loss": 0.0061416057869791985,
          "validation_loss": 0.005121405236423016,
          "train_metric": 0.0006185600650496781,
          "validation_metric": 0.0005158090498298407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.464798413999233
        },
        {
          "epoch": 209,
          "updates": 836,
          "samples_seen": 214016,
          "online_loss": 0.006011599558405578,
          "train_loss": 0.00604252377524972,
          "validation_loss": 0.00523537490516901,
          "train_metric": 0.0006085808854550123,
          "validation_metric": 0.0005272877751849592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4746333619987126
        },
        {
          "epoch": 210,
          "updates": 840,
          "samples_seen": 215040,
          "online_loss": 0.006307460949756205,
          "train_loss": 0.0060132830403745174,
          "validation_loss": 0.005076644476503134,
          "train_metric": 0.0006056358688510954,
          "validation_metric": 0.0005113009829074144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4852093489989784
        },
        {
          "epoch": 211,
          "updates": 844,
          "samples_seen": 216064,
          "online_loss": 0.006875503808259964,
          "train_loss": 0.005895363632589579,
          "validation_loss": 0.0050212424248456955,
          "train_metric": 0.0005937594105489552,
          "validation_metric": 0.0005057211965322495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4957399559989426
        },
        {
          "epoch": 212,
          "updates": 848,
          "samples_seen": 217088,
          "online_loss": 0.006918992614373565,
          "train_loss": 0.005952837411314249,
          "validation_loss": 0.005046125501394272,
          "train_metric": 0.0005995480460114777,
          "validation_metric": 0.0005082272691652179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5057826489992294
        },
        {
          "epoch": 213,
          "updates": 852,
          "samples_seen": 218112,
          "online_loss": 0.006172984489239752,
          "train_loss": 0.005987681448459625,
          "validation_loss": 0.005161609034985304,
          "train_metric": 0.0006030573858879507,
          "validation_metric": 0.0005198583821766078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.516731138999603
        },
        {
          "epoch": 214,
          "updates": 856,
          "samples_seen": 219136,
          "online_loss": 0.006298676715232432,
          "train_loss": 0.006008293014019728,
          "validation_loss": 0.005039581097662449,
          "train_metric": 0.0006051333039067686,
          "validation_metric": 0.0005075681256130338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.526501008998821
        },
        {
          "epoch": 215,
          "updates": 860,
          "samples_seen": 220160,
          "online_loss": 0.006152152083814144,
          "train_loss": 0.005866880062967539,
          "validation_loss": 0.005029991734772921,
          "train_metric": 0.000590890645980835,
          "validation_metric": 0.0005066022858954966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5360584179998114
        },
        {
          "epoch": 216,
          "updates": 864,
          "samples_seen": 221184,
          "online_loss": 0.00603113928809762,
          "train_loss": 0.005712863989174366,
          "validation_loss": 0.004957020748406649,
          "train_metric": 0.0005753787118010223,
          "validation_metric": 0.0004992529866285622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5464932220002083
        },
        {
          "epoch": 217,
          "updates": 868,
          "samples_seen": 222208,
          "online_loss": 0.006266834796406329,
          "train_loss": 0.005759899504482746,
          "validation_loss": 0.005020899698138237,
          "train_metric": 0.0005801160586997867,
          "validation_metric": 0.0005056866211816669,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5568121710002742
        },
        {
          "epoch": 218,
          "updates": 872,
          "samples_seen": 223232,
          "online_loss": 0.00666351686231792,
          "train_loss": 0.006177326198667288,
          "validation_loss": 0.0050720637664198875,
          "train_metric": 0.0006221576477400959,
          "validation_metric": 0.0005108396289870143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.566858697000498
        },
        {
          "epoch": 219,
          "updates": 876,
          "samples_seen": 224256,
          "online_loss": 0.0065526594407856464,
          "train_loss": 0.005881655029952526,
          "validation_loss": 0.00521077262237668,
          "train_metric": 0.0005923787830397487,
          "validation_metric": 0.0005248099332675338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.576449791999039
        },
        {
          "epoch": 220,
          "updates": 880,
          "samples_seen": 225280,
          "online_loss": 0.005790814058855176,
          "train_loss": 0.005664109718054533,
          "validation_loss": 0.0047146109864115715,
          "train_metric": 0.0005704683717340231,
          "validation_metric": 0.00047483836533501744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5876602930002264
        },
        {
          "epoch": 221,
          "updates": 884,
          "samples_seen": 226304,
          "online_loss": 0.006433013593778014,
          "train_loss": 0.005446576047688723,
          "validation_loss": 0.004685375839471817,
          "train_metric": 0.0005485591827891767,
          "validation_metric": 0.0004718939308077097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.598564317999262
        },
        {
          "epoch": 222,
          "updates": 888,
          "samples_seen": 227328,
          "online_loss": 0.005820697173476219,
          "train_loss": 0.0054405550472438335,
          "validation_loss": 0.004584336653351784,
          "train_metric": 0.0005479527171701193,
          "validation_metric": 0.00046171763096936047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6091296289996535
        },
        {
          "epoch": 223,
          "updates": 892,
          "samples_seen": 228352,
          "online_loss": 0.005466287140734494,
          "train_loss": 0.005454649683088064,
          "validation_loss": 0.004769566934555769,
          "train_metric": 0.0005493722856044769,
          "validation_metric": 0.00048037327360361814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.619084138999824
        },
        {
          "epoch": 224,
          "updates": 896,
          "samples_seen": 229376,
          "online_loss": 0.005503455409780145,
          "train_loss": 0.005589682143181562,
          "validation_loss": 0.004622182343155146,
          "train_metric": 0.0005629723891615868,
          "validation_metric": 0.00046552933054044843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.639037101998838
        },
        {
          "epoch": 225,
          "updates": 900,
          "samples_seen": 230400,
          "online_loss": 0.005698155611753464,
          "train_loss": 0.005318320356309414,
          "validation_loss": 0.00460421945899725,
          "train_metric": 0.00053564173867926,
          "validation_metric": 0.0004637200909201056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.648708077000265
        },
        {
          "epoch": 226,
          "updates": 904,
          "samples_seen": 231424,
          "online_loss": 0.005667669756803662,
          "train_loss": 0.005342741496860981,
          "validation_loss": 0.004467308521270752,
          "train_metric": 0.0005381013615988195,
          "validation_metric": 0.0004499309288803488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6589838179988874
        },
        {
          "epoch": 227,
          "updates": 908,
          "samples_seen": 232448,
          "online_loss": 0.006360000814311206,
          "train_loss": 0.005390094127506018,
          "validation_loss": 0.004609695635735989,
          "train_metric": 0.0005428706062957644,
          "validation_metric": 0.0004642716667149216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6693057230004342
        },
        {
          "epoch": 228,
          "updates": 912,
          "samples_seen": 233472,
          "online_loss": 0.0063132476061582565,
          "train_loss": 0.005326525308191776,
          "validation_loss": 0.004451633896678686,
          "train_metric": 0.0005364681710489094,
          "validation_metric": 0.00044835222070105374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.679984988999422
        },
        {
          "epoch": 229,
          "updates": 916,
          "samples_seen": 234496,
          "online_loss": 0.00724874809384346,
          "train_loss": 0.005599904805421829,
          "validation_loss": 0.004650639835745096,
          "train_metric": 0.0005640019080601633,
          "validation_metric": 0.0004683953884523362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6901523460001044
        },
        {
          "epoch": 230,
          "updates": 920,
          "samples_seen": 235520,
          "online_loss": 0.006939609418623149,
          "train_loss": 0.005714235361665487,
          "validation_loss": 0.004941822960972786,
          "train_metric": 0.00057551683858037,
          "validation_metric": 0.0004977222997695208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6997064419992967
        },
        {
          "epoch": 231,
          "updates": 924,
          "samples_seen": 236544,
          "online_loss": 0.006471008178777993,
          "train_loss": 0.005501961801201105,
          "validation_loss": 0.004614663776010275,
          "train_metric": 0.000554137397557497,
          "validation_metric": 0.0004647720779757947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7089078089993563
        },
        {
          "epoch": 232,
          "updates": 928,
          "samples_seen": 237568,
          "online_loss": 0.005659452988766134,
          "train_loss": 0.005348172038793564,
          "validation_loss": 0.004503632429987192,
          "train_metric": 0.0005386482807807624,
          "validation_metric": 0.00045358928036876023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.718402837999747
        },
        {
          "epoch": 233,
          "updates": 932,
          "samples_seen": 238592,
          "online_loss": 0.005536787677556276,
          "train_loss": 0.00527991633862257,
          "validation_loss": 0.004387908615171909,
          "train_metric": 0.0005317737814038992,
          "validation_metric": 0.0004419340693857521,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7282627029999276
        },
        {
          "epoch": 234,
          "updates": 936,
          "samples_seen": 239616,
          "online_loss": 0.005294401315040886,
          "train_loss": 0.005218051839619875,
          "validation_loss": 0.004341981373727322,
          "train_metric": 0.0005255431169643998,
          "validation_metric": 0.000437308510299772,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7389698419992783
        },
        {
          "epoch": 235,
          "updates": 940,
          "samples_seen": 240640,
          "online_loss": 0.005439436121378094,
          "train_loss": 0.005160981323570013,
          "validation_loss": 0.004494987893849611,
          "train_metric": 0.0005197951686568558,
          "validation_metric": 0.00045271875569596887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.748920089999956
        },
        {
          "epoch": 236,
          "updates": 944,
          "samples_seen": 241664,
          "online_loss": 0.005326475133188069,
          "train_loss": 0.005042757373303175,
          "validation_loss": 0.0042115869000554085,
          "train_metric": 0.0005078880349174142,
          "validation_metric": 0.00042417561053298414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.759338091000245
        },
        {
          "epoch": 237,
          "updates": 948,
          "samples_seen": 242688,
          "online_loss": 0.005116643384099007,
          "train_loss": 0.004923507105559111,
          "validation_loss": 0.004173318389803171,
          "train_metric": 0.000495877698995173,
          "validation_metric": 0.0004203213029541075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7704528510003
        },
        {
          "epoch": 238,
          "updates": 952,
          "samples_seen": 243712,
          "online_loss": 0.005173912330064923,
          "train_loss": 0.004947128240019083,
          "validation_loss": 0.0041684857569634914,
          "train_metric": 0.0004982565878890455,
          "validation_metric": 0.00041983459959737957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.780814540999927
        },
        {
          "epoch": 239,
          "updates": 956,
          "samples_seen": 244736,
          "online_loss": 0.005145486327819526,
          "train_loss": 0.004906264133751392,
          "validation_loss": 0.004225713200867176,
          "train_metric": 0.0004941410152241588,
          "validation_metric": 0.000425598380388692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7906014870004583
        },
        {
          "epoch": 240,
          "updates": 960,
          "samples_seen": 245760,
          "online_loss": 0.004987857886590064,
          "train_loss": 0.004902413114905357,
          "validation_loss": 0.004291872959583998,
          "train_metric": 0.0004937530611641705,
          "validation_metric": 0.0004322617023717612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.800171909999335
        }
      ],
      "best": {
        "epoch": 238,
        "updates": 952
      },
      "best_validation_metric": 0.00041983459959737957,
      "config": {
        "label": "Batch 256",
        "batch": 256,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.02048986577792494
    },
    "1fb401058891_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 1,
          "samples_seen": 1024,
          "online_loss": 0.9560257196426392,
          "train_loss": 0.9161807298660278,
          "validation_loss": 0.6619400978088379,
          "train_metric": 0.09227436780929565,
          "validation_metric": 0.06666817516088486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.013978955999846221
        },
        {
          "epoch": 2,
          "updates": 2,
          "samples_seen": 2048,
          "online_loss": 0.9161807298660278,
          "train_loss": 0.8742673993110657,
          "validation_loss": 0.630653440952301,
          "train_metric": 0.08805300295352936,
          "validation_metric": 0.06351710855960846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02231569200012018
        },
        {
          "epoch": 3,
          "updates": 3,
          "samples_seen": 3072,
          "online_loss": 0.8742673397064209,
          "train_loss": 0.8294147253036499,
          "validation_loss": 0.5974236726760864,
          "train_metric": 0.08353561162948608,
          "validation_metric": 0.06017033010721207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.030301516999315936
        },
        {
          "epoch": 4,
          "updates": 4,
          "samples_seen": 4096,
          "online_loss": 0.8294147849082947,
          "train_loss": 0.7809800505638123,
          "validation_loss": 0.5616785883903503,
          "train_metric": 0.07865744829177856,
          "validation_metric": 0.05657021328806877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1411802139991778
        },
        {
          "epoch": 5,
          "updates": 5,
          "samples_seen": 5120,
          "online_loss": 0.7809799909591675,
          "train_loss": 0.7286146879196167,
          "validation_loss": 0.5231126546859741,
          "train_metric": 0.0733834058046341,
          "validation_metric": 0.05268598347902298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1502413329999399
        },
        {
          "epoch": 6,
          "updates": 6,
          "samples_seen": 6144,
          "online_loss": 0.7286146879196167,
          "train_loss": 0.6722691059112549,
          "validation_loss": 0.48164600133895874,
          "train_metric": 0.06770848482847214,
          "validation_metric": 0.048509616404771805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15866837099929398
        },
        {
          "epoch": 7,
          "updates": 7,
          "samples_seen": 7168,
          "online_loss": 0.6722691059112549,
          "train_loss": 0.6122482419013977,
          "validation_loss": 0.4374586343765259,
          "train_metric": 0.06166340038180351,
          "validation_metric": 0.044059231877326965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16679716899852792
        },
        {
          "epoch": 8,
          "updates": 8,
          "samples_seen": 8192,
          "online_loss": 0.6122481822967529,
          "train_loss": 0.5493510365486145,
          "validation_loss": 0.39109256863594055,
          "train_metric": 0.05532862991094589,
          "validation_metric": 0.039389412850141525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17520977499953005
        },
        {
          "epoch": 9,
          "updates": 9,
          "samples_seen": 9216,
          "online_loss": 0.5493510365486145,
          "train_loss": 0.4850430190563202,
          "validation_loss": 0.343569815158844,
          "train_metric": 0.048851754516363144,
          "validation_metric": 0.034603092819452286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18332810299943958
        },
        {
          "epoch": 10,
          "updates": 10,
          "samples_seen": 10240,
          "online_loss": 0.4850430488586426,
          "train_loss": 0.4216090738773346,
          "validation_loss": 0.2965115010738373,
          "train_metric": 0.042462922632694244,
          "validation_metric": 0.029863551259040833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19140032099858217
        },
        {
          "epoch": 11,
          "updates": 11,
          "samples_seen": 11264,
          "online_loss": 0.4216090738773346,
          "train_loss": 0.36223751306533813,
          "validation_loss": 0.2522137463092804,
          "train_metric": 0.03648323193192482,
          "validation_metric": 0.025402043014764786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19966906699846731
        },
        {
          "epoch": 12,
          "updates": 12,
          "samples_seen": 12288,
          "online_loss": 0.36223748326301575,
          "train_loss": 0.3108953833580017,
          "validation_loss": 0.2135777324438095,
          "train_metric": 0.031312245875597,
          "validation_metric": 0.021510768681764603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2079414359996008
        },
        {
          "epoch": 13,
          "updates": 13,
          "samples_seen": 13312,
          "online_loss": 0.3108953833580017,
          "train_loss": 0.2716892957687378,
          "validation_loss": 0.18369662761688232,
          "train_metric": 0.027363549917936325,
          "validation_metric": 0.01850125379860401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21628156999940984
        },
        {
          "epoch": 14,
          "updates": 14,
          "samples_seen": 14336,
          "online_loss": 0.2716892659664154,
          "train_loss": 0.24716106057167053,
          "validation_loss": 0.1647392064332962,
          "train_metric": 0.024893153458833694,
          "validation_metric": 0.016591928899288177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2243535139987216
        },
        {
          "epoch": 15,
          "updates": 15,
          "samples_seen": 15360,
          "online_loss": 0.24716104567050934,
          "train_loss": 0.23577168583869934,
          "validation_loss": 0.15625663101673126,
          "train_metric": 0.023746062070131302,
          "validation_metric": 0.015737595036625862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23257652099891857
        },
        {
          "epoch": 16,
          "updates": 16,
          "samples_seen": 16384,
          "online_loss": 0.23577168583869934,
          "train_loss": 0.23156219720840454,
          "validation_loss": 0.1548166573047638,
          "train_metric": 0.023322096094489098,
          "validation_metric": 0.015592565760016441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24067651399855094
        },
        {
          "epoch": 17,
          "updates": 17,
          "samples_seen": 17408,
          "online_loss": 0.23156219720840454,
          "train_loss": 0.2286653220653534,
          "validation_loss": 0.1568123698234558,
          "train_metric": 0.02303033322095871,
          "validation_metric": 0.015793567523360252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2483017189988459
        },
        {
          "epoch": 18,
          "updates": 18,
          "samples_seen": 18432,
          "online_loss": 0.2286653220653534,
          "train_loss": 0.22529615461826324,
          "validation_loss": 0.1610189974308014,
          "train_metric": 0.02269100211560726,
          "validation_metric": 0.016217242926359177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2557813099992927
        },
        {
          "epoch": 19,
          "updates": 19,
          "samples_seen": 19456,
          "online_loss": 0.22529613971710205,
          "train_loss": 0.22261163592338562,
          "validation_loss": 0.16769887506961823,
          "train_metric": 0.022420627996325493,
          "validation_metric": 0.016890017315745354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2630619109986583
        },
        {
          "epoch": 20,
          "updates": 20,
          "samples_seen": 20480,
          "online_loss": 0.22261162102222443,
          "train_loss": 0.22140321135520935,
          "validation_loss": 0.1761833280324936,
          "train_metric": 0.022298922762274742,
          "validation_metric": 0.01774454116821289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27050996299840335
        },
        {
          "epoch": 21,
          "updates": 21,
          "samples_seen": 21504,
          "online_loss": 0.22140321135520935,
          "train_loss": 0.22034846246242523,
          "validation_loss": 0.18385130167007446,
          "train_metric": 0.022192690521478653,
          "validation_metric": 0.01851683109998703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2778972359992622
        },
        {
          "epoch": 22,
          "updates": 22,
          "samples_seen": 22528,
          "online_loss": 0.22034846246242523,
          "train_loss": 0.21711480617523193,
          "validation_loss": 0.1876053810119629,
          "train_metric": 0.02186700887978077,
          "validation_metric": 0.018894927576184273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2855455709996022
        },
        {
          "epoch": 23,
          "updates": 23,
          "samples_seen": 23552,
          "online_loss": 0.21711480617523193,
          "train_loss": 0.2103533148765564,
          "validation_loss": 0.18585535883903503,
          "train_metric": 0.021186014637351036,
          "validation_metric": 0.018718672916293144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2927334839987452
        },
        {
          "epoch": 24,
          "updates": 24,
          "samples_seen": 24576,
          "online_loss": 0.2103532999753952,
          "train_loss": 0.20036010444164276,
          "validation_loss": 0.17898492515087128,
          "train_metric": 0.020179536193609238,
          "validation_metric": 0.01802670769393444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30008142899896484
        },
        {
          "epoch": 25,
          "updates": 25,
          "samples_seen": 25600,
          "online_loss": 0.20036010444164276,
          "train_loss": 0.18852029740810394,
          "validation_loss": 0.16862811148166656,
          "train_metric": 0.018987072631716728,
          "validation_metric": 0.016983604058623314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3072302780001337
        },
        {
          "epoch": 26,
          "updates": 26,
          "samples_seen": 26624,
          "online_loss": 0.18852029740810394,
          "train_loss": 0.17649151384830475,
          "validation_loss": 0.15676411986351013,
          "train_metric": 0.017775580286979675,
          "validation_metric": 0.01578870788216591,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3147039429986762
        },
        {
          "epoch": 27,
          "updates": 27,
          "samples_seen": 27648,
          "online_loss": 0.17649152874946594,
          "train_loss": 0.16562947630882263,
          "validation_loss": 0.14510296285152435,
          "train_metric": 0.016681594774127007,
          "validation_metric": 0.014614238403737545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3225743610000791
        },
        {
          "epoch": 28,
          "updates": 28,
          "samples_seen": 28672,
          "online_loss": 0.16562947630882263,
          "train_loss": 0.15673771500587463,
          "validation_loss": 0.13480976223945618,
          "train_metric": 0.015786049887537956,
          "validation_metric": 0.013577544130384922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3312015519986744
        },
        {
          "epoch": 29,
          "updates": 29,
          "samples_seen": 29696,
          "online_loss": 0.15673771500587463,
          "train_loss": 0.15006347000598907,
          "validation_loss": 0.12647628784179688,
          "train_metric": 0.01511384453624487,
          "validation_metric": 0.012738226912915707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33958704699944064
        },
        {
          "epoch": 30,
          "updates": 30,
          "samples_seen": 30720,
          "online_loss": 0.15006348490715027,
          "train_loss": 0.14543038606643677,
          "validation_loss": 0.12022890895605087,
          "train_metric": 0.014647215604782104,
          "validation_metric": 0.012109015136957169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3486877870000171
        },
        {
          "epoch": 31,
          "updates": 31,
          "samples_seen": 31744,
          "online_loss": 0.14543038606643677,
          "train_loss": 0.14241059124469757,
          "validation_loss": 0.1158810555934906,
          "train_metric": 0.014343073591589928,
          "validation_metric": 0.011671116575598717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3568289669983642
        },
        {
          "epoch": 32,
          "updates": 32,
          "samples_seen": 32768,
          "online_loss": 0.14241057634353638,
          "train_loss": 0.14047309756278992,
          "validation_loss": 0.11307072639465332,
          "train_metric": 0.014147937297821045,
          "validation_metric": 0.011388069950044155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3652472849989863
        },
        {
          "epoch": 33,
          "updates": 33,
          "samples_seen": 33792,
          "online_loss": 0.1404731124639511,
          "train_loss": 0.13908520340919495,
          "validation_loss": 0.11136018484830856,
          "train_metric": 0.014008152298629284,
          "validation_metric": 0.01121578924357891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3733621569990646
        },
        {
          "epoch": 34,
          "updates": 34,
          "samples_seen": 34816,
          "online_loss": 0.13908520340919495,
          "train_loss": 0.13777372241020203,
          "validation_loss": 0.1103023961186409,
          "train_metric": 0.013876065611839294,
          "validation_metric": 0.011109251528978348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38168699400011974
        },
        {
          "epoch": 35,
          "updates": 35,
          "samples_seen": 35840,
          "online_loss": 0.13777372241020203,
          "train_loss": 0.13616566359996796,
          "validation_loss": 0.10949002951383591,
          "train_metric": 0.013714107684791088,
          "validation_metric": 0.011027433909475803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3900272020000557
        },
        {
          "epoch": 36,
          "updates": 36,
          "samples_seen": 36864,
          "online_loss": 0.13616566359996796,
          "train_loss": 0.13402269780635834,
          "validation_loss": 0.10860000550746918,
          "train_metric": 0.013498274609446526,
          "validation_metric": 0.010937795042991638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39863788899856445
        },
        {
          "epoch": 37,
          "updates": 37,
          "samples_seen": 37888,
          "online_loss": 0.13402268290519714,
          "train_loss": 0.1312716156244278,
          "validation_loss": 0.10743442177772522,
          "train_metric": 0.013221196830272675,
          "validation_metric": 0.010820401832461357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40682295499937027
        },
        {
          "epoch": 38,
          "updates": 38,
          "samples_seen": 38912,
          "online_loss": 0.1312716156244278,
          "train_loss": 0.12801945209503174,
          "validation_loss": 0.10594778507947922,
          "train_metric": 0.012893651612102985,
          "validation_metric": 0.010670673102140427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4151082209991728
        },
        {
          "epoch": 39,
          "updates": 39,
          "samples_seen": 39936,
          "online_loss": 0.12801945209503174,
          "train_loss": 0.12453638017177582,
          "validation_loss": 0.10424444079399109,
          "train_metric": 0.012542849406599998,
          "validation_metric": 0.010499119758605957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42354759499903594
        },
        {
          "epoch": 40,
          "updates": 40,
          "samples_seen": 40960,
          "online_loss": 0.12453639507293701,
          "train_loss": 0.12119479477405548,
          "validation_loss": 0.10253626108169556,
          "train_metric": 0.01220629457384348,
          "validation_metric": 0.010327075608074665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43193038099889236
        },
        {
          "epoch": 41,
          "updates": 41,
          "samples_seen": 41984,
          "online_loss": 0.12119479477405548,
          "train_loss": 0.11836740374565125,
          "validation_loss": 0.10106290131807327,
          "train_metric": 0.011921532452106476,
          "validation_metric": 0.010178686119616032,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4403875230000267
        },
        {
          "epoch": 42,
          "updates": 42,
          "samples_seen": 43008,
          "online_loss": 0.11836740374565125,
          "train_loss": 0.11630696058273315,
          "validation_loss": 0.09999421238899231,
          "train_metric": 0.011714011430740356,
          "validation_metric": 0.010071051307022572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4487736279988894
        },
        {
          "epoch": 43,
          "updates": 43,
          "samples_seen": 44032,
          "online_loss": 0.11630696803331375,
          "train_loss": 0.11504848301410675,
          "validation_loss": 0.09934888780117035,
          "train_metric": 0.011587261222302914,
          "validation_metric": 0.010006056167185307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5549442719984654
        },
        {
          "epoch": 44,
          "updates": 44,
          "samples_seen": 45056,
          "online_loss": 0.11504847556352615,
          "train_loss": 0.11438239365816116,
          "validation_loss": 0.09896872192621231,
          "train_metric": 0.011520175263285637,
          "validation_metric": 0.009967767633497715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5627701229986997
        },
        {
          "epoch": 45,
          "updates": 45,
          "samples_seen": 46080,
          "online_loss": 0.11438238620758057,
          "train_loss": 0.11392681300640106,
          "validation_loss": 0.09857125580310822,
          "train_metric": 0.011474290862679482,
          "validation_metric": 0.009927735663950443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5711532909990638
        },
        {
          "epoch": 46,
          "updates": 46,
          "samples_seen": 47104,
          "online_loss": 0.11392681300640106,
          "train_loss": 0.11327804625034332,
          "validation_loss": 0.09786372631788254,
          "train_metric": 0.011408950202167034,
          "validation_metric": 0.009856476448476315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5793302900001436
        },
        {
          "epoch": 47,
          "updates": 47,
          "samples_seen": 48128,
          "online_loss": 0.11327804625034332,
          "train_loss": 0.11216676980257034,
          "validation_loss": 0.09666233509778976,
          "train_metric": 0.011297025717794895,
          "validation_metric": 0.009735476225614548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5877089289988362
        },
        {
          "epoch": 48,
          "updates": 48,
          "samples_seen": 49152,
          "online_loss": 0.11216676980257034,
          "train_loss": 0.11054156720638275,
          "validation_loss": 0.09495529532432556,
          "train_metric": 0.011133342050015926,
          "validation_metric": 0.009563549421727657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5956827789996169
        },
        {
          "epoch": 49,
          "updates": 49,
          "samples_seen": 50176,
          "online_loss": 0.11054157465696335,
          "train_loss": 0.10854777693748474,
          "validation_loss": 0.09288579970598221,
          "train_metric": 0.01093253493309021,
          "validation_metric": 0.009355118498206139,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6037933560000965
        },
        {
          "epoch": 50,
          "updates": 50,
          "samples_seen": 51200,
          "online_loss": 0.10854779183864594,
          "train_loss": 0.10643056035041809,
          "validation_loss": 0.0906766727566719,
          "train_metric": 0.010719296522438526,
          "validation_metric": 0.009132622741162777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6116627840001456
        },
        {
          "epoch": 51,
          "updates": 51,
          "samples_seen": 52224,
          "online_loss": 0.1064305454492569,
          "train_loss": 0.10442199558019638,
          "validation_loss": 0.0885419249534607,
          "train_metric": 0.010516999289393425,
          "validation_metric": 0.008917619474232197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6202083239986678
        },
        {
          "epoch": 52,
          "updates": 52,
          "samples_seen": 53248,
          "online_loss": 0.10442198812961578,
          "train_loss": 0.10266201198101044,
          "validation_loss": 0.08662351965904236,
          "train_metric": 0.010339743457734585,
          "validation_metric": 0.008724404498934746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6282057579992397
        },
        {
          "epoch": 53,
          "updates": 53,
          "samples_seen": 54272,
          "online_loss": 0.10266203433275223,
          "train_loss": 0.1011735200881958,
          "validation_loss": 0.08496905863285065,
          "train_metric": 0.010189826600253582,
          "validation_metric": 0.008557774126529694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6367236499991122
        },
        {
          "epoch": 54,
          "updates": 54,
          "samples_seen": 55296,
          "online_loss": 0.1011735275387764,
          "train_loss": 0.09988442063331604,
          "validation_loss": 0.08354541659355164,
          "train_metric": 0.010059994645416737,
          "validation_metric": 0.008414389565587044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6447835589988244
        },
        {
          "epoch": 55,
          "updates": 55,
          "samples_seen": 56320,
          "online_loss": 0.09988442808389664,
          "train_loss": 0.09867511689662933,
          "validation_loss": 0.0822724774479866,
          "train_metric": 0.009938196279108524,
          "validation_metric": 0.008286181837320328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6530603239989432
        },
        {
          "epoch": 56,
          "updates": 56,
          "samples_seen": 57344,
          "online_loss": 0.09867511689662933,
          "train_loss": 0.09742829203605652,
          "validation_loss": 0.08106020092964172,
          "train_metric": 0.009812620468437672,
          "validation_metric": 0.008164087310433388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.661120459999438
        },
        {
          "epoch": 57,
          "updates": 57,
          "samples_seen": 58368,
          "online_loss": 0.09742830693721771,
          "train_loss": 0.09606557339429855,
          "validation_loss": 0.07983757555484772,
          "train_metric": 0.009675372391939163,
          "validation_metric": 0.008040948770940304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6694602549996489
        },
        {
          "epoch": 58,
          "updates": 58,
          "samples_seen": 59392,
          "online_loss": 0.09606557339429855,
          "train_loss": 0.09456409513950348,
          "validation_loss": 0.07856772094964981,
          "train_metric": 0.00952414982020855,
          "validation_metric": 0.007913053967058659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6775632399985625
        },
        {
          "epoch": 59,
          "updates": 59,
          "samples_seen": 60416,
          "online_loss": 0.09456409513950348,
          "train_loss": 0.09295393526554108,
          "validation_loss": 0.07724975794553757,
          "train_metric": 0.009361980482935905,
          "validation_metric": 0.007780313026160002,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6861019649986702
        },
        {
          "epoch": 60,
          "updates": 60,
          "samples_seen": 61440,
          "online_loss": 0.09295394271612167,
          "train_loss": 0.09130071103572845,
          "validation_loss": 0.075909823179245,
          "train_metric": 0.009195473045110703,
          "validation_metric": 0.007645359728485346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6942056739990221
        },
        {
          "epoch": 61,
          "updates": 61,
          "samples_seen": 62464,
          "online_loss": 0.09130071103572845,
          "train_loss": 0.08968067914247513,
          "validation_loss": 0.07458603382110596,
          "train_metric": 0.009032309986650944,
          "validation_metric": 0.007512031123042107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7027343399986421
        },
        {
          "epoch": 62,
          "updates": 62,
          "samples_seen": 63488,
          "online_loss": 0.08968067914247513,
          "train_loss": 0.08815548568964005,
          "validation_loss": 0.07331190258264542,
          "train_metric": 0.008878697641193867,
          "validation_metric": 0.007383706048130989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7118579609996232
        },
        {
          "epoch": 63,
          "updates": 63,
          "samples_seen": 64512,
          "online_loss": 0.08815548568964005,
          "train_loss": 0.08675387501716614,
          "validation_loss": 0.07210339605808258,
          "train_metric": 0.008737532421946526,
          "validation_metric": 0.007261989638209343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7239072799984569
        },
        {
          "epoch": 64,
          "updates": 64,
          "samples_seen": 65536,
          "online_loss": 0.08675387501716614,
          "train_loss": 0.08546563237905502,
          "validation_loss": 0.07095348834991455,
          "train_metric": 0.008607786148786545,
          "validation_metric": 0.007146175019443035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7344942449999508
        },
        {
          "epoch": 65,
          "updates": 65,
          "samples_seen": 66560,
          "online_loss": 0.08546563237905502,
          "train_loss": 0.08424965292215347,
          "validation_loss": 0.06983564049005508,
          "train_metric": 0.00848531723022461,
          "validation_metric": 0.007033590227365494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7429247109994321
        },
        {
          "epoch": 66,
          "updates": 66,
          "samples_seen": 67584,
          "online_loss": 0.08424966037273407,
          "train_loss": 0.08305266499519348,
          "validation_loss": 0.06871530413627625,
          "train_metric": 0.008364761248230934,
          "validation_metric": 0.006920753978192806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7541662539988465
        },
        {
          "epoch": 67,
          "updates": 67,
          "samples_seen": 68608,
          "online_loss": 0.08305266499519348,
          "train_loss": 0.08183129876852036,
          "validation_loss": 0.06756411492824554,
          "train_metric": 0.008241748437285423,
          "validation_metric": 0.006804809905588627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7625656289983453
        },
        {
          "epoch": 68,
          "updates": 68,
          "samples_seen": 69632,
          "online_loss": 0.08183129876852036,
          "train_loss": 0.08056823909282684,
          "validation_loss": 0.06637135148048401,
          "train_metric": 0.00811453815549612,
          "validation_metric": 0.006684679538011551,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.770874380999885
        },
        {
          "epoch": 69,
          "updates": 69,
          "samples_seen": 70656,
          "online_loss": 0.08056823909282684,
          "train_loss": 0.07927633821964264,
          "validation_loss": 0.06514780223369598,
          "train_metric": 0.007984422147274017,
          "validation_metric": 0.006561446934938431,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7793059589985205
        },
        {
          "epoch": 70,
          "updates": 70,
          "samples_seen": 71680,
          "online_loss": 0.07927633821964264,
          "train_loss": 0.0779900997877121,
          "validation_loss": 0.06392112374305725,
          "train_metric": 0.007854877039790154,
          "validation_metric": 0.0064379009418189526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7878513420000672
        },
        {
          "epoch": 71,
          "updates": 71,
          "samples_seen": 72704,
          "online_loss": 0.0779900923371315,
          "train_loss": 0.07674983143806458,
          "validation_loss": 0.06272576749324799,
          "train_metric": 0.0077299620024859905,
          "validation_metric": 0.00631750887259841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.796017867000046
        },
        {
          "epoch": 72,
          "updates": 72,
          "samples_seen": 73728,
          "online_loss": 0.07674982398748398,
          "train_loss": 0.07558578252792358,
          "validation_loss": 0.0615919828414917,
          "train_metric": 0.007612723391503096,
          "validation_metric": 0.0062033189460635185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8042768039995281
        },
        {
          "epoch": 73,
          "updates": 73,
          "samples_seen": 74752,
          "online_loss": 0.07558578997850418,
          "train_loss": 0.07450868934392929,
          "validation_loss": 0.06053847447037697,
          "train_metric": 0.007504242472350597,
          "validation_metric": 0.006097213365137577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8129809189995285
        },
        {
          "epoch": 74,
          "updates": 74,
          "samples_seen": 75776,
          "online_loss": 0.0745086818933487,
          "train_loss": 0.07350917160511017,
          "validation_loss": 0.059570372104644775,
          "train_metric": 0.007403573486953974,
          "validation_metric": 0.0059997099451720715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.821316941999612
        },
        {
          "epoch": 75,
          "updates": 75,
          "samples_seen": 76800,
          "online_loss": 0.07350917160511017,
          "train_loss": 0.07256462424993515,
          "validation_loss": 0.058682337403297424,
          "train_metric": 0.007308443076908588,
          "validation_metric": 0.005910269916057587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8303028880000056
        },
        {
          "epoch": 76,
          "updates": 76,
          "samples_seen": 77824,
          "online_loss": 0.07256462424993515,
          "train_loss": 0.07164978981018066,
          "validation_loss": 0.05786418169736862,
          "train_metric": 0.007216304540634155,
          "validation_metric": 0.005827868822962046,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8539338719983789
        },
        {
          "epoch": 77,
          "updates": 77,
          "samples_seen": 78848,
          "online_loss": 0.07164978235960007,
          "train_loss": 0.07074650377035141,
          "validation_loss": 0.0571066215634346,
          "train_metric": 0.007125329691916704,
          "validation_metric": 0.005751570221036673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.862555876999977
        },
        {
          "epoch": 78,
          "updates": 78,
          "samples_seen": 79872,
          "online_loss": 0.070746511220932,
          "train_loss": 0.06984960287809372,
          "validation_loss": 0.05640493705868721,
          "train_metric": 0.0070349955931305885,
          "validation_metric": 0.005680898204445839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8745906829990417
        },
        {
          "epoch": 79,
          "updates": 79,
          "samples_seen": 80896,
          "online_loss": 0.06984960287809372,
          "train_loss": 0.06896737217903137,
          "validation_loss": 0.055759597569704056,
          "train_metric": 0.006946141831576824,
          "validation_metric": 0.005615903064608574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8883410029993684
        },
        {
          "epoch": 80,
          "updates": 80,
          "samples_seen": 81920,
          "online_loss": 0.06896737217903137,
          "train_loss": 0.06811722368001938,
          "validation_loss": 0.05517415329813957,
          "train_metric": 0.006860516965389252,
          "validation_metric": 0.005556938704103231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8968704129983962
        },
        {
          "epoch": 81,
          "updates": 81,
          "samples_seen": 82944,
          "online_loss": 0.06811722368001938,
          "train_loss": 0.06731852889060974,
          "validation_loss": 0.05465142801403999,
          "train_metric": 0.006780075840651989,
          "validation_metric": 0.005504291970282793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0372639189990878
        },
        {
          "epoch": 82,
          "updates": 82,
          "samples_seen": 83968,
          "online_loss": 0.06731852889060974,
          "train_loss": 0.06658535450696945,
          "validation_loss": 0.0541897714138031,
          "train_metric": 0.006706233136355877,
          "validation_metric": 0.005457795225083828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0457206629998836
        },
        {
          "epoch": 83,
          "updates": 83,
          "samples_seen": 84992,
          "online_loss": 0.06658535450696945,
          "train_loss": 0.06592170894145966,
          "validation_loss": 0.05378088727593422,
          "train_metric": 0.006639393046498299,
          "validation_metric": 0.005416614003479481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0593574719987373
        },
        {
          "epoch": 84,
          "updates": 84,
          "samples_seen": 86016,
          "online_loss": 0.06592170894145966,
          "train_loss": 0.06532096862792969,
          "validation_loss": 0.05341033264994621,
          "train_metric": 0.006578887812793255,
          "validation_metric": 0.005379293113946915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0721838689987635
        },
        {
          "epoch": 85,
          "updates": 85,
          "samples_seen": 87040,
          "online_loss": 0.06532096117734909,
          "train_loss": 0.06476933509111404,
          "validation_loss": 0.053060635924339294,
          "train_metric": 0.0065233297646045685,
          "validation_metric": 0.005344072822481394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0802554549991328
        },
        {
          "epoch": 86,
          "updates": 86,
          "samples_seen": 88064,
          "online_loss": 0.06476932764053345,
          "train_loss": 0.0642518624663353,
          "validation_loss": 0.05271579325199127,
          "train_metric": 0.006471212022006512,
          "validation_metric": 0.005309341941028833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0960480799985817
        },
        {
          "epoch": 87,
          "updates": 87,
          "samples_seen": 89088,
          "online_loss": 0.0642518550157547,
          "train_loss": 0.06375806778669357,
          "validation_loss": 0.05236562341451645,
          "train_metric": 0.006421479396522045,
          "validation_metric": 0.005274074152112007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.108091534999403
        },
        {
          "epoch": 88,
          "updates": 88,
          "samples_seen": 90112,
          "online_loss": 0.06375806778669357,
          "train_loss": 0.06328494846820831,
          "validation_loss": 0.05200795829296112,
          "train_metric": 0.006373827811330557,
          "validation_metric": 0.005238051526248455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.121546093998404
        },
        {
          "epoch": 89,
          "updates": 89,
          "samples_seen": 91136,
          "online_loss": 0.06328494846820831,
          "train_loss": 0.0628361627459526,
          "validation_loss": 0.05164818465709686,
          "train_metric": 0.006328628398478031,
          "validation_metric": 0.005201816093176603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.130365475999497
        },
        {
          "epoch": 90,
          "updates": 90,
          "samples_seen": 92160,
          "online_loss": 0.0628361701965332,
          "train_loss": 0.062418244779109955,
          "validation_loss": 0.051296256482601166,
          "train_metric": 0.00628653634339571,
          "validation_metric": 0.005166370887309313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.139079442998991
        },
        {
          "epoch": 91,
          "updates": 91,
          "samples_seen": 93184,
          "online_loss": 0.062418244779109955,
          "train_loss": 0.06203577667474747,
          "validation_loss": 0.05096273869276047,
          "train_metric": 0.006248016841709614,
          "validation_metric": 0.005132779944688082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1474692319989117
        },
        {
          "epoch": 92,
          "updates": 92,
          "samples_seen": 94208,
          "online_loss": 0.06203577667474747,
          "train_loss": 0.061688050627708435,
          "validation_loss": 0.0506555512547493,
          "train_metric": 0.006212995387613773,
          "validation_metric": 0.0051018414087593555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.15573922799922
        },
        {
          "epoch": 93,
          "updates": 93,
          "samples_seen": 95232,
          "online_loss": 0.061688054352998734,
          "train_loss": 0.06136844679713249,
          "validation_loss": 0.05037828907370567,
          "train_metric": 0.006180805154144764,
          "validation_metric": 0.005073916167020798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1643894859989814
        },
        {
          "epoch": 94,
          "updates": 94,
          "samples_seen": 96256,
          "online_loss": 0.06136845052242279,
          "train_loss": 0.06106661632657051,
          "validation_loss": 0.05013055354356766,
          "train_metric": 0.0061504049226641655,
          "validation_metric": 0.0050489651039242744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.172579300999132
        },
        {
          "epoch": 95,
          "updates": 95,
          "samples_seen": 97280,
          "online_loss": 0.06106661260128021,
          "train_loss": 0.0607721321284771,
          "validation_loss": 0.04990949109196663,
          "train_metric": 0.00612074788659811,
          "validation_metric": 0.005026700906455517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1808022320001328
        },
        {
          "epoch": 96,
          "updates": 96,
          "samples_seen": 98304,
          "online_loss": 0.060772139579057693,
          "train_loss": 0.06047789379954338,
          "validation_loss": 0.04971146583557129,
          "train_metric": 0.006091112270951271,
          "validation_metric": 0.005006756633520126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1889879199989082
        },
        {
          "epoch": 97,
          "updates": 97,
          "samples_seen": 99328,
          "online_loss": 0.060477904975414276,
          "train_loss": 0.060181625187397,
          "validation_loss": 0.04953300207853317,
          "train_metric": 0.006061272695660591,
          "validation_metric": 0.004988782107830048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1974668379989453
        },
        {
          "epoch": 98,
          "updates": 98,
          "samples_seen": 100352,
          "online_loss": 0.060181621462106705,
          "train_loss": 0.05988515913486481,
          "validation_loss": 0.049370553344488144,
          "train_metric": 0.006031414493918419,
          "validation_metric": 0.00497242109850049,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2057677100001456
        },
        {
          "epoch": 99,
          "updates": 99,
          "samples_seen": 101376,
          "online_loss": 0.05988515913486481,
          "train_loss": 0.05959207937121391,
          "validation_loss": 0.04921960085630417,
          "train_metric": 0.006001895759254694,
          "validation_metric": 0.00495721772313118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2145480049985053
        },
        {
          "epoch": 100,
          "updates": 100,
          "samples_seen": 102400,
          "online_loss": 0.05959208309650421,
          "train_loss": 0.05930487811565399,
          "validation_loss": 0.049073681235313416,
          "train_metric": 0.005972969811409712,
          "validation_metric": 0.00494252098724246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2229395889989974
        },
        {
          "epoch": 101,
          "updates": 101,
          "samples_seen": 103424,
          "online_loss": 0.059304870665073395,
          "train_loss": 0.05902320146560669,
          "validation_loss": 0.04892423376441002,
          "train_metric": 0.00594460079446435,
          "validation_metric": 0.0049274698831140995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2313540609993652
        },
        {
          "epoch": 102,
          "updates": 102,
          "samples_seen": 104448,
          "online_loss": 0.05902320519089699,
          "train_loss": 0.058743856847286224,
          "validation_loss": 0.04876178130507469,
          "train_metric": 0.005916466470807791,
          "validation_metric": 0.0049111079424619675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2395908839989715
        },
        {
          "epoch": 103,
          "updates": 103,
          "samples_seen": 105472,
          "online_loss": 0.05874386057257652,
          "train_loss": 0.05846228823065758,
          "validation_loss": 0.048577819019556046,
          "train_metric": 0.005888108164072037,
          "validation_metric": 0.004892580211162567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.247594531998402
        },
        {
          "epoch": 104,
          "updates": 104,
          "samples_seen": 106496,
          "online_loss": 0.05846228823065758,
          "train_loss": 0.05817470699548721,
          "validation_loss": 0.04836694896221161,
          "train_metric": 0.005859144032001495,
          "validation_metric": 0.004871341399848461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.255902385999434
        },
        {
          "epoch": 105,
          "updates": 105,
          "samples_seen": 107520,
          "online_loss": 0.05817471072077751,
          "train_loss": 0.05787958949804306,
          "validation_loss": 0.04812813177704811,
          "train_metric": 0.005829420872032642,
          "validation_metric": 0.004847289063036442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.264271383999585
        },
        {
          "epoch": 106,
          "updates": 106,
          "samples_seen": 108544,
          "online_loss": 0.05787959322333336,
          "train_loss": 0.05757787078619003,
          "validation_loss": 0.047864656895399094,
          "train_metric": 0.005799032282084227,
          "validation_metric": 0.004820752888917923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2724354279998806
        },
        {
          "epoch": 107,
          "updates": 107,
          "samples_seen": 109568,
          "online_loss": 0.05757787451148033,
          "train_loss": 0.057271845638751984,
          "validation_loss": 0.047582902014255524,
          "train_metric": 0.005768210627138615,
          "validation_metric": 0.004792376421391964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2802945029998227
        },
        {
          "epoch": 108,
          "updates": 108,
          "samples_seen": 110592,
          "online_loss": 0.05727185308933258,
          "train_loss": 0.05696363374590874,
          "validation_loss": 0.04729047045111656,
          "train_metric": 0.0057371677830815315,
          "validation_metric": 0.0047629233449697495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2884603809998225
        },
        {
          "epoch": 109,
          "updates": 109,
          "samples_seen": 111616,
          "online_loss": 0.05696363374590874,
          "train_loss": 0.056654006242752075,
          "validation_loss": 0.046994488686323166,
          "train_metric": 0.0057059843093156815,
          "validation_metric": 0.00473311310634017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.296750281999266
        },
        {
          "epoch": 110,
          "updates": 110,
          "samples_seen": 112640,
          "online_loss": 0.056654006242752075,
          "train_loss": 0.05634216219186783,
          "validation_loss": 0.046700604259967804,
          "train_metric": 0.005674576386809349,
          "validation_metric": 0.004703513812273741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3053740010000183
        },
        {
          "epoch": 111,
          "updates": 111,
          "samples_seen": 113664,
          "online_loss": 0.05634216591715813,
          "train_loss": 0.056026436388492584,
          "validation_loss": 0.04641265422105789,
          "train_metric": 0.005642778240144253,
          "validation_metric": 0.004674512427300215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3141626999986329
        },
        {
          "epoch": 112,
          "updates": 112,
          "samples_seen": 114688,
          "online_loss": 0.056026436388492584,
          "train_loss": 0.05570529028773308,
          "validation_loss": 0.046132851392030716,
          "train_metric": 0.005610432475805283,
          "validation_metric": 0.004646332003176212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3228836039997987
        },
        {
          "epoch": 113,
          "updates": 113,
          "samples_seen": 115712,
          "online_loss": 0.05570528656244278,
          "train_loss": 0.05537811666727066,
          "validation_loss": 0.04586197808384895,
          "train_metric": 0.005577481351792812,
          "validation_metric": 0.0046190498396754265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3312302609992912
        },
        {
          "epoch": 114,
          "updates": 114,
          "samples_seen": 116736,
          "online_loss": 0.05537812039256096,
          "train_loss": 0.05504540354013443,
          "validation_loss": 0.045599520206451416,
          "train_metric": 0.005543971434235573,
          "validation_metric": 0.00459261704236269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3395161109983746
        },
        {
          "epoch": 115,
          "updates": 115,
          "samples_seen": 117760,
          "online_loss": 0.05504540354013443,
          "train_loss": 0.05470822751522064,
          "validation_loss": 0.045343637466430664,
          "train_metric": 0.005510011687874794,
          "validation_metric": 0.0045668454840779305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3475652809993335
        },
        {
          "epoch": 116,
          "updates": 116,
          "samples_seen": 118784,
          "online_loss": 0.054708220064640045,
          "train_loss": 0.05436762049794197,
          "validation_loss": 0.045091256499290466,
          "train_metric": 0.00547570688650012,
          "validation_metric": 0.00454142689704895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3561551259990665
        },
        {
          "epoch": 117,
          "updates": 117,
          "samples_seen": 119808,
          "online_loss": 0.05436762422323227,
          "train_loss": 0.05402407422661781,
          "validation_loss": 0.04483841732144356,
          "train_metric": 0.005441107787191868,
          "validation_metric": 0.004515961278229952,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3649050629992416
        },
        {
          "epoch": 118,
          "updates": 118,
          "samples_seen": 120832,
          "online_loss": 0.05402407795190811,
          "train_loss": 0.05367743968963623,
          "validation_loss": 0.04458093270659447,
          "train_metric": 0.005406195297837257,
          "validation_metric": 0.0044900281354784966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4809485379992111
        },
        {
          "epoch": 119,
          "updates": 119,
          "samples_seen": 121856,
          "online_loss": 0.05367743968963623,
          "train_loss": 0.05332721769809723,
          "validation_loss": 0.044315192848443985,
          "train_metric": 0.005370921455323696,
          "validation_metric": 0.004463263787329197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4988393610001367
        },
        {
          "epoch": 120,
          "updates": 120,
          "samples_seen": 122880,
          "online_loss": 0.05332722142338753,
          "train_loss": 0.0529729425907135,
          "validation_loss": 0.04403886944055557,
          "train_metric": 0.005335241090506315,
          "validation_metric": 0.00443543354049325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5115106379998906
        },
        {
          "epoch": 121,
          "updates": 121,
          "samples_seen": 123904,
          "online_loss": 0.0529729463160038,
          "train_loss": 0.052614495158195496,
          "validation_loss": 0.04375121369957924,
          "train_metric": 0.0052991388365626335,
          "validation_metric": 0.004406461957842112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5173028419994807
        },
        {
          "epoch": 122,
          "updates": 122,
          "samples_seen": 124928,
          "online_loss": 0.0526144914329052,
          "train_loss": 0.05225200578570366,
          "validation_loss": 0.04345312714576721,
          "train_metric": 0.005262630991637707,
          "validation_metric": 0.004376440308988094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5230235819999507
        },
        {
          "epoch": 123,
          "updates": 123,
          "samples_seen": 125952,
          "online_loss": 0.05225200578570366,
          "train_loss": 0.05188574641942978,
          "validation_loss": 0.04314665123820305,
          "train_metric": 0.00522574270144105,
          "validation_metric": 0.004345572553575039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.528743320999638
        },
        {
          "epoch": 124,
          "updates": 124,
          "samples_seen": 126976,
          "online_loss": 0.05188575014472008,
          "train_loss": 0.05151589214801788,
          "validation_loss": 0.04283452406525612,
          "train_metric": 0.005188492126762867,
          "validation_metric": 0.004314136691391468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5345991989997856
        },
        {
          "epoch": 125,
          "updates": 125,
          "samples_seen": 128000,
          "online_loss": 0.05151589587330818,
          "train_loss": 0.05114242061972618,
          "validation_loss": 0.04251964017748833,
          "train_metric": 0.005150876939296722,
          "validation_metric": 0.0042824228294193745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5402758789987274
        },
        {
          "epoch": 126,
          "updates": 126,
          "samples_seen": 129024,
          "online_loss": 0.05114242434501648,
          "train_loss": 0.05076514557003975,
          "validation_loss": 0.04220451042056084,
          "train_metric": 0.005112878978252411,
          "validation_metric": 0.004250683356076479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5458824839988665
        },
        {
          "epoch": 127,
          "updates": 127,
          "samples_seen": 130048,
          "online_loss": 0.05076514184474945,
          "train_loss": 0.050383780151605606,
          "validation_loss": 0.041890908032655716,
          "train_metric": 0.005074470303952694,
          "validation_metric": 0.004219098947942257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.551641189998918
        },
        {
          "epoch": 128,
          "updates": 128,
          "samples_seen": 131072,
          "online_loss": 0.050383783876895905,
          "train_loss": 0.04999801516532898,
          "validation_loss": 0.041579559445381165,
          "train_metric": 0.005035617388784885,
          "validation_metric": 0.004187741316854954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5573229229994467
        },
        {
          "epoch": 129,
          "updates": 129,
          "samples_seen": 132096,
          "online_loss": 0.04999801889061928,
          "train_loss": 0.0496075265109539,
          "validation_loss": 0.04127011448144913,
          "train_metric": 0.004996289033442736,
          "validation_metric": 0.004156574606895447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5633100389986794
        },
        {
          "epoch": 130,
          "updates": 130,
          "samples_seen": 133120,
          "online_loss": 0.0496075265109539,
          "train_loss": 0.0492120236158371,
          "validation_loss": 0.040961213409900665,
          "train_metric": 0.004956455901265144,
          "validation_metric": 0.004125463776290417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5690060709985119
        },
        {
          "epoch": 131,
          "updates": 131,
          "samples_seen": 134144,
          "online_loss": 0.049212031066417694,
          "train_loss": 0.04881126061081886,
          "validation_loss": 0.04065096378326416,
          "train_metric": 0.004916091449558735,
          "validation_metric": 0.0040942165069282055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5747709779989236
        },
        {
          "epoch": 132,
          "updates": 132,
          "samples_seen": 135168,
          "online_loss": 0.04881126061081886,
          "train_loss": 0.048405107110738754,
          "validation_loss": 0.04033726081252098,
          "train_metric": 0.004875184968113899,
          "validation_metric": 0.004062621854245663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5805999969998084
        },
        {
          "epoch": 133,
          "updates": 133,
          "samples_seen": 136192,
          "online_loss": 0.048405103385448456,
          "train_loss": 0.04799354821443558,
          "validation_loss": 0.040018290281295776,
          "train_metric": 0.0048337336629629135,
          "validation_metric": 0.0040304954163730145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5862697389984532
        },
        {
          "epoch": 134,
          "updates": 134,
          "samples_seen": 137216,
          "online_loss": 0.04799354076385498,
          "train_loss": 0.04757656157016754,
          "validation_loss": 0.03969267010688782,
          "train_metric": 0.004791737999767065,
          "validation_metric": 0.003997700288891792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5920492009990994
        },
        {
          "epoch": 135,
          "updates": 135,
          "samples_seen": 138240,
          "online_loss": 0.04757656902074814,
          "train_loss": 0.04715413227677345,
          "validation_loss": 0.039359599351882935,
          "train_metric": 0.004749190993607044,
          "validation_metric": 0.003964154981076717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5978181379996386
        },
        {
          "epoch": 136,
          "updates": 136,
          "samples_seen": 139264,
          "online_loss": 0.04715413600206375,
          "train_loss": 0.04672609642148018,
          "validation_loss": 0.03901877626776695,
          "train_metric": 0.004706081468611956,
          "validation_metric": 0.00392982829362154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.603574316999584
        },
        {
          "epoch": 137,
          "updates": 137,
          "samples_seen": 140288,
          "online_loss": 0.04672609642148018,
          "train_loss": 0.0462922677397728,
          "validation_loss": 0.038670431822538376,
          "train_metric": 0.004662388004362583,
          "validation_metric": 0.0038947444409132004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6093121339999925
        },
        {
          "epoch": 138,
          "updates": 138,
          "samples_seen": 141312,
          "online_loss": 0.0462922677397728,
          "train_loss": 0.04585254192352295,
          "validation_loss": 0.03831511363387108,
          "train_metric": 0.004618100356310606,
          "validation_metric": 0.003858958138152957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6151587269996526
        },
        {
          "epoch": 139,
          "updates": 139,
          "samples_seen": 142336,
          "online_loss": 0.04585254192352295,
          "train_loss": 0.045406971126794815,
          "validation_loss": 0.037953659892082214,
          "train_metric": 0.0045732236467301846,
          "validation_metric": 0.003822553437203169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.620916927999133
        },
        {
          "epoch": 140,
          "updates": 140,
          "samples_seen": 143360,
          "online_loss": 0.04540696740150452,
          "train_loss": 0.044955700635910034,
          "validation_loss": 0.03758692368865013,
          "train_metric": 0.004527773708105087,
          "validation_metric": 0.0037856174167245626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6266538639993087
        },
        {
          "epoch": 141,
          "updates": 141,
          "samples_seen": 144384,
          "online_loss": 0.044955700635910034,
          "train_loss": 0.04449890926480293,
          "validation_loss": 0.037215523421764374,
          "train_metric": 0.004481767304241657,
          "validation_metric": 0.0037482113111764193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.633100694998575
        },
        {
          "epoch": 142,
          "updates": 142,
          "samples_seen": 145408,
          "online_loss": 0.044498905539512634,
          "train_loss": 0.0440366230905056,
          "validation_loss": 0.03683977574110031,
          "train_metric": 0.004435207229107618,
          "validation_metric": 0.003710367251187563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.638879045000067
        },
        {
          "epoch": 143,
          "updates": 143,
          "samples_seen": 146432,
          "online_loss": 0.0440366193652153,
          "train_loss": 0.043568823486566544,
          "validation_loss": 0.036459796130657196,
          "train_metric": 0.004388092551380396,
          "validation_metric": 0.0036720966454595327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6446110689994384
        },
        {
          "epoch": 144,
          "updates": 144,
          "samples_seen": 147456,
          "online_loss": 0.04356881603598595,
          "train_loss": 0.043095484375953674,
          "validation_loss": 0.03607569634914398,
          "train_metric": 0.0043404195457696915,
          "validation_metric": 0.003633412066847086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6505510129991308
        },
        {
          "epoch": 145,
          "updates": 145,
          "samples_seen": 148480,
          "online_loss": 0.043095484375953674,
          "train_loss": 0.04261673614382744,
          "validation_loss": 0.03568778932094574,
          "train_metric": 0.004292201716452837,
          "validation_metric": 0.0035943433176726103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6562971769999422
        },
        {
          "epoch": 146,
          "updates": 146,
          "samples_seen": 149504,
          "online_loss": 0.04261673241853714,
          "train_loss": 0.04213280975818634,
          "validation_loss": 0.03529653698205948,
          "train_metric": 0.004243462346494198,
          "validation_metric": 0.0035549381282180548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6621068549993652
        },
        {
          "epoch": 147,
          "updates": 147,
          "samples_seen": 150528,
          "online_loss": 0.04213280975818634,
          "train_loss": 0.04164402186870575,
          "validation_loss": 0.034902423620224,
          "train_metric": 0.004194233100861311,
          "validation_metric": 0.0035152442287653685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6680805119995057
        },
        {
          "epoch": 148,
          "updates": 148,
          "samples_seen": 151552,
          "online_loss": 0.04164402186870575,
          "train_loss": 0.04115065187215805,
          "validation_loss": 0.03450560197234154,
          "train_metric": 0.004144543316215277,
          "validation_metric": 0.00347527745179832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6737901589985995
        },
        {
          "epoch": 149,
          "updates": 149,
          "samples_seen": 152576,
          "online_loss": 0.04115065187215805,
          "train_loss": 0.040652986615896225,
          "validation_loss": 0.034105800092220306,
          "train_metric": 0.004094419535249472,
          "validation_metric": 0.003435011487454176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.679423969999334
        },
        {
          "epoch": 150,
          "updates": 150,
          "samples_seen": 153600,
          "online_loss": 0.04065297916531563,
          "train_loss": 0.040151309221982956,
          "validation_loss": 0.03370228782296181,
          "train_metric": 0.004043893422931433,
          "validation_metric": 0.003394371597096324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.685396099999707
        },
        {
          "epoch": 151,
          "updates": 151,
          "samples_seen": 154624,
          "online_loss": 0.040151312947273254,
          "train_loss": 0.03964599221944809,
          "validation_loss": 0.03329404816031456,
          "train_metric": 0.003992999438196421,
          "validation_metric": 0.003353254869580269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6912684679991798
        },
        {
          "epoch": 152,
          "updates": 152,
          "samples_seen": 155648,
          "online_loss": 0.03964599221944809,
          "train_loss": 0.03913738951086998,
          "validation_loss": 0.03288005292415619,
          "train_metric": 0.003941774368286133,
          "validation_metric": 0.0033115586265921593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6971202509994328
        },
        {
          "epoch": 153,
          "updates": 153,
          "samples_seen": 156672,
          "online_loss": 0.03913738951086998,
          "train_loss": 0.03862586244940758,
          "validation_loss": 0.03245968371629715,
          "train_metric": 0.003890255931764841,
          "validation_metric": 0.0032692207023501396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7030877899996995
        },
        {
          "epoch": 154,
          "updates": 154,
          "samples_seen": 157696,
          "online_loss": 0.038625869899988174,
          "train_loss": 0.03811182081699371,
          "validation_loss": 0.03203314542770386,
          "train_metric": 0.003838482778519392,
          "validation_metric": 0.003226260654628277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7090801679987635
        },
        {
          "epoch": 155,
          "updates": 155,
          "samples_seen": 158720,
          "online_loss": 0.038111817091703415,
          "train_loss": 0.03759568929672241,
          "validation_loss": 0.03160158172249794,
          "train_metric": 0.003786499612033367,
          "validation_metric": 0.003182795364409685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7148324609988777
        },
        {
          "epoch": 156,
          "updates": 156,
          "samples_seen": 159744,
          "online_loss": 0.037595681846141815,
          "train_loss": 0.03707795590162277,
          "validation_loss": 0.031166914850473404,
          "train_metric": 0.003734356025233865,
          "validation_metric": 0.0031390171498060226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.801954835998913
        },
        {
          "epoch": 157,
          "updates": 157,
          "samples_seen": 160768,
          "online_loss": 0.03707795590162277,
          "train_loss": 0.0365590900182724,
          "validation_loss": 0.0307312048971653,
          "train_metric": 0.003682097652927041,
          "validation_metric": 0.0030951346270740032,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8080727759988804
        },
        {
          "epoch": 158,
          "updates": 158,
          "samples_seen": 161792,
          "online_loss": 0.0365590900182724,
          "train_loss": 0.036039549857378006,
          "validation_loss": 0.030295955017209053,
          "train_metric": 0.0036297712940722704,
          "validation_metric": 0.00305129773914814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8139060649991734
        },
        {
          "epoch": 159,
          "updates": 159,
          "samples_seen": 162816,
          "online_loss": 0.036039549857378006,
          "train_loss": 0.03551984950900078,
          "validation_loss": 0.029861686751246452,
          "train_metric": 0.0035774291027337313,
          "validation_metric": 0.0030075598042458296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8196099249998952
        },
        {
          "epoch": 160,
          "updates": 160,
          "samples_seen": 163840,
          "online_loss": 0.03551984950900078,
          "train_loss": 0.035000547766685486,
          "validation_loss": 0.029427718371152878,
          "train_metric": 0.0035251271910965443,
          "validation_metric": 0.0029638519044965506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8255711549991247
        },
        {
          "epoch": 161,
          "updates": 161,
          "samples_seen": 164864,
          "online_loss": 0.035000551491975784,
          "train_loss": 0.03448221832513809,
          "validation_loss": 0.028992759063839912,
          "train_metric": 0.0034729233011603355,
          "validation_metric": 0.002920044120401144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8315660519983794
        },
        {
          "epoch": 162,
          "updates": 162,
          "samples_seen": 165888,
          "online_loss": 0.03448222205042839,
          "train_loss": 0.033965401351451874,
          "validation_loss": 0.02855565771460533,
          "train_metric": 0.003420870751142502,
          "validation_metric": 0.002876021433621645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8373790629993891
        },
        {
          "epoch": 163,
          "updates": 163,
          "samples_seen": 166912,
          "online_loss": 0.03396540507674217,
          "train_loss": 0.033450670540332794,
          "validation_loss": 0.028116250410676003,
          "train_metric": 0.0033690291456878185,
          "validation_metric": 0.002831765916198492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8431779549991916
        },
        {
          "epoch": 164,
          "updates": 164,
          "samples_seen": 167936,
          "online_loss": 0.033450666815042496,
          "train_loss": 0.0329386405646801,
          "validation_loss": 0.027675678953528404,
          "train_metric": 0.003317459486424923,
          "validation_metric": 0.0027873930521309376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8488922050000838
        },
        {
          "epoch": 165,
          "updates": 165,
          "samples_seen": 168960,
          "online_loss": 0.0329386442899704,
          "train_loss": 0.032429955899715424,
          "validation_loss": 0.027235977351665497,
          "train_metric": 0.003266226267442107,
          "validation_metric": 0.0027431081980466843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8546286749988212
        },
        {
          "epoch": 166,
          "updates": 166,
          "samples_seen": 169984,
          "online_loss": 0.032429955899715424,
          "train_loss": 0.031925223767757416,
          "validation_loss": 0.026799123734235764,
          "train_metric": 0.0032153914216905832,
          "validation_metric": 0.002699109259992838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.860248558999956
        },
        {
          "epoch": 167,
          "updates": 167,
          "samples_seen": 171008,
          "online_loss": 0.03192522004246712,
          "train_loss": 0.03142505884170532,
          "validation_loss": 0.026366285979747772,
          "train_metric": 0.0031650171149522066,
          "validation_metric": 0.002655515680089593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8659330219998083
        },
        {
          "epoch": 168,
          "updates": 168,
          "samples_seen": 172032,
          "online_loss": 0.03142506256699562,
          "train_loss": 0.030930064618587494,
          "validation_loss": 0.025937583297491074,
          "train_metric": 0.003115162719041109,
          "validation_metric": 0.0026123386342078447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8716691319987149
        },
        {
          "epoch": 169,
          "updates": 169,
          "samples_seen": 173056,
          "online_loss": 0.030930066481232643,
          "train_loss": 0.030440794304013252,
          "validation_loss": 0.0255126953125,
          "train_metric": 0.003065885277464986,
          "validation_metric": 0.002569545293226838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8774380429986195
        },
        {
          "epoch": 170,
          "updates": 170,
          "samples_seen": 174080,
          "online_loss": 0.030440792441368103,
          "train_loss": 0.02995777502655983,
          "validation_loss": 0.025091754272580147,
          "train_metric": 0.003017237177118659,
          "validation_metric": 0.0025271496269851923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8833801869986928
        },
        {
          "epoch": 171,
          "updates": 171,
          "samples_seen": 175104,
          "online_loss": 0.02995777502655983,
          "train_loss": 0.02948158048093319,
          "validation_loss": 0.02467595785856247,
          "train_metric": 0.0029692770913243294,
          "validation_metric": 0.0024852720089256763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.889139148999675
        },
        {
          "epoch": 172,
          "updates": 172,
          "samples_seen": 176128,
          "online_loss": 0.02948158234357834,
          "train_loss": 0.029012786224484444,
          "validation_loss": 0.024266991764307022,
          "train_metric": 0.0029220618307590485,
          "validation_metric": 0.002444082871079445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8951731990000553
        },
        {
          "epoch": 173,
          "updates": 173,
          "samples_seen": 177152,
          "online_loss": 0.029012786224484444,
          "train_loss": 0.028551889583468437,
          "validation_loss": 0.023865897208452225,
          "train_metric": 0.002875641919672489,
          "validation_metric": 0.002403685823082924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9011873089984874
        },
        {
          "epoch": 174,
          "updates": 174,
          "samples_seen": 178176,
          "online_loss": 0.028551891446113586,
          "train_loss": 0.028099365532398224,
          "validation_loss": 0.023472392931580544,
          "train_metric": 0.002830065321177244,
          "validation_metric": 0.002364053390920162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.907009429000027
        },
        {
          "epoch": 175,
          "updates": 175,
          "samples_seen": 179200,
          "online_loss": 0.028099369257688522,
          "train_loss": 0.027655644342303276,
          "validation_loss": 0.02308523654937744,
          "train_metric": 0.002785374876111746,
          "validation_metric": 0.0023250600788742304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9128400519985007
        },
        {
          "epoch": 176,
          "updates": 176,
          "samples_seen": 180224,
          "online_loss": 0.027655644342303276,
          "train_loss": 0.027221055701375008,
          "validation_loss": 0.022703608497977257,
          "train_metric": 0.0027416052762418985,
          "validation_metric": 0.0022866243962198496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9187891769997805
        },
        {
          "epoch": 177,
          "updates": 177,
          "samples_seen": 181248,
          "online_loss": 0.02722105383872986,
          "train_loss": 0.026795951649546623,
          "validation_loss": 0.02232828177511692,
          "train_metric": 0.0026987900491803885,
          "validation_metric": 0.0022488227114081383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9245846119993075
        },
        {
          "epoch": 178,
          "updates": 178,
          "samples_seen": 182272,
          "online_loss": 0.026795949786901474,
          "train_loss": 0.026380637660622597,
          "validation_loss": 0.021961314603686333,
          "train_metric": 0.0026569608598947525,
          "validation_metric": 0.0022118634078651667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9303767559995322
        },
        {
          "epoch": 179,
          "updates": 179,
          "samples_seen": 183296,
          "online_loss": 0.026380637660622597,
          "train_loss": 0.025975331664085388,
          "validation_loss": 0.021604424342513084,
          "train_metric": 0.002616140292957425,
          "validation_metric": 0.002175918547436595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9361327149999852
        },
        {
          "epoch": 180,
          "updates": 180,
          "samples_seen": 184320,
          "online_loss": 0.025975331664085388,
          "train_loss": 0.02558022178709507,
          "validation_loss": 0.021257640793919563,
          "train_metric": 0.0025763465091586113,
          "validation_metric": 0.0021409918554127216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9418521409988898
        },
        {
          "epoch": 181,
          "updates": 181,
          "samples_seen": 185344,
          "online_loss": 0.02558022178709507,
          "train_loss": 0.025195438414812088,
          "validation_loss": 0.020919619128108025,
          "train_metric": 0.002537592314183712,
          "validation_metric": 0.002106947358697653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9478089819986053
        },
        {
          "epoch": 182,
          "updates": 182,
          "samples_seen": 186368,
          "online_loss": 0.02519543655216694,
          "train_loss": 0.024821080267429352,
          "validation_loss": 0.020589401945471764,
          "train_metric": 0.002499887952581048,
          "validation_metric": 0.002073689131066203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9538803199993708
        },
        {
          "epoch": 183,
          "updates": 183,
          "samples_seen": 187392,
          "online_loss": 0.024821078404784203,
          "train_loss": 0.024457227438688278,
          "validation_loss": 0.020267533138394356,
          "train_metric": 0.002463242504745722,
          "validation_metric": 0.002041271887719631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9596733869984746
        },
        {
          "epoch": 184,
          "updates": 184,
          "samples_seen": 188416,
          "online_loss": 0.024457227438688278,
          "train_loss": 0.024103917181491852,
          "validation_loss": 0.01995517686009407,
          "train_metric": 0.00242765829898417,
          "validation_metric": 0.0020098122768104076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9655579079990275
        },
        {
          "epoch": 185,
          "updates": 185,
          "samples_seen": 189440,
          "online_loss": 0.024103917181491852,
          "train_loss": 0.023761125281453133,
          "validation_loss": 0.019652405753731728,
          "train_metric": 0.002393133472651243,
          "validation_metric": 0.0019793184474110603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9711765569991258
        },
        {
          "epoch": 186,
          "updates": 186,
          "samples_seen": 190464,
          "online_loss": 0.023761127144098282,
          "train_loss": 0.02342882566154003,
          "validation_loss": 0.019358089193701744,
          "train_metric": 0.002359665697440505,
          "validation_metric": 0.0019496759632602334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9769424989990512
        },
        {
          "epoch": 187,
          "updates": 187,
          "samples_seen": 191488,
          "online_loss": 0.02342882752418518,
          "train_loss": 0.023106925189495087,
          "validation_loss": 0.019071459770202637,
          "train_metric": 0.0023272449616342783,
          "validation_metric": 0.0019208076409995556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9828582149984868
        },
        {
          "epoch": 188,
          "updates": 188,
          "samples_seen": 192512,
          "online_loss": 0.023106925189495087,
          "train_loss": 0.02279531955718994,
          "validation_loss": 0.018793577328324318,
          "train_metric": 0.0022958610206842422,
          "validation_metric": 0.001892820349894464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9886627359992417
        },
        {
          "epoch": 189,
          "updates": 189,
          "samples_seen": 193536,
          "online_loss": 0.02279532141983509,
          "train_loss": 0.02249387465417385,
          "validation_loss": 0.018526069819927216,
          "train_metric": 0.002265501068904996,
          "validation_metric": 0.0018658781191334128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9945282310000039
        },
        {
          "epoch": 190,
          "updates": 190,
          "samples_seen": 194560,
          "online_loss": 0.022493876516819,
          "train_loss": 0.022202426567673683,
          "validation_loss": 0.018268991261720657,
          "train_metric": 0.0022361474111676216,
          "validation_metric": 0.0018399859545752406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0002530019992264
        },
        {
          "epoch": 191,
          "updates": 191,
          "samples_seen": 195584,
          "online_loss": 0.022202428430318832,
          "train_loss": 0.021920789033174515,
          "validation_loss": 0.01802077516913414,
          "train_metric": 0.0022077816538512707,
          "validation_metric": 0.001814986695535481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0059421579990158
        },
        {
          "epoch": 192,
          "updates": 192,
          "samples_seen": 196608,
          "online_loss": 0.021920789033174515,
          "train_loss": 0.02164873480796814,
          "validation_loss": 0.01778046041727066,
          "train_metric": 0.0021803812123835087,
          "validation_metric": 0.001790782785974443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0117495819995383
        },
        {
          "epoch": 193,
          "updates": 193,
          "samples_seen": 197632,
          "online_loss": 0.02164873480796814,
          "train_loss": 0.021386023610830307,
          "validation_loss": 0.017548788338899612,
          "train_metric": 0.0021539221052080393,
          "validation_metric": 0.0017674496630206704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.09622109799966
        },
        {
          "epoch": 194,
          "updates": 194,
          "samples_seen": 198656,
          "online_loss": 0.021386023610830307,
          "train_loss": 0.021132394671440125,
          "validation_loss": 0.017326459288597107,
          "train_metric": 0.0021283773239701986,
          "validation_metric": 0.0017450572922825813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.102357709998614
        },
        {
          "epoch": 195,
          "updates": 195,
          "samples_seen": 199680,
          "online_loss": 0.021132394671440125,
          "train_loss": 0.020887546241283417,
          "validation_loss": 0.017112668603658676,
          "train_metric": 0.002103717066347599,
          "validation_metric": 0.0017235249979421496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1081435139985842
        },
        {
          "epoch": 196,
          "updates": 196,
          "samples_seen": 200704,
          "online_loss": 0.020887546241283417,
          "train_loss": 0.02065117098391056,
          "validation_loss": 0.016906360164284706,
          "train_metric": 0.002079910133033991,
          "validation_metric": 0.0017027466092258692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.119745021998824
        },
        {
          "epoch": 197,
          "updates": 197,
          "samples_seen": 201728,
          "online_loss": 0.02065116912126541,
          "train_loss": 0.020422937348484993,
          "validation_loss": 0.016708076000213623,
          "train_metric": 0.0020569232292473316,
          "validation_metric": 0.0016827763756737113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1399006609990465
        },
        {
          "epoch": 198,
          "updates": 198,
          "samples_seen": 202752,
          "online_loss": 0.020422935485839844,
          "train_loss": 0.020202511921525,
          "validation_loss": 0.016518671065568924,
          "train_metric": 0.002034723060205579,
          "validation_metric": 0.0016637000953778625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.14692544799982
        },
        {
          "epoch": 199,
          "updates": 199,
          "samples_seen": 203776,
          "online_loss": 0.020202510058879852,
          "train_loss": 0.019989557564258575,
          "validation_loss": 0.016337184235453606,
          "train_metric": 0.0020132751669734716,
          "validation_metric": 0.00164542137645185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1550589999988006
        },
        {
          "epoch": 200,
          "updates": 200,
          "samples_seen": 204800,
          "online_loss": 0.019989555701613426,
          "train_loss": 0.01978372223675251,
          "validation_loss": 0.016161827370524406,
          "train_metric": 0.0019925436936318874,
          "validation_metric": 0.0016277600079774857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.160904323998693
        },
        {
          "epoch": 201,
          "updates": 201,
          "samples_seen": 205824,
          "online_loss": 0.01978371851146221,
          "train_loss": 0.019584650173783302,
          "validation_loss": 0.015992432832717896,
          "train_metric": 0.001972494414076209,
          "validation_metric": 0.0016106993425637484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1687037339997914
        },
        {
          "epoch": 202,
          "updates": 202,
          "samples_seen": 206848,
          "online_loss": 0.01958465203642845,
          "train_loss": 0.019391996785998344,
          "validation_loss": 0.01582961343228817,
          "train_metric": 0.001953090657480061,
          "validation_metric": 0.0015943007310852408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.174453720999736
        },
        {
          "epoch": 203,
          "updates": 203,
          "samples_seen": 207872,
          "online_loss": 0.019391993060708046,
          "train_loss": 0.01920541562139988,
          "validation_loss": 0.015672655776143074,
          "train_metric": 0.0019342991290614009,
          "validation_metric": 0.0015784924617037177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1805212019989995
        },
        {
          "epoch": 204,
          "updates": 204,
          "samples_seen": 208896,
          "online_loss": 0.01920541748404503,
          "train_loss": 0.01902456395328045,
          "validation_loss": 0.015520447865128517,
          "train_metric": 0.0019160843221470714,
          "validation_metric": 0.0015631626592949033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1863724069989985
        },
        {
          "epoch": 205,
          "updates": 205,
          "samples_seen": 209920,
          "online_loss": 0.01902456395328045,
          "train_loss": 0.018849104642868042,
          "validation_loss": 0.015373315662145615,
          "train_metric": 0.0018984127091243863,
          "validation_metric": 0.0015483441529795527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1923548219983786
        },
        {
          "epoch": 206,
          "updates": 206,
          "samples_seen": 210944,
          "online_loss": 0.018849104642868042,
          "train_loss": 0.018678706139326096,
          "validation_loss": 0.015231634490191936,
          "train_metric": 0.0018812509952113032,
          "validation_metric": 0.0015340744284912944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.198174968998501
        },
        {
          "epoch": 207,
          "updates": 207,
          "samples_seen": 211968,
          "online_loss": 0.018678706139326096,
          "train_loss": 0.01851305365562439,
          "validation_loss": 0.015094072557985783,
          "train_metric": 0.0018645668169483542,
          "validation_metric": 0.0015202196082100272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2040799019996484
        },
        {
          "epoch": 208,
          "updates": 208,
          "samples_seen": 212992,
          "online_loss": 0.01851305179297924,
          "train_loss": 0.018351847305893898,
          "validation_loss": 0.014959624037146568,
          "train_metric": 0.0018483306048437953,
          "validation_metric": 0.0015066786436364055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2099051759996655
        },
        {
          "epoch": 209,
          "updates": 209,
          "samples_seen": 214016,
          "online_loss": 0.01835184544324875,
          "train_loss": 0.018194779753684998,
          "validation_loss": 0.01482880488038063,
          "train_metric": 0.0018325112760066986,
          "validation_metric": 0.0014935029903426766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.216072932998941
        },
        {
          "epoch": 210,
          "updates": 210,
          "samples_seen": 215040,
          "online_loss": 0.018194779753684998,
          "train_loss": 0.01804158464074135,
          "validation_loss": 0.014701683074235916,
          "train_metric": 0.0018170821713283658,
          "validation_metric": 0.0014806996332481503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.222954128999845
        },
        {
          "epoch": 211,
          "updates": 211,
          "samples_seen": 216064,
          "online_loss": 0.01804158464074135,
          "train_loss": 0.017891990020871162,
          "validation_loss": 0.014577329158782959,
          "train_metric": 0.0018020154675468802,
          "validation_metric": 0.001468175440095365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.228803574998892
        },
        {
          "epoch": 212,
          "updates": 212,
          "samples_seen": 217088,
          "online_loss": 0.017891988158226013,
          "train_loss": 0.01774575188755989,
          "validation_loss": 0.014455876313149929,
          "train_metric": 0.0017872871831059456,
          "validation_metric": 0.0014559428673237562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2348501199994644
        },
        {
          "epoch": 213,
          "updates": 213,
          "samples_seen": 218112,
          "online_loss": 0.01774575188755989,
          "train_loss": 0.01760263927280903,
          "validation_loss": 0.014337806031107903,
          "train_metric": 0.0017728733364492655,
          "validation_metric": 0.0014440515078604221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.240588184999069
        },
        {
          "epoch": 214,
          "updates": 214,
          "samples_seen": 219136,
          "online_loss": 0.017602641135454178,
          "train_loss": 0.017462443560361862,
          "validation_loss": 0.01422226894646883,
          "train_metric": 0.0017587533220648766,
          "validation_metric": 0.0014324148651212454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.246459648999007
        },
        {
          "epoch": 215,
          "updates": 215,
          "samples_seen": 220160,
          "online_loss": 0.01746244542300701,
          "train_loss": 0.017324956133961678,
          "validation_loss": 0.014108626171946526,
          "train_metric": 0.0017449060687795281,
          "validation_metric": 0.0014209692599251866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2525959969989344
        },
        {
          "epoch": 216,
          "updates": 216,
          "samples_seen": 221184,
          "online_loss": 0.017324956133961678,
          "train_loss": 0.01719001494348049,
          "validation_loss": 0.013997439295053482,
          "train_metric": 0.0017313151620328426,
          "validation_metric": 0.0014097709208726883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.258414570998866
        },
        {
          "epoch": 217,
          "updates": 217,
          "samples_seen": 222208,
          "online_loss": 0.017190013080835342,
          "train_loss": 0.01705744117498398,
          "validation_loss": 0.013888632878661156,
          "train_metric": 0.0017179630231112242,
          "validation_metric": 0.0013988121645525098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.264380839998921
        },
        {
          "epoch": 218,
          "updates": 218,
          "samples_seen": 223232,
          "online_loss": 0.01705744117498398,
          "train_loss": 0.016927087679505348,
          "validation_loss": 0.013781637884676456,
          "train_metric": 0.0017048345180228353,
          "validation_metric": 0.0013880361802875996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2702378229987517
        },
        {
          "epoch": 219,
          "updates": 219,
          "samples_seen": 224256,
          "online_loss": 0.016927089542150497,
          "train_loss": 0.0167988333851099,
          "validation_loss": 0.013676902279257774,
          "train_metric": 0.0016919169574975967,
          "validation_metric": 0.0013774875551462173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2761538999984623
        },
        {
          "epoch": 220,
          "updates": 220,
          "samples_seen": 225280,
          "online_loss": 0.0167988333851099,
          "train_loss": 0.016672546043992043,
          "validation_loss": 0.013574469834566116,
          "train_metric": 0.0016791976522654295,
          "validation_metric": 0.0013671707129105926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.282098027999382
        },
        {
          "epoch": 221,
          "updates": 221,
          "samples_seen": 226304,
          "online_loss": 0.016672546043992043,
          "train_loss": 0.016548123210668564,
          "validation_loss": 0.013473549857735634,
          "train_metric": 0.0016666664741933346,
          "validation_metric": 0.0013570066075772047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.287869123998462
        },
        {
          "epoch": 222,
          "updates": 222,
          "samples_seen": 227328,
          "online_loss": 0.016548123210668564,
          "train_loss": 0.016425468027591705,
          "validation_loss": 0.0133742680773139,
          "train_metric": 0.0016543129459023476,
          "validation_metric": 0.0013470071135088801,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.293716210999264
        },
        {
          "epoch": 223,
          "updates": 223,
          "samples_seen": 228352,
          "online_loss": 0.016425466164946556,
          "train_loss": 0.01630450040102005,
          "validation_loss": 0.013276929967105389,
          "train_metric": 0.0016421295003965497,
          "validation_metric": 0.0013372038956731558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.300451279999834
        },
        {
          "epoch": 224,
          "updates": 224,
          "samples_seen": 229376,
          "online_loss": 0.01630450040102005,
          "train_loss": 0.016185149550437927,
          "validation_loss": 0.013181084766983986,
          "train_metric": 0.0016301089199259877,
          "validation_metric": 0.0013275506207719445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3064769169996
        },
        {
          "epoch": 225,
          "updates": 225,
          "samples_seen": 230400,
          "online_loss": 0.016185151413083076,
          "train_loss": 0.01606736145913601,
          "validation_loss": 0.013086820021271706,
          "train_metric": 0.001618245616555214,
          "validation_metric": 0.0013180566020309925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3123868629991193
        },
        {
          "epoch": 226,
          "updates": 226,
          "samples_seen": 231424,
          "online_loss": 0.01606736145913601,
          "train_loss": 0.015951069071888924,
          "validation_loss": 0.012994443997740746,
          "train_metric": 0.0016065333038568497,
          "validation_metric": 0.0013087530387565494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.318672814999445
        },
        {
          "epoch": 227,
          "updates": 227,
          "samples_seen": 232448,
          "online_loss": 0.015951069071888924,
          "train_loss": 0.01583625003695488,
          "validation_loss": 0.012903437949717045,
          "train_metric": 0.0015949688386172056,
          "validation_metric": 0.0012995870783925056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3245038479999494
        },
        {
          "epoch": 228,
          "updates": 228,
          "samples_seen": 233472,
          "online_loss": 0.01583624817430973,
          "train_loss": 0.01572285033762455,
          "validation_loss": 0.012813728302717209,
          "train_metric": 0.0015835477970540524,
          "validation_metric": 0.0012905518524348736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3306649909991393
        },
        {
          "epoch": 229,
          "updates": 229,
          "samples_seen": 234496,
          "online_loss": 0.01572285033762455,
          "train_loss": 0.015610852278769016,
          "validation_loss": 0.012725596316158772,
          "train_metric": 0.0015722678508609533,
          "validation_metric": 0.0012816755333915353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3364492589989823
        },
        {
          "epoch": 230,
          "updates": 230,
          "samples_seen": 235520,
          "online_loss": 0.015610852278769016,
          "train_loss": 0.015500236302614212,
          "validation_loss": 0.012638724409043789,
          "train_metric": 0.0015611270209774375,
          "validation_metric": 0.0012729258742183447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3420615499999258
        },
        {
          "epoch": 231,
          "updates": 231,
          "samples_seen": 236544,
          "online_loss": 0.015500234439969063,
          "train_loss": 0.015390983782708645,
          "validation_loss": 0.01255321130156517,
          "train_metric": 0.0015501236775889993,
          "validation_metric": 0.001264313468709588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.348129815998618
        },
        {
          "epoch": 232,
          "updates": 232,
          "samples_seen": 237568,
          "online_loss": 0.015390986576676369,
          "train_loss": 0.015283076092600822,
          "validation_loss": 0.012469255365431309,
          "train_metric": 0.001539255608804524,
          "validation_metric": 0.0012558576418086886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.354202589000124
        },
        {
          "epoch": 233,
          "updates": 233,
          "samples_seen": 238592,
          "online_loss": 0.015283077023923397,
          "train_loss": 0.015176510438323021,
          "validation_loss": 0.012386498041450977,
          "train_metric": 0.001528522465378046,
          "validation_metric": 0.0012475226540118456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.440911404999497
        },
        {
          "epoch": 234,
          "updates": 234,
          "samples_seen": 239616,
          "online_loss": 0.015176510438323021,
          "train_loss": 0.015071277506649494,
          "validation_loss": 0.012304961681365967,
          "train_metric": 0.0015179237816482782,
          "validation_metric": 0.0012393107172101736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.44688696799858
        },
        {
          "epoch": 235,
          "updates": 235,
          "samples_seen": 240640,
          "online_loss": 0.015071275644004345,
          "train_loss": 0.014967373572289944,
          "validation_loss": 0.012224791571497917,
          "train_metric": 0.0015074589755386114,
          "validation_metric": 0.001231236383318901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4532817660001456
        },
        {
          "epoch": 236,
          "updates": 236,
          "samples_seen": 241664,
          "online_loss": 0.014967373572289944,
          "train_loss": 0.014864782802760601,
          "validation_loss": 0.01214572973549366,
          "train_metric": 0.00149712641723454,
          "validation_metric": 0.001223273342475295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4590579790001357
        },
        {
          "epoch": 237,
          "updates": 237,
          "samples_seen": 242688,
          "online_loss": 0.014864781871438026,
          "train_loss": 0.01476352196186781,
          "validation_loss": 0.01206795871257782,
          "train_metric": 0.0014869278529658914,
          "validation_metric": 0.0012154405703768134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4652884930001164
        },
        {
          "epoch": 238,
          "updates": 238,
          "samples_seen": 243712,
          "online_loss": 0.014763521030545235,
          "train_loss": 0.014663580805063248,
          "validation_loss": 0.011991524137556553,
          "train_metric": 0.0014768620021641254,
          "validation_metric": 0.0012077424908056855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.471201922999171
        },
        {
          "epoch": 239,
          "updates": 239,
          "samples_seen": 244736,
          "online_loss": 0.014663581736385822,
          "train_loss": 0.014564956538379192,
          "validation_loss": 0.011916174553334713,
          "train_metric": 0.001466928981244564,
          "validation_metric": 0.0012001533759757876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4771129099990503
        },
        {
          "epoch": 240,
          "updates": 240,
          "samples_seen": 245760,
          "online_loss": 0.014564955607056618,
          "train_loss": 0.01446764450520277,
          "validation_loss": 0.011842071078717709,
          "train_metric": 0.00145712832454592,
          "validation_metric": 0.001192689873278141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4833912959984445
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 240
      },
      "best_validation_metric": 0.001192689873278141,
      "config": {
        "label": "Full batch",
        "batch": 1024,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.034535342379628166
    },
    "6a3fa29d2adc_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.725183792412281,
          "train_loss": 0.271520733833313,
          "validation_loss": 0.22441114485263824,
          "train_metric": 0.027346575632691383,
          "validation_metric": 0.022601868957281113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.012577266999869607
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.19251352921128273,
          "train_loss": 0.13974319398403168,
          "validation_loss": 0.1179056391119957,
          "train_metric": 0.014074422419071198,
          "validation_metric": 0.011875025928020477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.024875118000636576
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12273477180860937,
          "train_loss": 0.10032490640878677,
          "validation_loss": 0.08398673683404922,
          "train_metric": 0.010104356333613396,
          "validation_metric": 0.008458836935460567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0368557410001813
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09152746060863137,
          "train_loss": 0.08048801124095917,
          "validation_loss": 0.06612011045217514,
          "train_metric": 0.008106458932161331,
          "validation_metric": 0.006659375037997961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04849434000061592
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07133611291646957,
          "train_loss": 0.06516025960445404,
          "validation_loss": 0.05265787988901138,
          "train_metric": 0.0065627028234303,
          "validation_metric": 0.005303508602082729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.060471692000646726
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06069809605833143,
          "train_loss": 0.05711313337087631,
          "validation_loss": 0.047087281942367554,
          "train_metric": 0.005752225872129202,
          "validation_metric": 0.004742457997053862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07256131500071206
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05503121088258922,
          "train_loss": 0.05024784803390503,
          "validation_loss": 0.041228801012039185,
          "train_metric": 0.005060779396444559,
          "validation_metric": 0.004152413923293352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08443490300123813
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04905405070167035,
          "train_loss": 0.04574151337146759,
          "validation_loss": 0.03793175891041756,
          "train_metric": 0.004606917500495911,
          "validation_metric": 0.0038203480653464794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09630337999988114
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04519221524242312,
          "train_loss": 0.05024033412337303,
          "validation_loss": 0.04192633181810379,
          "train_metric": 0.005060022231191397,
          "validation_metric": 0.004222666844725609,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11016259499956504
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.049359321128576994,
          "train_loss": 0.046037573367357254,
          "validation_loss": 0.039142146706581116,
          "train_metric": 0.004636736121028662,
          "validation_metric": 0.003942254465073347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.122790597999483
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04093247506534681,
          "train_loss": 0.03829994797706604,
          "validation_loss": 0.03261260688304901,
          "train_metric": 0.0038574300706386566,
          "validation_metric": 0.003284622449427843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13470523999967554
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.036824317183345556,
          "train_loss": 0.032820843160152435,
          "validation_loss": 0.028341077268123627,
          "train_metric": 0.003305594902485609,
          "validation_metric": 0.002854409161955118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14637159500125563
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03441062942147255,
          "train_loss": 0.03700372204184532,
          "validation_loss": 0.02941253036260605,
          "train_metric": 0.0037268793676048517,
          "validation_metric": 0.0029623224399983883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15781071899982635
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.032163326046429574,
          "train_loss": 0.02908608876168728,
          "validation_loss": 0.024541635066270828,
          "train_metric": 0.0029294448904693127,
          "validation_metric": 0.0024717431515455246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16964741500123637
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03053868602728471,
          "train_loss": 0.02759496122598648,
          "validation_loss": 0.023318752646446228,
          "train_metric": 0.0027792633045464754,
          "validation_metric": 0.0023485796991735697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18159412800014252
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.026482005312573165,
          "train_loss": 0.027983058243989944,
          "validation_loss": 0.023301327601075172,
          "train_metric": 0.002818351611495018,
          "validation_metric": 0.002346824621781707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1935068050006521
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.02441761939553544,
          "train_loss": 0.023983746767044067,
          "validation_loss": 0.02047683112323284,
          "train_metric": 0.002415555063635111,
          "validation_metric": 0.0020623516757041216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2057420290002483
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.023378114216029644,
          "train_loss": 0.02086450532078743,
          "validation_loss": 0.017464669421315193,
          "train_metric": 0.002101396443322301,
          "validation_metric": 0.0017589774215593934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21758032100115088
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.02140462800161913,
          "train_loss": 0.02289021760225296,
          "validation_loss": 0.018739325925707817,
          "train_metric": 0.002305419184267521,
          "validation_metric": 0.001887356280349195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22884577100012393
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02563406975241378,
          "train_loss": 0.02189042419195175,
          "validation_loss": 0.017806872725486755,
          "train_metric": 0.0022047236561775208,
          "validation_metric": 0.001793443108908832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24029510500076867
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023129807494115084,
          "train_loss": 0.020913323387503624,
          "validation_loss": 0.017478393390774727,
          "train_metric": 0.0021063131280243397,
          "validation_metric": 0.00176035996992141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2513334089999262
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02696609584381804,
          "train_loss": 0.026205450296401978,
          "validation_loss": 0.021345825865864754,
          "train_metric": 0.0026393167208880186,
          "validation_metric": 0.002149873413145542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.262370168000416
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.019721148069947958,
          "train_loss": 0.01606700010597706,
          "validation_loss": 0.01406087540090084,
          "train_metric": 0.0016182095278054476,
          "validation_metric": 0.0014161599101498723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27429968300020846
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.017273633362492546,
          "train_loss": 0.014825965277850628,
          "validation_loss": 0.012481079436838627,
          "train_metric": 0.0014932170743122697,
          "validation_metric": 0.0012570485705509782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2861230300004536
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019757675006985664,
          "train_loss": 0.01432263944298029,
          "validation_loss": 0.012306571938097477,
          "train_metric": 0.001442523906007409,
          "validation_metric": 0.0012394727673381567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29859581900018384
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.017510230012703687,
          "train_loss": 0.014158128760755062,
          "validation_loss": 0.01238947082310915,
          "train_metric": 0.0014259549789130688,
          "validation_metric": 0.0012478221906349063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31268683100097405
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015726784709841013,
          "train_loss": 0.013463691808283329,
          "validation_loss": 0.01134172547608614,
          "train_metric": 0.0013560139341279864,
          "validation_metric": 0.001142297056503594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3245317849996354
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.013707112200791016,
          "train_loss": 0.012812243774533272,
          "validation_loss": 0.01123434491455555,
          "train_metric": 0.001290402258746326,
          "validation_metric": 0.0011314821895211935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3370261190011661
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016051386279286817,
          "train_loss": 0.02502411976456642,
          "validation_loss": 0.020168717950582504,
          "train_metric": 0.0025203379336744547,
          "validation_metric": 0.002031319309026003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3488584870010527
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01542237150715664,
          "train_loss": 0.014315691776573658,
          "validation_loss": 0.011095796711742878,
          "train_metric": 0.001441823784261942,
          "validation_metric": 0.001117528066970408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36048203900099907
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.015414124500239268,
          "train_loss": 0.021503446623682976,
          "validation_loss": 0.018202824518084526,
          "train_metric": 0.0021657487377524376,
          "validation_metric": 0.0018333217594772577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37245290700047917
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.020806275395443663,
          "train_loss": 0.017160432413220406,
          "validation_loss": 0.012870152480900288,
          "train_metric": 0.0017283360939472914,
          "validation_metric": 0.0012962345499545336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38395921400115185
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01630209616268985,
          "train_loss": 0.011724606156349182,
          "validation_loss": 0.009741163812577724,
          "train_metric": 0.0011808593990281224,
          "validation_metric": 0.0009810944320634007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3956770740005595
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013805997223244049,
          "train_loss": 0.023595713078975677,
          "validation_loss": 0.019702864810824394,
          "train_metric": 0.0023764735087752342,
          "validation_metric": 0.0019844004418700933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40744897300101
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.015536964187049307,
          "train_loss": 0.011063741520047188,
          "validation_loss": 0.00977888610213995,
          "train_metric": 0.0011142995208501816,
          "validation_metric": 0.000984893529675901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4182426210008998
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.012316050342633389,
          "train_loss": 0.011551490053534508,
          "validation_loss": 0.0093732550740242,
          "train_metric": 0.0011634236434474587,
          "validation_metric": 0.0009440399007871747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.429826679001053
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.01322234922554344,
          "train_loss": 0.01286236196756363,
          "validation_loss": 0.010114856995642185,
          "train_metric": 0.0012954497942700982,
          "validation_metric": 0.001018731389194727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44125166999947396
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011155385203892365,
          "train_loss": 0.011489743366837502,
          "validation_loss": 0.010611145757138729,
          "train_metric": 0.0011572049697861075,
          "validation_metric": 0.0010687157046049833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45227475899991987
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.011601212609093636,
          "train_loss": 0.014726401306688786,
          "validation_loss": 0.010290934704244137,
          "train_metric": 0.0014831890584900975,
          "validation_metric": 0.0010364654008299112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4636076429997047
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012412100812071003,
          "train_loss": 0.012658714316785336,
          "validation_loss": 0.01127684861421585,
          "train_metric": 0.0012749392772093415,
          "validation_metric": 0.00113576278090477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4748377360010636
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011675719724735245,
          "train_loss": 0.009901653975248337,
          "validation_loss": 0.007692788261920214,
          "train_metric": 0.0009972583502531052,
          "validation_metric": 0.0007747894851490855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4866531420011597
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010013210485340096,
          "train_loss": 0.008797609247267246,
          "validation_loss": 0.007866949774324894,
          "train_metric": 0.0008860630914568901,
          "validation_metric": 0.0007923304219730198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4981093080004939
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009386818463099189,
          "train_loss": 0.013002975843846798,
          "validation_loss": 0.009492841549217701,
          "train_metric": 0.001309612300246954,
          "validation_metric": 0.0009560841717757285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5124796100008098
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01224605915194843,
          "train_loss": 0.009541220031678677,
          "validation_loss": 0.0073880041018128395,
          "train_metric": 0.0009609567932784557,
          "validation_metric": 0.0007440927438437939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5250621830000455
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.01069485695916228,
          "train_loss": 0.01067455392330885,
          "validation_loss": 0.009969848208129406,
          "train_metric": 0.0010751018999144435,
          "validation_metric": 0.0010041267378255725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5367191140012437
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010661405161954463,
          "train_loss": 0.009650606662034988,
          "validation_loss": 0.008344125933945179,
          "train_metric": 0.0009719737572595477,
          "validation_metric": 0.0008403898100368679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5483034680000856
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011100387433543801,
          "train_loss": 0.015353727154433727,
          "validation_loss": 0.010953432880342007,
          "train_metric": 0.0015463712625205517,
          "validation_metric": 0.0011031896574422717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5599500200005423
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010565859702182934,
          "train_loss": 0.008216824382543564,
          "validation_loss": 0.007050653453916311,
          "train_metric": 0.0008275684667751193,
          "validation_metric": 0.0007101160590536892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5719077920002746
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00831423222552985,
          "train_loss": 0.007751990109682083,
          "validation_loss": 0.007047897204756737,
          "train_metric": 0.0007807521033100784,
          "validation_metric": 0.000709838408511132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5840091509999183
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.008484421006869525,
          "train_loss": 0.008210123516619205,
          "validation_loss": 0.006662710104137659,
          "train_metric": 0.0008268934907391667,
          "validation_metric": 0.0006710437010042369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5965172810010699
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009121570168645121,
          "train_loss": 0.008138127624988556,
          "validation_loss": 0.006662833970040083,
          "train_metric": 0.0008196425042115152,
          "validation_metric": 0.0006710562738589942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6085914659997798
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.012653169003897347,
          "train_loss": 0.008039749227464199,
          "validation_loss": 0.006858522072434425,
          "train_metric": 0.0008097341051325202,
          "validation_metric": 0.0006907652714289725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6203038740004558
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008644956105854362,
          "train_loss": 0.008325216360390186,
          "validation_loss": 0.00712231732904911,
          "train_metric": 0.0008384853135794401,
          "validation_metric": 0.0007173336343839765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6320054360003269
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.00779974230681546,
          "train_loss": 0.008252358995378017,
          "validation_loss": 0.007306789513677359,
          "train_metric": 0.0008311473648063838,
          "validation_metric": 0.0007359131122939289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6438596640000469
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.00925846655445639,
          "train_loss": 0.00652270158752799,
          "validation_loss": 0.0057301875203847885,
          "train_metric": 0.0006569424876943231,
          "validation_metric": 0.0005771234282292426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6560350310010108
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.012231946529936977,
          "train_loss": 0.00858000572770834,
          "validation_loss": 0.00622341874986887,
          "train_metric": 0.0008641468011774123,
          "validation_metric": 0.0006267998833209276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6677834010006336
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008446793217444792,
          "train_loss": 0.00689212791621685,
          "validation_loss": 0.006349670700728893,
          "train_metric": 0.0006941498722881079,
          "validation_metric": 0.0006395156960934401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6797941830009222
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.008316159102832898,
          "train_loss": 0.007118558511137962,
          "validation_loss": 0.006148797459900379,
          "train_metric": 0.0007169551681727171,
          "validation_metric": 0.0006192844011820853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6949392919996171
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007200873791589402,
          "train_loss": 0.006566303316503763,
          "validation_loss": 0.005335439462214708,
          "train_metric": 0.0006613340810872614,
          "validation_metric": 0.0005373659078031778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7095565520012315
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006584036775166169,
          "train_loss": 0.006239462643861771,
          "validation_loss": 0.005712360609322786,
          "train_metric": 0.0006284158444032073,
          "validation_metric": 0.0005753280129283667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7210715000001073
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.006660603481577709,
          "train_loss": 0.006314736790955067,
          "validation_loss": 0.005067717283964157,
          "train_metric": 0.0006359972176142037,
          "validation_metric": 0.0005104019655846059,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7352938149997499
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.006349309740471654,
          "train_loss": 0.005507147405296564,
          "validation_loss": 0.004698514938354492,
          "train_metric": 0.0005546596948988736,
          "validation_metric": 0.00047321722377091646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7658382179997716
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.008569663259549998,
          "train_loss": 0.015207894146442413,
          "validation_loss": 0.0099509721621871,
          "train_metric": 0.001531683374196291,
          "validation_metric": 0.0010022253263741732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7883393940010137
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.010702520652557723,
          "train_loss": 0.0073378500528633595,
          "validation_loss": 0.007015361916273832,
          "train_metric": 0.0007390414248220623,
          "validation_metric": 0.0007065616082400084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8209316740012582
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.007163544389186427,
          "train_loss": 0.0053948210552334785,
          "validation_loss": 0.0044744606129825115,
          "train_metric": 0.0005433466285467148,
          "validation_metric": 0.00045065133599564433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8618062490004377
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005669868871336803,
          "train_loss": 0.005300572142004967,
          "validation_loss": 0.0045133717358112335,
          "train_metric": 0.0005338541814126074,
          "validation_metric": 0.00045457022497430444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8868657969997003
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0069226891282596625,
          "train_loss": 0.007420714013278484,
          "validation_loss": 0.006919790990650654,
          "train_metric": 0.0007473871810361743,
          "validation_metric": 0.000696935981977731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9161741770003573
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00731872973847203,
          "train_loss": 0.006452545523643494,
          "validation_loss": 0.0051038870587944984,
          "train_metric": 0.00064987683435902,
          "validation_metric": 0.0005140448338352144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9947886999998445
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.009972294181352481,
          "train_loss": 0.0056060985662043095,
          "validation_loss": 0.004278527572751045,
          "train_metric": 0.0005646257195621729,
          "validation_metric": 0.00043091760016977787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0159553909998067
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.007715228144661523,
          "train_loss": 0.00778129231184721,
          "validation_loss": 0.007210019044578075,
          "train_metric": 0.0007837032317183912,
          "validation_metric": 0.0007261668215505779,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0357240659996023
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006288037599006202,
          "train_loss": 0.005706281866878271,
          "validation_loss": 0.00420799758285284,
          "train_metric": 0.0005747157265432179,
          "validation_metric": 0.00042381405364722013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0571725640002114
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00576221395022003,
          "train_loss": 0.005137464962899685,
          "validation_loss": 0.004070205148309469,
          "train_metric": 0.000517426582518965,
          "validation_metric": 0.0004099361540284008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.078170045000661
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.005247840948868543,
          "train_loss": 0.006295354105532169,
          "validation_loss": 0.005427327938377857,
          "train_metric": 0.0006340449908748269,
          "validation_metric": 0.0005466206348501146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0976068660002056
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005847955937497318,
          "train_loss": 0.0056092459708452225,
          "validation_loss": 0.004303757566958666,
          "train_metric": 0.0005649426602758467,
          "validation_metric": 0.0004334586556069553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1357552520003082
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00761535503261257,
          "train_loss": 0.0046490756794810295,
          "validation_loss": 0.004038749262690544,
          "train_metric": 0.00046823782031424344,
          "validation_metric": 0.00040676796925254166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1630575680010224
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0055943846455193125,
          "train_loss": 0.005045777186751366,
          "validation_loss": 0.003924061544239521,
          "train_metric": 0.000508192228153348,
          "validation_metric": 0.0003952171537093818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1922197570002027
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.00631263930699788,
          "train_loss": 0.005989368073642254,
          "validation_loss": 0.005381503142416477,
          "train_metric": 0.0006032272358424962,
          "validation_metric": 0.0005420053494162858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.217024706000302
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.005129424927872606,
          "train_loss": 0.004593792837113142,
          "validation_loss": 0.0038168870378285646,
          "train_metric": 0.00046266999561339617,
          "validation_metric": 0.0003844228631351143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2403950990010344
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00618596728600096,
          "train_loss": 0.004681388847529888,
          "validation_loss": 0.004123073071241379,
          "train_metric": 0.0004714923270512372,
          "validation_metric": 0.00041526081622578204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2574821729995165
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004919681974570267,
          "train_loss": 0.004786643199622631,
          "validation_loss": 0.0038449480198323727,
          "train_metric": 0.00048209322267211974,
          "validation_metric": 0.00038724910700693727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2745655580001767
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006906665556016378,
          "train_loss": 0.012297742068767548,
          "validation_loss": 0.009444266557693481,
          "train_metric": 0.0012385835871100426,
          "validation_metric": 0.0009511919342912734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2921656250000524
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.010686725392588414,
          "train_loss": 0.0064812321215868,
          "validation_loss": 0.0057849641889333725,
          "train_metric": 0.00065276597160846,
          "validation_metric": 0.0005826404667459428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3085371930010297
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.007700789232330862,
          "train_loss": 0.004105992149561644,
          "validation_loss": 0.003409368684515357,
          "train_metric": 0.0004135404305998236,
          "validation_metric": 0.0003433791280258447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.326111686999866
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.004603917019267101,
          "train_loss": 0.004845749586820602,
          "validation_loss": 0.0035455594770610332,
          "train_metric": 0.0004880461492575705,
          "validation_metric": 0.00035709579242393374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3429897950009035
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.007054458052152768,
          "train_loss": 0.015696968883275986,
          "validation_loss": 0.012972714379429817,
          "train_metric": 0.0015809413744136691,
          "validation_metric": 0.0013065641978755593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3602315680000174
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.011451014113845304,
          "train_loss": 0.020557669922709465,
          "validation_loss": 0.013468781486153603,
          "train_metric": 0.0020704930648207664,
          "validation_metric": 0.0013565262779593468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3774377280005865
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.015485930867725983,
          "train_loss": 0.00581009965389967,
          "validation_loss": 0.004421717952936888,
          "train_metric": 0.0005851719761267304,
          "validation_metric": 0.0004453392466530204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3943433490003372
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.006737215720931999,
          "train_loss": 0.0061674462631344795,
          "validation_loss": 0.0055349646136164665,
          "train_metric": 0.0006211625295691192,
          "validation_metric": 0.0005574613460339606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4145256730007532
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005034035624703392,
          "train_loss": 0.0039005510043352842,
          "validation_loss": 0.0032008232083171606,
          "train_metric": 0.00039284920785576105,
          "validation_metric": 0.000322375213727355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4390077749994816
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004518714900768828,
          "train_loss": 0.0038000443018972874,
          "validation_loss": 0.0031561877112835646,
          "train_metric": 0.0003827265463769436,
          "validation_metric": 0.0003178796614520252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.457939271000214
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.006248384044738486,
          "train_loss": 0.005066498648375273,
          "validation_loss": 0.004222129005938768,
          "train_metric": 0.0005102792056277394,
          "validation_metric": 0.00042523740557953715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4872176070002752
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.005300535114656668,
          "train_loss": 0.004341588355600834,
          "validation_loss": 0.0038378548342734575,
          "train_metric": 0.0004372688417788595,
          "validation_metric": 0.00038653466617688537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5037818230011908
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004943848747643642,
          "train_loss": 0.0077809556387364864,
          "validation_loss": 0.005293975584208965,
          "train_metric": 0.0007836693548597395,
          "validation_metric": 0.0005331897991709411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5201812160012196
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004287212868803181,
          "train_loss": 0.008774102665483952,
          "validation_loss": 0.007212408818304539,
          "train_metric": 0.0008836955530568957,
          "validation_metric": 0.0007264073938131332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5365223349999724
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0035014182722079568,
          "train_loss": 0.004472360014915466,
          "validation_loss": 0.004246356897056103,
          "train_metric": 0.00045043969294056296,
          "validation_metric": 0.0004276775289326906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5528499930005637
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004747023289382923,
          "train_loss": 0.00342956674285233,
          "validation_loss": 0.0028597929049283266,
          "train_metric": 0.000345413398463279,
          "validation_metric": 0.00028802783344872296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5697019979998004
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005751226031861734,
          "train_loss": 0.006062327418476343,
          "validation_loss": 0.004158920142799616,
          "train_metric": 0.0006105754291638732,
          "validation_metric": 0.0004188711754977703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5863522560011916
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.006527146659209393,
          "train_loss": 0.004064810462296009,
          "validation_loss": 0.003415434155613184,
          "train_metric": 0.000409392814617604,
          "validation_metric": 0.00034399001742713153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.602920774001177
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.005003856465918943,
          "train_loss": 0.0043891118839383125,
          "validation_loss": 0.0036584429908543825,
          "train_metric": 0.00044205531594343483,
          "validation_metric": 0.000368465029168874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6331259300004604
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.005998564207402524,
          "train_loss": 0.004563015885651112,
          "validation_loss": 0.004187696613371372,
          "train_metric": 0.00045957029215060174,
          "validation_metric": 0.0004217694513499737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.674658036999972
        }
      ],
      "best": {
        "epoch": 96,
        "updates": 1536
      },
      "best_validation_metric": 0.00028802783344872296,
      "config": {
        "label": "Batch 64",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01697138277951219
    },
    "1fb401058891_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 1,
          "samples_seen": 1024,
          "online_loss": 0.9984928965568542,
          "train_loss": 0.9801727533340454,
          "validation_loss": 0.707870364189148,
          "train_metric": 0.09871941059827805,
          "validation_metric": 0.07129411399364471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.011126183000669698
        },
        {
          "epoch": 2,
          "updates": 2,
          "samples_seen": 2048,
          "online_loss": 0.9801727533340454,
          "train_loss": 0.9603869915008545,
          "validation_loss": 0.6948423981666565,
          "train_metric": 0.0967266634106636,
          "validation_metric": 0.06998198479413986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01994863900108612
        },
        {
          "epoch": 3,
          "updates": 3,
          "samples_seen": 3072,
          "online_loss": 0.9603869915008545,
          "train_loss": 0.9384909868240356,
          "validation_loss": 0.6804297566413879,
          "train_metric": 0.09452136605978012,
          "validation_metric": 0.06853038817644119,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.028126966000854736
        },
        {
          "epoch": 4,
          "updates": 4,
          "samples_seen": 4096,
          "online_loss": 0.9384909272193909,
          "train_loss": 0.9138616919517517,
          "validation_loss": 0.6641983389854431,
          "train_metric": 0.09204079955816269,
          "validation_metric": 0.06689562648534775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1615793510009098
        },
        {
          "epoch": 5,
          "updates": 5,
          "samples_seen": 5120,
          "online_loss": 0.9138616919517517,
          "train_loss": 0.8859156370162964,
          "validation_loss": 0.6457259654998779,
          "train_metric": 0.08922617882490158,
          "validation_metric": 0.0650351494550705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1701863440011948
        },
        {
          "epoch": 6,
          "updates": 6,
          "samples_seen": 6144,
          "online_loss": 0.8859155774116516,
          "train_loss": 0.8541439175605774,
          "validation_loss": 0.6246242523193359,
          "train_metric": 0.08602624386548996,
          "validation_metric": 0.06290987133979797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.178605435999998
        },
        {
          "epoch": 7,
          "updates": 7,
          "samples_seen": 7168,
          "online_loss": 0.8541439771652222,
          "train_loss": 0.818126916885376,
          "validation_loss": 0.6005696654319763,
          "train_metric": 0.08239874243736267,
          "validation_metric": 0.06048717349767685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18644203500116419
        },
        {
          "epoch": 8,
          "updates": 8,
          "samples_seen": 8192,
          "online_loss": 0.8181267976760864,
          "train_loss": 0.7775458693504333,
          "validation_loss": 0.573311448097229,
          "train_metric": 0.07831156998872757,
          "validation_metric": 0.05774183198809624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19471262200022466
        },
        {
          "epoch": 9,
          "updates": 9,
          "samples_seen": 9216,
          "online_loss": 0.7775458097457886,
          "train_loss": 0.7321982979774475,
          "validation_loss": 0.5426748991012573,
          "train_metric": 0.07374434173107147,
          "validation_metric": 0.05465622991323471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2029997560002812
        },
        {
          "epoch": 10,
          "updates": 10,
          "samples_seen": 10240,
          "online_loss": 0.7321983575820923,
          "train_loss": 0.6820284724235535,
          "validation_loss": 0.5085822343826294,
          "train_metric": 0.06869141012430191,
          "validation_metric": 0.0512225367128849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21119552300115174
        },
        {
          "epoch": 11,
          "updates": 11,
          "samples_seen": 11264,
          "online_loss": 0.6820285320281982,
          "train_loss": 0.6271827816963196,
          "validation_loss": 0.4710938334465027,
          "train_metric": 0.06316754966974258,
          "validation_metric": 0.0474468432366848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21974776500064763
        },
        {
          "epoch": 12,
          "updates": 12,
          "samples_seen": 12288,
          "online_loss": 0.6271827220916748,
          "train_loss": 0.5680776834487915,
          "validation_loss": 0.4304536283016205,
          "train_metric": 0.05721469968557358,
          "validation_metric": 0.04335371404886246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22813949500050512
        },
        {
          "epoch": 13,
          "updates": 13,
          "samples_seen": 13312,
          "online_loss": 0.5680776238441467,
          "train_loss": 0.5054815411567688,
          "validation_loss": 0.38714104890823364,
          "train_metric": 0.05091025307774544,
          "validation_metric": 0.038991428911685944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23623436200068682
        },
        {
          "epoch": 14,
          "updates": 14,
          "samples_seen": 14336,
          "online_loss": 0.5054816007614136,
          "train_loss": 0.4406189024448395,
          "validation_loss": 0.3419446349143982,
          "train_metric": 0.04437752440571785,
          "validation_metric": 0.034439411014318466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2441963700002816
        },
        {
          "epoch": 15,
          "updates": 15,
          "samples_seen": 15360,
          "online_loss": 0.44061893224716187,
          "train_loss": 0.37529289722442627,
          "validation_loss": 0.2960549294948578,
          "train_metric": 0.03779812902212143,
          "validation_metric": 0.029817568138241768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2523969110006874
        },
        {
          "epoch": 16,
          "updates": 16,
          "samples_seen": 16384,
          "online_loss": 0.37529292702674866,
          "train_loss": 0.31199777126312256,
          "validation_loss": 0.25114503502845764,
          "train_metric": 0.03142327442765236,
          "validation_metric": 0.025294410064816475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2604893820007419
        },
        {
          "epoch": 17,
          "updates": 17,
          "samples_seen": 17408,
          "online_loss": 0.31199777126312256,
          "train_loss": 0.25395697355270386,
          "validation_loss": 0.2093973308801651,
          "train_metric": 0.025577614083886147,
          "validation_metric": 0.02108973078429699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26890426200043294
        },
        {
          "epoch": 18,
          "updates": 18,
          "samples_seen": 18432,
          "online_loss": 0.25395697355270386,
          "train_loss": 0.20501048862934113,
          "validation_loss": 0.17344781756401062,
          "train_metric": 0.02064790576696396,
          "validation_metric": 0.017469028010964394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.277227758000663
        },
        {
          "epoch": 19,
          "updates": 19,
          "samples_seen": 19456,
          "online_loss": 0.20501047372817993,
          "train_loss": 0.1692221313714981,
          "validation_loss": 0.1461627185344696,
          "train_metric": 0.017043432220816612,
          "validation_metric": 0.014720974490046501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2851363890003995
        },
        {
          "epoch": 20,
          "updates": 20,
          "samples_seen": 20480,
          "online_loss": 0.1692221313714981,
          "train_loss": 0.1499544233083725,
          "validation_loss": 0.1300608068704605,
          "train_metric": 0.01510285958647728,
          "validation_metric": 0.013099249452352524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29310964600153966
        },
        {
          "epoch": 21,
          "updates": 21,
          "samples_seen": 21504,
          "online_loss": 0.1499544233083725,
          "train_loss": 0.14820151031017303,
          "validation_loss": 0.12625624239444733,
          "train_metric": 0.014926312491297722,
          "validation_metric": 0.012716066092252731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3012675999998464
        },
        {
          "epoch": 22,
          "updates": 22,
          "samples_seen": 22528,
          "online_loss": 0.14820151031017303,
          "train_loss": 0.16058266162872314,
          "validation_loss": 0.13311734795570374,
          "train_metric": 0.01617329753935337,
          "validation_metric": 0.013407093472778797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3086849820010684
        },
        {
          "epoch": 23,
          "updates": 23,
          "samples_seen": 23552,
          "online_loss": 0.16058264672756195,
          "train_loss": 0.17885594069957733,
          "validation_loss": 0.14566545188426971,
          "train_metric": 0.018013713881373405,
          "validation_metric": 0.01467089168727398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31608589900133666
        },
        {
          "epoch": 24,
          "updates": 24,
          "samples_seen": 24576,
          "online_loss": 0.17885592579841614,
          "train_loss": 0.1931406855583191,
          "validation_loss": 0.15717023611068726,
          "train_metric": 0.019452422857284546,
          "validation_metric": 0.01582960970699787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3236649190002936
        },
        {
          "epoch": 25,
          "updates": 25,
          "samples_seen": 25600,
          "online_loss": 0.1931406855583191,
          "train_loss": 0.19712764024734497,
          "validation_loss": 0.16245920956134796,
          "train_metric": 0.019853973761200905,
          "validation_metric": 0.016362296417355537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3311783990011463
        },
        {
          "epoch": 26,
          "updates": 26,
          "samples_seen": 26624,
          "online_loss": 0.19712765514850616,
          "train_loss": 0.19032138586044312,
          "validation_loss": 0.1599966287612915,
          "train_metric": 0.019168473780155182,
          "validation_metric": 0.01611427403986454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3383836180000799
        },
        {
          "epoch": 27,
          "updates": 27,
          "samples_seen": 27648,
          "online_loss": 0.19032138586044312,
          "train_loss": 0.1762879490852356,
          "validation_loss": 0.15137168765068054,
          "train_metric": 0.01775507815182209,
          "validation_metric": 0.015245603397488594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3456148170007509
        },
        {
          "epoch": 28,
          "updates": 28,
          "samples_seen": 28672,
          "online_loss": 0.1762879341840744,
          "train_loss": 0.159781813621521,
          "validation_loss": 0.13958294689655304,
          "train_metric": 0.016092637553811073,
          "validation_metric": 0.014058283530175686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35277721600141376
        },
        {
          "epoch": 29,
          "updates": 29,
          "samples_seen": 29696,
          "online_loss": 0.1597817987203598,
          "train_loss": 0.14472904801368713,
          "validation_loss": 0.12753748893737793,
          "train_metric": 0.014576580375432968,
          "validation_metric": 0.012845107354223728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36015774200132
        },
        {
          "epoch": 30,
          "updates": 30,
          "samples_seen": 30720,
          "online_loss": 0.14472904801368713,
          "train_loss": 0.13338318467140198,
          "validation_loss": 0.11724204570055008,
          "train_metric": 0.013433866202831268,
          "validation_metric": 0.011808189563453197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36844958100118674
        },
        {
          "epoch": 31,
          "updates": 31,
          "samples_seen": 31744,
          "online_loss": 0.13338318467140198,
          "train_loss": 0.12634509801864624,
          "validation_loss": 0.10961552709341049,
          "train_metric": 0.012725015170872211,
          "validation_metric": 0.011040073819458485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3767128760009655
        },
        {
          "epoch": 32,
          "updates": 32,
          "samples_seen": 32768,
          "online_loss": 0.12634509801864624,
          "train_loss": 0.12302938848733902,
          "validation_loss": 0.10467509180307388,
          "train_metric": 0.012391069903969765,
          "validation_metric": 0.0105424914509058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3865156380015833
        },
        {
          "epoch": 33,
          "updates": 33,
          "samples_seen": 33792,
          "online_loss": 0.12302938848733902,
          "train_loss": 0.12224037200212479,
          "validation_loss": 0.1018718034029007,
          "train_metric": 0.01231160294264555,
          "validation_metric": 0.010260154493153095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39516931700018176
        },
        {
          "epoch": 34,
          "updates": 34,
          "samples_seen": 34816,
          "online_loss": 0.12224036455154419,
          "train_loss": 0.12265750020742416,
          "validation_loss": 0.10042364150285721,
          "train_metric": 0.012353615835309029,
          "validation_metric": 0.01011430099606514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40349622200119484
        },
        {
          "epoch": 35,
          "updates": 35,
          "samples_seen": 35840,
          "online_loss": 0.12265750765800476,
          "train_loss": 0.12315250188112259,
          "validation_loss": 0.09956672787666321,
          "train_metric": 0.012403469532728195,
          "validation_metric": 0.010027995333075523,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4121739840011287
        },
        {
          "epoch": 36,
          "updates": 36,
          "samples_seen": 36864,
          "online_loss": 0.12315249443054199,
          "train_loss": 0.12294331192970276,
          "validation_loss": 0.09870384633541107,
          "train_metric": 0.012382399290800095,
          "validation_metric": 0.009941089898347855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4208162539998739
        },
        {
          "epoch": 37,
          "updates": 37,
          "samples_seen": 37888,
          "online_loss": 0.12294329702854156,
          "train_loss": 0.12162825465202332,
          "validation_loss": 0.09746500849723816,
          "train_metric": 0.012249953113496304,
          "validation_metric": 0.00981631875038147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4292769280000357
        },
        {
          "epoch": 38,
          "updates": 38,
          "samples_seen": 38912,
          "online_loss": 0.12162826210260391,
          "train_loss": 0.11914759129285812,
          "validation_loss": 0.09570664912462234,
          "train_metric": 0.012000108137726784,
          "validation_metric": 0.009639223106205463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4382436579999194
        },
        {
          "epoch": 39,
          "updates": 39,
          "samples_seen": 39936,
          "online_loss": 0.11914759129285812,
          "train_loss": 0.11570756882429123,
          "validation_loss": 0.09347426891326904,
          "train_metric": 0.011653644032776356,
          "validation_metric": 0.009414385072886944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44688533400039887
        },
        {
          "epoch": 40,
          "updates": 40,
          "samples_seen": 40960,
          "online_loss": 0.11570758372545242,
          "train_loss": 0.11168835312128067,
          "validation_loss": 0.09094579517841339,
          "train_metric": 0.011248841881752014,
          "validation_metric": 0.009159727953374386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45519691099980264
        },
        {
          "epoch": 41,
          "updates": 41,
          "samples_seen": 41984,
          "online_loss": 0.11168834567070007,
          "train_loss": 0.10754545778036118,
          "validation_loss": 0.08836610615253448,
          "train_metric": 0.01083158515393734,
          "validation_metric": 0.008899910375475883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46364824100055557
        },
        {
          "epoch": 42,
          "updates": 42,
          "samples_seen": 43008,
          "online_loss": 0.10754545778036118,
          "train_loss": 0.10371240228414536,
          "validation_loss": 0.08597948402166367,
          "train_metric": 0.010445533320307732,
          "validation_metric": 0.008659538812935352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4724519310002506
        },
        {
          "epoch": 43,
          "updates": 43,
          "samples_seen": 44032,
          "online_loss": 0.10371240228414536,
          "train_loss": 0.10051248967647552,
          "validation_loss": 0.08396806567907333,
          "train_metric": 0.010123251006007195,
          "validation_metric": 0.008456956595182419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5793504720004421
        },
        {
          "epoch": 44,
          "updates": 44,
          "samples_seen": 45056,
          "online_loss": 0.10051250457763672,
          "train_loss": 0.0980939120054245,
          "validation_loss": 0.08240629732608795,
          "train_metric": 0.009879658930003643,
          "validation_metric": 0.008299660868942738,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5875640900012513
        },
        {
          "epoch": 45,
          "updates": 45,
          "samples_seen": 46080,
          "online_loss": 0.0980939045548439,
          "train_loss": 0.09640473127365112,
          "validation_loss": 0.0812433734536171,
          "train_metric": 0.009709532372653484,
          "validation_metric": 0.008182534947991371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5956410630005848
        },
        {
          "epoch": 46,
          "updates": 46,
          "samples_seen": 47104,
          "online_loss": 0.09640473127365112,
          "train_loss": 0.09522069990634918,
          "validation_loss": 0.08032234013080597,
          "train_metric": 0.009590279310941696,
          "validation_metric": 0.008089771494269371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6041606150010921
        },
        {
          "epoch": 47,
          "updates": 47,
          "samples_seen": 48128,
          "online_loss": 0.09522069990634918,
          "train_loss": 0.09422354400157928,
          "validation_loss": 0.07943321764469147,
          "train_metric": 0.00948985107243061,
          "validation_metric": 0.008000223897397518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6119277730012982
        },
        {
          "epoch": 48,
          "updates": 48,
          "samples_seen": 49152,
          "online_loss": 0.09422355890274048,
          "train_loss": 0.09310679882764816,
          "validation_loss": 0.07838327437639236,
          "train_metric": 0.0093773752450943,
          "validation_metric": 0.007894476875662804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6200395690011646
        },
        {
          "epoch": 49,
          "updates": 49,
          "samples_seen": 50176,
          "online_loss": 0.09310679882764816,
          "train_loss": 0.091671422123909,
          "validation_loss": 0.07705798745155334,
          "train_metric": 0.009232809767127037,
          "validation_metric": 0.007760997861623764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6285048059999099
        },
        {
          "epoch": 50,
          "updates": 50,
          "samples_seen": 51200,
          "online_loss": 0.0916714146733284,
          "train_loss": 0.08987609297037125,
          "validation_loss": 0.07545018196105957,
          "train_metric": 0.009051991626620293,
          "validation_metric": 0.007599066011607647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6366916759998276
        },
        {
          "epoch": 51,
          "updates": 51,
          "samples_seen": 52224,
          "online_loss": 0.08987608551979065,
          "train_loss": 0.08782747387886047,
          "validation_loss": 0.07364872843027115,
          "train_metric": 0.008845660835504532,
          "validation_metric": 0.0074176304042339325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6447101949997887
        },
        {
          "epoch": 52,
          "updates": 52,
          "samples_seen": 53248,
          "online_loss": 0.08782747387886047,
          "train_loss": 0.08572174608707428,
          "validation_loss": 0.071796715259552,
          "train_metric": 0.008633580058813095,
          "validation_metric": 0.007231102325022221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6528578010002093
        },
        {
          "epoch": 53,
          "updates": 53,
          "samples_seen": 54272,
          "online_loss": 0.08572175353765488,
          "train_loss": 0.08376579731702805,
          "validation_loss": 0.07003885507583618,
          "train_metric": 0.008436583913862705,
          "validation_metric": 0.007054056506603956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6606266710005002
        },
        {
          "epoch": 54,
          "updates": 54,
          "samples_seen": 55296,
          "online_loss": 0.08376579731702805,
          "train_loss": 0.08210919797420502,
          "validation_loss": 0.06847812980413437,
          "train_metric": 0.008269738405942917,
          "validation_metric": 0.0068968660198152065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6688181839999743
        },
        {
          "epoch": 55,
          "updates": 55,
          "samples_seen": 56320,
          "online_loss": 0.08210919797420502,
          "train_loss": 0.08080670237541199,
          "validation_loss": 0.06715337932109833,
          "train_metric": 0.00813855417072773,
          "validation_metric": 0.0067634424194693565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6766269590007141
        },
        {
          "epoch": 56,
          "updates": 56,
          "samples_seen": 57344,
          "online_loss": 0.08080669492483139,
          "train_loss": 0.07981666922569275,
          "validation_loss": 0.06604038178920746,
          "train_metric": 0.008038842119276524,
          "validation_metric": 0.006651346106082201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6850182480011426
        },
        {
          "epoch": 57,
          "updates": 57,
          "samples_seen": 58368,
          "online_loss": 0.07981666177511215,
          "train_loss": 0.07902746647596359,
          "validation_loss": 0.06507012993097305,
          "train_metric": 0.007959357462823391,
          "validation_metric": 0.00655362568795681,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6932202340012736
        },
        {
          "epoch": 58,
          "updates": 58,
          "samples_seen": 59392,
          "online_loss": 0.07902747392654419,
          "train_loss": 0.07829805463552475,
          "validation_loss": 0.06415574997663498,
          "train_metric": 0.007885892875492573,
          "validation_metric": 0.006461531855165958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7015932930007693
        },
        {
          "epoch": 59,
          "updates": 59,
          "samples_seen": 60416,
          "online_loss": 0.07829805463552475,
          "train_loss": 0.07749872654676437,
          "validation_loss": 0.06321925669908524,
          "train_metric": 0.007805387023836374,
          "validation_metric": 0.0063672116957604885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7097180030013988
        },
        {
          "epoch": 60,
          "updates": 60,
          "samples_seen": 61440,
          "online_loss": 0.07749871909618378,
          "train_loss": 0.07654216140508652,
          "validation_loss": 0.06221172958612442,
          "train_metric": 0.007709046360105276,
          "validation_metric": 0.006265738047659397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7180731000007654
        },
        {
          "epoch": 61,
          "updates": 61,
          "samples_seen": 62464,
          "online_loss": 0.07654216885566711,
          "train_loss": 0.07539970427751541,
          "validation_loss": 0.06112385541200638,
          "train_metric": 0.007593982852995396,
          "validation_metric": 0.006156171206384897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7261926179999136
        },
        {
          "epoch": 62,
          "updates": 62,
          "samples_seen": 63488,
          "online_loss": 0.07539971172809601,
          "train_loss": 0.07410182803869247,
          "validation_loss": 0.059985652565956116,
          "train_metric": 0.007463264279067516,
          "validation_metric": 0.006041535176336765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7346252930001356
        },
        {
          "epoch": 63,
          "updates": 63,
          "samples_seen": 64512,
          "online_loss": 0.07410182058811188,
          "train_loss": 0.07272505760192871,
          "validation_loss": 0.058856528252363205,
          "train_metric": 0.007324601523578167,
          "validation_metric": 0.00592781463637948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7480847740007448
        },
        {
          "epoch": 64,
          "updates": 64,
          "samples_seen": 65536,
          "online_loss": 0.07272505760192871,
          "train_loss": 0.07136928290128708,
          "validation_loss": 0.05780830234289169,
          "train_metric": 0.007188053335994482,
          "validation_metric": 0.005822240374982357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7582634000009421
        },
        {
          "epoch": 65,
          "updates": 65,
          "samples_seen": 66560,
          "online_loss": 0.07136929035186768,
          "train_loss": 0.07013064622879028,
          "validation_loss": 0.05690465122461319,
          "train_metric": 0.007063302211463451,
          "validation_metric": 0.005731228739023209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.76850504200047
        },
        {
          "epoch": 66,
          "updates": 66,
          "samples_seen": 67584,
          "online_loss": 0.07013064622879028,
          "train_loss": 0.06907621771097183,
          "validation_loss": 0.05618225410580635,
          "train_metric": 0.006957103032618761,
          "validation_metric": 0.005658470559865236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7769063249997998
        },
        {
          "epoch": 67,
          "updates": 67,
          "samples_seen": 68608,
          "online_loss": 0.06907621026039124,
          "train_loss": 0.06822699308395386,
          "validation_loss": 0.05563873052597046,
          "train_metric": 0.006871572230011225,
          "validation_metric": 0.005603729747235775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7851909390010405
        },
        {
          "epoch": 68,
          "updates": 68,
          "samples_seen": 69632,
          "online_loss": 0.06822699308395386,
          "train_loss": 0.06755425035953522,
          "validation_loss": 0.05523163080215454,
          "train_metric": 0.006803817581385374,
          "validation_metric": 0.005562727805227041,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7932160370000929
        },
        {
          "epoch": 69,
          "updates": 69,
          "samples_seen": 70656,
          "online_loss": 0.06755425035953522,
          "train_loss": 0.06699102371931076,
          "validation_loss": 0.054889969527721405,
          "train_metric": 0.006747090723365545,
          "validation_metric": 0.005528316367417574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8025260660015192
        },
        {
          "epoch": 70,
          "updates": 70,
          "samples_seen": 71680,
          "online_loss": 0.06699103116989136,
          "train_loss": 0.06645527482032776,
          "validation_loss": 0.054535090923309326,
          "train_metric": 0.006693131756037474,
          "validation_metric": 0.005492574535310268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8116153080009099
        },
        {
          "epoch": 71,
          "updates": 71,
          "samples_seen": 72704,
          "online_loss": 0.06645527482032776,
          "train_loss": 0.06587681919336319,
          "validation_loss": 0.054103780537843704,
          "train_metric": 0.0066348714753985405,
          "validation_metric": 0.005449134390801191,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8212554570000066
        },
        {
          "epoch": 72,
          "updates": 72,
          "samples_seen": 73728,
          "online_loss": 0.06587681919336319,
          "train_loss": 0.06521832942962646,
          "validation_loss": 0.053565338253974915,
          "train_metric": 0.00656855059787631,
          "validation_metric": 0.005394905339926481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8299702320000506
        },
        {
          "epoch": 73,
          "updates": 73,
          "samples_seen": 74752,
          "online_loss": 0.06521833688020706,
          "train_loss": 0.0644831731915474,
          "validation_loss": 0.05292689800262451,
          "train_metric": 0.006494509056210518,
          "validation_metric": 0.00533060310408473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8383829129998048
        },
        {
          "epoch": 74,
          "updates": 74,
          "samples_seen": 75776,
          "online_loss": 0.0644831657409668,
          "train_loss": 0.06370873004198074,
          "validation_loss": 0.05222616344690323,
          "train_metric": 0.006416509859263897,
          "validation_metric": 0.005260028410702944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8470505889999913
        },
        {
          "epoch": 75,
          "updates": 75,
          "samples_seen": 76800,
          "online_loss": 0.06370873004198074,
          "train_loss": 0.06294945627450943,
          "validation_loss": 0.05151605233550072,
          "train_metric": 0.00634003896266222,
          "validation_metric": 0.0051885079592466354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8557264090013632
        },
        {
          "epoch": 76,
          "updates": 76,
          "samples_seen": 77824,
          "online_loss": 0.06294944882392883,
          "train_loss": 0.06225680932402611,
          "validation_loss": 0.05084725841879845,
          "train_metric": 0.006270278245210648,
          "validation_metric": 0.005121150054037571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8653824160010117
        },
        {
          "epoch": 77,
          "updates": 77,
          "samples_seen": 78848,
          "online_loss": 0.06225680559873581,
          "train_loss": 0.06166328862309456,
          "validation_loss": 0.05025501921772957,
          "train_metric": 0.006210501305758953,
          "validation_metric": 0.005061501637101173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8738821060014743
        },
        {
          "epoch": 78,
          "updates": 78,
          "samples_seen": 79872,
          "online_loss": 0.06166329234838486,
          "train_loss": 0.06117480248212814,
          "validation_loss": 0.049752868711948395,
          "train_metric": 0.006161301862448454,
          "validation_metric": 0.0050109270960092545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8825740170013887
        },
        {
          "epoch": 79,
          "updates": 79,
          "samples_seen": 80896,
          "online_loss": 0.061174798756837845,
          "train_loss": 0.06077233701944351,
          "validation_loss": 0.04933410882949829,
          "train_metric": 0.00612076697871089,
          "validation_metric": 0.004968751221895218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8911624430002121
        },
        {
          "epoch": 80,
          "updates": 80,
          "samples_seen": 81920,
          "online_loss": 0.06077233701944351,
          "train_loss": 0.06042107567191124,
          "validation_loss": 0.04897862672805786,
          "train_metric": 0.006085389759391546,
          "validation_metric": 0.004932948388159275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8997691680015123
        },
        {
          "epoch": 81,
          "updates": 81,
          "samples_seen": 82944,
          "online_loss": 0.06042107939720154,
          "train_loss": 0.06008296459913254,
          "validation_loss": 0.04866224154829979,
          "train_metric": 0.006051335949450731,
          "validation_metric": 0.004901082720607519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.007768995999868
        },
        {
          "epoch": 82,
          "updates": 82,
          "samples_seen": 83968,
          "online_loss": 0.06008296459913254,
          "train_loss": 0.05972862243652344,
          "validation_loss": 0.04836529865860939,
          "train_metric": 0.006015648599714041,
          "validation_metric": 0.00487117562443018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0164767670012225
        },
        {
          "epoch": 83,
          "updates": 83,
          "samples_seen": 84992,
          "online_loss": 0.05972862243652344,
          "train_loss": 0.059344932436943054,
          "validation_loss": 0.04807794466614723,
          "train_metric": 0.005977005232125521,
          "validation_metric": 0.004842234309762716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.024706899001103
        },
        {
          "epoch": 84,
          "updates": 84,
          "samples_seen": 86016,
          "online_loss": 0.059344932436943054,
          "train_loss": 0.05893638730049133,
          "validation_loss": 0.04780101031064987,
          "train_metric": 0.005935857072472572,
          "validation_metric": 0.004814343061298132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0329272340004536
        },
        {
          "epoch": 85,
          "updates": 85,
          "samples_seen": 87040,
          "online_loss": 0.05893639102578163,
          "train_loss": 0.05852045118808746,
          "validation_loss": 0.047542721033096313,
          "train_metric": 0.005893965717405081,
          "validation_metric": 0.004788328427821398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0415019490010309
        },
        {
          "epoch": 86,
          "updates": 86,
          "samples_seen": 88064,
          "online_loss": 0.05852045118808746,
          "train_loss": 0.05811929702758789,
          "validation_loss": 0.04731298238039017,
          "train_metric": 0.0058535621501505375,
          "validation_metric": 0.004765191115438938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0497790220015304
        },
        {
          "epoch": 87,
          "updates": 87,
          "samples_seen": 89088,
          "online_loss": 0.05811929330229759,
          "train_loss": 0.057751115411520004,
          "validation_loss": 0.04711766168475151,
          "train_metric": 0.005816480610519648,
          "validation_metric": 0.004745518788695335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0581643510013237
        },
        {
          "epoch": 88,
          "updates": 88,
          "samples_seen": 90112,
          "online_loss": 0.0577511191368103,
          "train_loss": 0.057424191385507584,
          "validation_loss": 0.046954862773418427,
          "train_metric": 0.005783554166555405,
          "validation_metric": 0.004729121923446655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0664555590010423
        },
        {
          "epoch": 89,
          "updates": 89,
          "samples_seen": 91136,
          "online_loss": 0.057424187660217285,
          "train_loss": 0.05713528394699097,
          "validation_loss": 0.04681447893381119,
          "train_metric": 0.005754456855356693,
          "validation_metric": 0.00471498304978013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0746950810007547
        },
        {
          "epoch": 90,
          "updates": 90,
          "samples_seen": 92160,
          "online_loss": 0.05713528394699097,
          "train_loss": 0.05687222257256508,
          "validation_loss": 0.04668070375919342,
          "train_metric": 0.005727962125092745,
          "validation_metric": 0.004701509140431881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0832059170006687
        },
        {
          "epoch": 91,
          "updates": 91,
          "samples_seen": 93184,
          "online_loss": 0.05687222629785538,
          "train_loss": 0.05661909282207489,
          "validation_loss": 0.04653644934296608,
          "train_metric": 0.005702468100935221,
          "validation_metric": 0.0046869805082678795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0915058750015305
        },
        {
          "epoch": 92,
          "updates": 92,
          "samples_seen": 94208,
          "online_loss": 0.05661910027265549,
          "train_loss": 0.05636191368103027,
          "validation_loss": 0.0463680773973465,
          "train_metric": 0.00567656522616744,
          "validation_metric": 0.0046700225211679935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1005010530006984
        },
        {
          "epoch": 93,
          "updates": 93,
          "samples_seen": 95232,
          "online_loss": 0.056361909955739975,
          "train_loss": 0.05609273910522461,
          "validation_loss": 0.04616893082857132,
          "train_metric": 0.0056494553573429585,
          "validation_metric": 0.004649965558201075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1088705490001303
        },
        {
          "epoch": 94,
          "updates": 94,
          "samples_seen": 96256,
          "online_loss": 0.05609273910522461,
          "train_loss": 0.05581117421388626,
          "validation_loss": 0.0459405891597271,
          "train_metric": 0.0056210970506072044,
          "validation_metric": 0.004626967944204807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1209027920012886
        },
        {
          "epoch": 95,
          "updates": 95,
          "samples_seen": 97280,
          "online_loss": 0.05581117421388626,
          "train_loss": 0.0555228628218174,
          "validation_loss": 0.045691512525081635,
          "train_metric": 0.005592059344053268,
          "validation_metric": 0.004601881839334965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1293737160012824
        },
        {
          "epoch": 96,
          "updates": 96,
          "samples_seen": 98304,
          "online_loss": 0.0555228628218174,
          "train_loss": 0.05523599311709404,
          "validation_loss": 0.04543362185359001,
          "train_metric": 0.0055631669238209724,
          "validation_metric": 0.004575907252728939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1379131070007134
        },
        {
          "epoch": 97,
          "updates": 97,
          "samples_seen": 99328,
          "online_loss": 0.05523598939180374,
          "train_loss": 0.0549573078751564,
          "validation_loss": 0.04517822712659836,
          "train_metric": 0.005535098724067211,
          "validation_metric": 0.004550185985863209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1460870810005872
        },
        {
          "epoch": 98,
          "updates": 98,
          "samples_seen": 100352,
          "online_loss": 0.0549573078751564,
          "train_loss": 0.054689329117536545,
          "validation_loss": 0.04493291303515434,
          "train_metric": 0.005508108995854855,
          "validation_metric": 0.004525478929281235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1549205030005396
        },
        {
          "epoch": 99,
          "updates": 99,
          "samples_seen": 101376,
          "online_loss": 0.05468933284282684,
          "train_loss": 0.05442975461483002,
          "validation_loss": 0.04470039904117584,
          "train_metric": 0.0054819658398628235,
          "validation_metric": 0.004502060357481241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1636736880009266
        },
        {
          "epoch": 100,
          "updates": 100,
          "samples_seen": 102400,
          "online_loss": 0.054429762065410614,
          "train_loss": 0.054173048585653305,
          "validation_loss": 0.044479478150606155,
          "train_metric": 0.005456110928207636,
          "validation_metric": 0.0044798096641898155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1728963420009677
        },
        {
          "epoch": 101,
          "updates": 101,
          "samples_seen": 103424,
          "online_loss": 0.054173048585653305,
          "train_loss": 0.05391303449869156,
          "validation_loss": 0.04426708444952965,
          "train_metric": 0.005429922603070736,
          "validation_metric": 0.00445841858163476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1816272890009714
        },
        {
          "epoch": 102,
          "updates": 102,
          "samples_seen": 104448,
          "online_loss": 0.05391303449869156,
          "train_loss": 0.053645357489585876,
          "validation_loss": 0.04406028985977173,
          "train_metric": 0.005402964539825916,
          "validation_metric": 0.004437590949237347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1899595680006314
        },
        {
          "epoch": 103,
          "updates": 103,
          "samples_seen": 105472,
          "online_loss": 0.05364536494016647,
          "train_loss": 0.05336873233318329,
          "validation_loss": 0.04385726898908615,
          "train_metric": 0.005375103559345007,
          "validation_metric": 0.004417143762111664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1983165850015212
        },
        {
          "epoch": 104,
          "updates": 104,
          "samples_seen": 106496,
          "online_loss": 0.05336873233318329,
          "train_loss": 0.05308463051915169,
          "validation_loss": 0.04365719109773636,
          "train_metric": 0.005346489138901234,
          "validation_metric": 0.004396992269903421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2077893860005133
        },
        {
          "epoch": 105,
          "updates": 105,
          "samples_seen": 107520,
          "online_loss": 0.053084634244441986,
          "train_loss": 0.05279609560966492,
          "validation_loss": 0.04345939680933952,
          "train_metric": 0.0053174300119280815,
          "validation_metric": 0.004377070814371109,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2163997540010314
        },
        {
          "epoch": 106,
          "updates": 106,
          "samples_seen": 108544,
          "online_loss": 0.052796099334955215,
          "train_loss": 0.052506159991025925,
          "validation_loss": 0.04326242208480835,
          "train_metric": 0.005288228392601013,
          "validation_metric": 0.004357232712209225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.224657892000323
        },
        {
          "epoch": 107,
          "updates": 107,
          "samples_seen": 109568,
          "online_loss": 0.05250616744160652,
          "train_loss": 0.052216529846191406,
          "validation_loss": 0.04306358844041824,
          "train_metric": 0.0052590565755963326,
          "validation_metric": 0.004337206482887268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2329125230007776
        },
        {
          "epoch": 108,
          "updates": 108,
          "samples_seen": 110592,
          "online_loss": 0.05221652239561081,
          "train_loss": 0.051927030086517334,
          "validation_loss": 0.0428590327501297,
          "train_metric": 0.005229900125414133,
          "validation_metric": 0.004316604696214199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2412459290007973
        },
        {
          "epoch": 109,
          "updates": 109,
          "samples_seen": 111616,
          "online_loss": 0.051927026361227036,
          "train_loss": 0.05163603648543358,
          "validation_loss": 0.04264454543590546,
          "train_metric": 0.005200591869652271,
          "validation_metric": 0.004295002203434706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2497983060002298
        },
        {
          "epoch": 110,
          "updates": 110,
          "samples_seen": 112640,
          "online_loss": 0.05163602530956268,
          "train_loss": 0.05134138837456703,
          "validation_loss": 0.042416688054800034,
          "train_metric": 0.005170917138457298,
          "validation_metric": 0.004272053483873606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2582667420010694
        },
        {
          "epoch": 111,
          "updates": 111,
          "samples_seen": 113664,
          "online_loss": 0.05134139209985733,
          "train_loss": 0.051041629165410995,
          "validation_loss": 0.042173851281404495,
          "train_metric": 0.005140725988894701,
          "validation_metric": 0.004247595556080341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2679840560012963
        },
        {
          "epoch": 112,
          "updates": 112,
          "samples_seen": 114688,
          "online_loss": 0.051041629165410995,
          "train_loss": 0.05073654651641846,
          "validation_loss": 0.041916802525520325,
          "train_metric": 0.0051099988631904125,
          "validation_metric": 0.004221706185489893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2772760600000765
        },
        {
          "epoch": 113,
          "updates": 113,
          "samples_seen": 115712,
          "online_loss": 0.050736550241708755,
          "train_loss": 0.0504271574318409,
          "validation_loss": 0.04164842143654823,
          "train_metric": 0.005078837741166353,
          "validation_metric": 0.004194676410406828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2856183540006896
        },
        {
          "epoch": 114,
          "updates": 114,
          "samples_seen": 116736,
          "online_loss": 0.0504271537065506,
          "train_loss": 0.05011497065424919,
          "validation_loss": 0.0413728766143322,
          "train_metric": 0.005047395825386047,
          "validation_metric": 0.004166924394667149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.293602344001556
        },
        {
          "epoch": 115,
          "updates": 115,
          "samples_seen": 117760,
          "online_loss": 0.05011496692895889,
          "train_loss": 0.049801211804151535,
          "validation_loss": 0.04109450802206993,
          "train_metric": 0.005015795584768057,
          "validation_metric": 0.004138888791203499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3018585959998745
        },
        {
          "epoch": 116,
          "updates": 116,
          "samples_seen": 118784,
          "online_loss": 0.049801211804151535,
          "train_loss": 0.04948633164167404,
          "validation_loss": 0.040816936641931534,
          "train_metric": 0.004984081722795963,
          "validation_metric": 0.004110931884497404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3101319369998237
        },
        {
          "epoch": 117,
          "updates": 117,
          "samples_seen": 119808,
          "online_loss": 0.04948632791638374,
          "train_loss": 0.04916992783546448,
          "validation_loss": 0.04054240509867668,
          "train_metric": 0.0049522146582603455,
          "validation_metric": 0.00408328277990222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.318318845000249
        },
        {
          "epoch": 118,
          "updates": 118,
          "samples_seen": 120832,
          "online_loss": 0.049169931560754776,
          "train_loss": 0.048851124942302704,
          "validation_loss": 0.040271807461977005,
          "train_metric": 0.004920106381177902,
          "validation_metric": 0.004056029487401247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4263095500009513
        },
        {
          "epoch": 119,
          "updates": 119,
          "samples_seen": 121856,
          "online_loss": 0.048851124942302704,
          "train_loss": 0.04852908477187157,
          "validation_loss": 0.04000496491789818,
          "train_metric": 0.004887671675533056,
          "validation_metric": 0.004029153846204281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.435054831999878
        },
        {
          "epoch": 120,
          "updates": 120,
          "samples_seen": 122880,
          "online_loss": 0.04852908477187157,
          "train_loss": 0.04820341616868973,
          "validation_loss": 0.03974107280373573,
          "train_metric": 0.004854870960116386,
          "validation_metric": 0.004002575296908617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4437116420012899
        },
        {
          "epoch": 121,
          "updates": 121,
          "samples_seen": 123904,
          "online_loss": 0.04820341244339943,
          "train_loss": 0.047874368727207184,
          "validation_loss": 0.039479225873947144,
          "train_metric": 0.0048217312432825565,
          "validation_metric": 0.003976202569901943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4524311430013768
        },
        {
          "epoch": 122,
          "updates": 122,
          "samples_seen": 124928,
          "online_loss": 0.047874368727207184,
          "train_loss": 0.04754269868135452,
          "validation_loss": 0.03921851888298988,
          "train_metric": 0.0047883265651762486,
          "validation_metric": 0.003949945792555809,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4636406760000682
        },
        {
          "epoch": 123,
          "updates": 123,
          "samples_seen": 125952,
          "online_loss": 0.047542694956064224,
          "train_loss": 0.047209229320287704,
          "validation_loss": 0.038958027958869934,
          "train_metric": 0.0047547402791678905,
          "validation_metric": 0.003923709969967604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.472746557001301
        },
        {
          "epoch": 124,
          "updates": 124,
          "samples_seen": 126976,
          "online_loss": 0.047209229320287704,
          "train_loss": 0.04687444865703583,
          "validation_loss": 0.03869657963514328,
          "train_metric": 0.004721023142337799,
          "validation_metric": 0.0038973777554929256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4810383660005755
        },
        {
          "epoch": 125,
          "updates": 125,
          "samples_seen": 128000,
          "online_loss": 0.04687444493174553,
          "train_loss": 0.04653835669159889,
          "validation_loss": 0.03843270614743233,
          "train_metric": 0.004687172826379538,
          "validation_metric": 0.0038708013016730547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4894239780005591
        },
        {
          "epoch": 126,
          "updates": 126,
          "samples_seen": 129024,
          "online_loss": 0.04653835669159889,
          "train_loss": 0.04620053991675377,
          "validation_loss": 0.03816484659910202,
          "train_metric": 0.00465314881876111,
          "validation_metric": 0.0038438239134848118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4976346170005854
        },
        {
          "epoch": 127,
          "updates": 127,
          "samples_seen": 130048,
          "online_loss": 0.04620054364204407,
          "train_loss": 0.045860469341278076,
          "validation_loss": 0.0378916971385479,
          "train_metric": 0.004618898965418339,
          "validation_metric": 0.003816313110291958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.505738714000472
        },
        {
          "epoch": 128,
          "updates": 128,
          "samples_seen": 131072,
          "online_loss": 0.045860469341278076,
          "train_loss": 0.04551783576607704,
          "validation_loss": 0.037612538784742355,
          "train_metric": 0.004584390204399824,
          "validation_metric": 0.003788196947425604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5139813750010944
        },
        {
          "epoch": 129,
          "updates": 129,
          "samples_seen": 132096,
          "online_loss": 0.04551784321665764,
          "train_loss": 0.04517269879579544,
          "validation_loss": 0.03732748329639435,
          "train_metric": 0.004549629054963589,
          "validation_metric": 0.003759487299248576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5229185070002131
        },
        {
          "epoch": 130,
          "updates": 130,
          "samples_seen": 133120,
          "online_loss": 0.04517269879579544,
          "train_loss": 0.04482542723417282,
          "validation_loss": 0.03703742474317551,
          "train_metric": 0.004514652770012617,
          "validation_metric": 0.003730273339897394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5314716910015704
        },
        {
          "epoch": 131,
          "updates": 131,
          "samples_seen": 134144,
          "online_loss": 0.04482542350888252,
          "train_loss": 0.044476527720689774,
          "validation_loss": 0.03674387186765671,
          "train_metric": 0.004479512572288513,
          "validation_metric": 0.003700708970427513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5395289510015573
        },
        {
          "epoch": 132,
          "updates": 132,
          "samples_seen": 135168,
          "online_loss": 0.04447651654481888,
          "train_loss": 0.04412638768553734,
          "validation_loss": 0.03644854202866554,
          "train_metric": 0.004444248043000698,
          "validation_metric": 0.0036709634587168694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5492618230000517
        },
        {
          "epoch": 133,
          "updates": 133,
          "samples_seen": 136192,
          "online_loss": 0.04412638768553734,
          "train_loss": 0.04377516731619835,
          "validation_loss": 0.036152999848127365,
          "train_metric": 0.004408874083310366,
          "validation_metric": 0.0036411979235708714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5578370060011366
        },
        {
          "epoch": 134,
          "updates": 134,
          "samples_seen": 137216,
          "online_loss": 0.04377516731619835,
          "train_loss": 0.04342272877693176,
          "validation_loss": 0.03585836663842201,
          "train_metric": 0.004373378586024046,
          "validation_metric": 0.0036115229595452547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5663335050012392
        },
        {
          "epoch": 135,
          "updates": 135,
          "samples_seen": 138240,
          "online_loss": 0.04342272877693176,
          "train_loss": 0.04306885227560997,
          "validation_loss": 0.03556516021490097,
          "train_metric": 0.004337737336754799,
          "validation_metric": 0.0035819928161799908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5747593750002125
        },
        {
          "epoch": 136,
          "updates": 136,
          "samples_seen": 139264,
          "online_loss": 0.04306885600090027,
          "train_loss": 0.04271339625120163,
          "validation_loss": 0.03527336195111275,
          "train_metric": 0.004301936831325293,
          "validation_metric": 0.0035526035353541374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.583121817000574
        },
        {
          "epoch": 137,
          "updates": 137,
          "samples_seen": 140288,
          "online_loss": 0.04271339252591133,
          "train_loss": 0.04235638305544853,
          "validation_loss": 0.034982506185770035,
          "train_metric": 0.0042659807950258255,
          "validation_metric": 0.003523309947922826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.591461119000087
        },
        {
          "epoch": 138,
          "updates": 138,
          "samples_seen": 141312,
          "online_loss": 0.04235638678073883,
          "train_loss": 0.04199808090925217,
          "validation_loss": 0.034691981971263885,
          "train_metric": 0.00422989297658205,
          "validation_metric": 0.0034940489567816257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5999016170007963
        },
        {
          "epoch": 139,
          "updates": 139,
          "samples_seen": 142336,
          "online_loss": 0.04199808090925217,
          "train_loss": 0.04163879528641701,
          "validation_loss": 0.03440107777714729,
          "train_metric": 0.004193706903606653,
          "validation_metric": 0.0034647504799067974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6083135120006773
        },
        {
          "epoch": 140,
          "updates": 140,
          "samples_seen": 143360,
          "online_loss": 0.04163879528641701,
          "train_loss": 0.04127883166074753,
          "validation_loss": 0.034109245985746384,
          "train_metric": 0.004157452378422022,
          "validation_metric": 0.0034353581722825766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6182835859999614
        },
        {
          "epoch": 141,
          "updates": 141,
          "samples_seen": 144384,
          "online_loss": 0.04127883166074753,
          "train_loss": 0.040918368846178055,
          "validation_loss": 0.03381602466106415,
          "train_metric": 0.004121148493140936,
          "validation_metric": 0.0034058259334415197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6272824420011602
        },
        {
          "epoch": 142,
          "updates": 142,
          "samples_seen": 145408,
          "online_loss": 0.04091837257146835,
          "train_loss": 0.04055747389793396,
          "validation_loss": 0.03352108597755432,
          "train_metric": 0.004084800835698843,
          "validation_metric": 0.0033761209342628717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6356522149999364
        },
        {
          "epoch": 143,
          "updates": 143,
          "samples_seen": 146432,
          "online_loss": 0.04055747762322426,
          "train_loss": 0.04019613564014435,
          "validation_loss": 0.03322424367070198,
          "train_metric": 0.004048408009111881,
          "validation_metric": 0.003346224082633853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6440430500006187
        },
        {
          "epoch": 144,
          "updates": 144,
          "samples_seen": 147456,
          "online_loss": 0.040196143090724945,
          "train_loss": 0.03983437269926071,
          "validation_loss": 0.03292541950941086,
          "train_metric": 0.004011972341686487,
          "validation_metric": 0.003316127462312579,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6525049980009499
        },
        {
          "epoch": 145,
          "updates": 145,
          "samples_seen": 148480,
          "online_loss": 0.03983437269926071,
          "train_loss": 0.03947228565812111,
          "validation_loss": 0.03262472152709961,
          "train_metric": 0.003975504543632269,
          "validation_metric": 0.0032858429476618767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6610636620007426
        },
        {
          "epoch": 146,
          "updates": 146,
          "samples_seen": 149504,
          "online_loss": 0.03947228565812111,
          "train_loss": 0.03911004960536957,
          "validation_loss": 0.03232242166996002,
          "train_metric": 0.003939020447432995,
          "validation_metric": 0.0032553961500525475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6696277130013186
        },
        {
          "epoch": 147,
          "updates": 147,
          "samples_seen": 150528,
          "online_loss": 0.03911004588007927,
          "train_loss": 0.03874783217906952,
          "validation_loss": 0.03201889991760254,
          "train_metric": 0.0039025400765240192,
          "validation_metric": 0.0032248261850327253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6811546520002594
        },
        {
          "epoch": 148,
          "updates": 148,
          "samples_seen": 151552,
          "online_loss": 0.03874783590435982,
          "train_loss": 0.03838580846786499,
          "validation_loss": 0.03171461820602417,
          "train_metric": 0.003866077633574605,
          "validation_metric": 0.0031941800843924284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6941202720008732
        },
        {
          "epoch": 149,
          "updates": 149,
          "samples_seen": 152576,
          "online_loss": 0.03838580846786499,
          "train_loss": 0.03802404925227165,
          "validation_loss": 0.03141006454825401,
          "train_metric": 0.003829642664641142,
          "validation_metric": 0.0031635062769055367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7029791720015055
        },
        {
          "epoch": 150,
          "updates": 150,
          "samples_seen": 153600,
          "online_loss": 0.038024045526981354,
          "train_loss": 0.03766259551048279,
          "validation_loss": 0.031105604022741318,
          "train_metric": 0.0037932386621832848,
          "validation_metric": 0.003132842481136322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7114849549998326
        },
        {
          "epoch": 151,
          "updates": 151,
          "samples_seen": 154624,
          "online_loss": 0.03766259178519249,
          "train_loss": 0.03730147331953049,
          "validation_loss": 0.03080148622393608,
          "train_metric": 0.0037568677216768265,
          "validation_metric": 0.003102213144302368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.720186773000023
        },
        {
          "epoch": 152,
          "updates": 152,
          "samples_seen": 155648,
          "online_loss": 0.03730147331953049,
          "train_loss": 0.036940742284059525,
          "validation_loss": 0.030497752130031586,
          "train_metric": 0.0037205363623797894,
          "validation_metric": 0.0030716219916939735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7288653280011204
        },
        {
          "epoch": 153,
          "updates": 153,
          "samples_seen": 156672,
          "online_loss": 0.036940742284059525,
          "train_loss": 0.036580491811037064,
          "validation_loss": 0.030194329097867012,
          "train_metric": 0.0036842534318566322,
          "validation_metric": 0.003041062271222472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7377648150013556
        },
        {
          "epoch": 154,
          "updates": 154,
          "samples_seen": 157696,
          "online_loss": 0.03658049553632736,
          "train_loss": 0.036220844835042953,
          "validation_loss": 0.02989107184112072,
          "train_metric": 0.003648030571639538,
          "validation_metric": 0.003010519314557314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.746730564000245
        },
        {
          "epoch": 155,
          "updates": 155,
          "samples_seen": 158720,
          "online_loss": 0.036220844835042953,
          "train_loss": 0.03586187958717346,
          "validation_loss": 0.02958785928785801,
          "train_metric": 0.003611877327784896,
          "validation_metric": 0.002979980781674385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7558474090001255
        },
        {
          "epoch": 156,
          "updates": 156,
          "samples_seen": 159744,
          "online_loss": 0.03586187958717346,
          "train_loss": 0.035503678023815155,
          "validation_loss": 0.029284661635756493,
          "train_metric": 0.0035758004523813725,
          "validation_metric": 0.002949443878605962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8673841430008906
        },
        {
          "epoch": 157,
          "updates": 157,
          "samples_seen": 160768,
          "online_loss": 0.035503678023815155,
          "train_loss": 0.03514626622200012,
          "validation_loss": 0.028981536626815796,
          "train_metric": 0.0035398032050579786,
          "validation_metric": 0.002918914193287492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8765597210003762
        },
        {
          "epoch": 158,
          "updates": 158,
          "samples_seen": 161792,
          "online_loss": 0.03514626622200012,
          "train_loss": 0.03478968143463135,
          "validation_loss": 0.02867860347032547,
          "train_metric": 0.003503889311105013,
          "validation_metric": 0.0028884040657430887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8855528370004322
        },
        {
          "epoch": 159,
          "updates": 159,
          "samples_seen": 162816,
          "online_loss": 0.03478968143463135,
          "train_loss": 0.034433964639902115,
          "validation_loss": 0.02837599813938141,
          "train_metric": 0.003468062961474061,
          "validation_metric": 0.002857926534488797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8939233550008794
        },
        {
          "epoch": 160,
          "updates": 160,
          "samples_seen": 163840,
          "online_loss": 0.034433964639902115,
          "train_loss": 0.0340791717171669,
          "validation_loss": 0.028073862195014954,
          "train_metric": 0.0034323290456086397,
          "validation_metric": 0.0028274967335164547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9024848650005879
        },
        {
          "epoch": 161,
          "updates": 161,
          "samples_seen": 164864,
          "online_loss": 0.0340791717171669,
          "train_loss": 0.03372538834810257,
          "validation_loss": 0.02777230367064476,
          "train_metric": 0.003396698273718357,
          "validation_metric": 0.002797124907374382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9111475300014718
        },
        {
          "epoch": 162,
          "updates": 162,
          "samples_seen": 165888,
          "online_loss": 0.03372538834810257,
          "train_loss": 0.03337269648909569,
          "validation_loss": 0.027471434324979782,
          "train_metric": 0.0033611757680773735,
          "validation_metric": 0.002766821999102831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9195508820012037
        },
        {
          "epoch": 163,
          "updates": 163,
          "samples_seen": 166912,
          "online_loss": 0.03337269276380539,
          "train_loss": 0.03302113711833954,
          "validation_loss": 0.027171334251761436,
          "train_metric": 0.0033257680479437113,
          "validation_metric": 0.0027365973219275475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9288863390011102
        },
        {
          "epoch": 164,
          "updates": 164,
          "samples_seen": 167936,
          "online_loss": 0.03302113711833954,
          "train_loss": 0.032670751214027405,
          "validation_loss": 0.02687207981944084,
          "train_metric": 0.0032904786057770252,
          "validation_metric": 0.002706457395106554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9375249510012509
        },
        {
          "epoch": 165,
          "updates": 165,
          "samples_seen": 168960,
          "online_loss": 0.032670751214027405,
          "train_loss": 0.03232157975435257,
          "validation_loss": 0.02657373994588852,
          "train_metric": 0.003255311166867614,
          "validation_metric": 0.002676410134881735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.946598365000682
        },
        {
          "epoch": 166,
          "updates": 166,
          "samples_seen": 169984,
          "online_loss": 0.03232157975435257,
          "train_loss": 0.03197367489337921,
          "validation_loss": 0.02627638727426529,
          "train_metric": 0.003220271784812212,
          "validation_metric": 0.0026464613620191813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9551744330001384
        },
        {
          "epoch": 167,
          "updates": 167,
          "samples_seen": 171008,
          "online_loss": 0.03197367116808891,
          "train_loss": 0.031627099961042404,
          "validation_loss": 0.025980109348893166,
          "train_metric": 0.00318536558188498,
          "validation_metric": 0.0026166215538978577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9640121870015719
        },
        {
          "epoch": 168,
          "updates": 168,
          "samples_seen": 172032,
          "online_loss": 0.0316271036863327,
          "train_loss": 0.03128194436430931,
          "validation_loss": 0.025685034692287445,
          "train_metric": 0.003150603035464883,
          "validation_metric": 0.002586902119219303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9729769620007573
        },
        {
          "epoch": 169,
          "updates": 169,
          "samples_seen": 173056,
          "online_loss": 0.03128194436430931,
          "train_loss": 0.030938295647501945,
          "validation_loss": 0.025391343981027603,
          "train_metric": 0.003115992061793804,
          "validation_metric": 0.002557323081418872,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9818778970002313
        },
        {
          "epoch": 170,
          "updates": 170,
          "samples_seen": 174080,
          "online_loss": 0.030938293784856796,
          "train_loss": 0.03059622459113598,
          "validation_loss": 0.02509925700724125,
          "train_metric": 0.00308153941296041,
          "validation_metric": 0.002527905162423849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9912193500003923
        },
        {
          "epoch": 171,
          "updates": 171,
          "samples_seen": 175104,
          "online_loss": 0.03059622459113598,
          "train_loss": 0.030255800113081932,
          "validation_loss": 0.02480899728834629,
          "train_metric": 0.0030472532380372286,
          "validation_metric": 0.002498671179637313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.000136761000249
        },
        {
          "epoch": 172,
          "updates": 172,
          "samples_seen": 176128,
          "online_loss": 0.03025580197572708,
          "train_loss": 0.02991708554327488,
          "validation_loss": 0.024520769715309143,
          "train_metric": 0.003013139357790351,
          "validation_metric": 0.002469642087817192,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0094070590002957
        },
        {
          "epoch": 173,
          "updates": 173,
          "samples_seen": 177152,
          "online_loss": 0.02991708368062973,
          "train_loss": 0.02958015352487564,
          "validation_loss": 0.024234741926193237,
          "train_metric": 0.0029792047571390867,
          "validation_metric": 0.0024408346507698298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0182684840001457
        },
        {
          "epoch": 174,
          "updates": 174,
          "samples_seen": 178176,
          "online_loss": 0.02958015352487564,
          "train_loss": 0.029245084151625633,
          "validation_loss": 0.02395101636648178,
          "train_metric": 0.0029454578179866076,
          "validation_metric": 0.0024122591130435467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.026648020000721
        },
        {
          "epoch": 175,
          "updates": 175,
          "samples_seen": 179200,
          "online_loss": 0.029245084151625633,
          "train_loss": 0.028911957517266273,
          "validation_loss": 0.023669691756367683,
          "train_metric": 0.0029119066894054413,
          "validation_metric": 0.0023839245550334454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0357221920003212
        },
        {
          "epoch": 176,
          "updates": 176,
          "samples_seen": 180224,
          "online_loss": 0.028911957517266273,
          "train_loss": 0.028580863028764725,
          "validation_loss": 0.023390810936689377,
          "train_metric": 0.0028785597532987595,
          "validation_metric": 0.002355836797505617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.044821565001257
        },
        {
          "epoch": 177,
          "updates": 177,
          "samples_seen": 181248,
          "online_loss": 0.028580859303474426,
          "train_loss": 0.028251860290765762,
          "validation_loss": 0.023114414885640144,
          "train_metric": 0.0028454242274165154,
          "validation_metric": 0.002327999332919717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0532674599999154
        },
        {
          "epoch": 178,
          "updates": 178,
          "samples_seen": 182272,
          "online_loss": 0.02825186215341091,
          "train_loss": 0.027925048023462296,
          "validation_loss": 0.02284049242734909,
          "train_metric": 0.002812508726492524,
          "validation_metric": 0.0023004107642918825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0616682859999855
        },
        {
          "epoch": 179,
          "updates": 179,
          "samples_seen": 183296,
          "online_loss": 0.027925049886107445,
          "train_loss": 0.027600480243563652,
          "validation_loss": 0.022568993270397186,
          "train_metric": 0.002779819071292877,
          "validation_metric": 0.0022730662021785975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.069871512001555
        },
        {
          "epoch": 180,
          "updates": 180,
          "samples_seen": 184320,
          "online_loss": 0.027600478380918503,
          "train_loss": 0.027278238907456398,
          "validation_loss": 0.022299926728010178,
          "train_metric": 0.0027473641093820333,
          "validation_metric": 0.0022459670435637236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.078511988000173
        },
        {
          "epoch": 181,
          "updates": 181,
          "samples_seen": 185344,
          "online_loss": 0.02727823704481125,
          "train_loss": 0.0269584059715271,
          "validation_loss": 0.022033311426639557,
          "train_metric": 0.0027151519898325205,
          "validation_metric": 0.002219114452600479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0868980049999664
        },
        {
          "epoch": 182,
          "updates": 182,
          "samples_seen": 186368,
          "online_loss": 0.0269584059715271,
          "train_loss": 0.026641061529517174,
          "validation_loss": 0.021769290789961815,
          "train_metric": 0.0026831901632249355,
          "validation_metric": 0.002192523330450058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.095518672000253
        },
        {
          "epoch": 183,
          "updates": 183,
          "samples_seen": 187392,
          "online_loss": 0.026641061529517174,
          "train_loss": 0.026326287537813187,
          "validation_loss": 0.021508030593395233,
          "train_metric": 0.0026514872442930937,
          "validation_metric": 0.002166209975257516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.103840110999954
        },
        {
          "epoch": 184,
          "updates": 184,
          "samples_seen": 188416,
          "online_loss": 0.026326289400458336,
          "train_loss": 0.026014156639575958,
          "validation_loss": 0.02124970033764839,
          "train_metric": 0.0026200509164482355,
          "validation_metric": 0.0021401920821517706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.112398977000339
        },
        {
          "epoch": 185,
          "updates": 185,
          "samples_seen": 189440,
          "online_loss": 0.026014156639575958,
          "train_loss": 0.025704752653837204,
          "validation_loss": 0.02099444344639778,
          "train_metric": 0.0025888881646096706,
          "validation_metric": 0.002114483155310154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1211581170009595
        },
        {
          "epoch": 186,
          "updates": 186,
          "samples_seen": 190464,
          "online_loss": 0.025704750791192055,
          "train_loss": 0.025398142635822296,
          "validation_loss": 0.020742319524288177,
          "train_metric": 0.002558007836341858,
          "validation_metric": 0.002089090645313263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1294834790005552
        },
        {
          "epoch": 187,
          "updates": 187,
          "samples_seen": 191488,
          "online_loss": 0.025398142635822296,
          "train_loss": 0.025094393640756607,
          "validation_loss": 0.020493362098932266,
          "train_metric": 0.002527415519580245,
          "validation_metric": 0.0020640164148062468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1397633430005953
        },
        {
          "epoch": 188,
          "updates": 188,
          "samples_seen": 192512,
          "online_loss": 0.025094393640756607,
          "train_loss": 0.02479359321296215,
          "validation_loss": 0.020247574895620346,
          "train_metric": 0.002497119829058647,
          "validation_metric": 0.0020392616279423237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.150160909000988
        },
        {
          "epoch": 189,
          "updates": 189,
          "samples_seen": 193536,
          "online_loss": 0.0247935950756073,
          "train_loss": 0.024495810270309448,
          "validation_loss": 0.020004991441965103,
          "train_metric": 0.0024671282153576612,
          "validation_metric": 0.0020148293115198612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1586422550008137
        },
        {
          "epoch": 190,
          "updates": 190,
          "samples_seen": 194560,
          "online_loss": 0.024495810270309448,
          "train_loss": 0.02420111373066902,
          "validation_loss": 0.019765622913837433,
          "train_metric": 0.0024374471977353096,
          "validation_metric": 0.001990721095353365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.167650058001527
        },
        {
          "epoch": 191,
          "updates": 191,
          "samples_seen": 195584,
          "online_loss": 0.02420111373066902,
          "train_loss": 0.023909570649266243,
          "validation_loss": 0.019529428333044052,
          "train_metric": 0.002408084226772189,
          "validation_metric": 0.0019669325556606054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.176231280000138
        },
        {
          "epoch": 192,
          "updates": 192,
          "samples_seen": 196608,
          "online_loss": 0.023909566923975945,
          "train_loss": 0.023621244356036186,
          "validation_loss": 0.019296351820230484,
          "train_metric": 0.002379045356065035,
          "validation_metric": 0.0019434577552601695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.185389482001483
        },
        {
          "epoch": 193,
          "updates": 193,
          "samples_seen": 197632,
          "online_loss": 0.023621242493391037,
          "train_loss": 0.023336203768849373,
          "validation_loss": 0.019066333770751953,
          "train_metric": 0.0023503368720412254,
          "validation_metric": 0.0019202912226319313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.293298181000864
        },
        {
          "epoch": 194,
          "updates": 194,
          "samples_seen": 198656,
          "online_loss": 0.023336203768849373,
          "train_loss": 0.02305450849235058,
          "validation_loss": 0.018839357420802116,
          "train_metric": 0.002321965526789427,
          "validation_metric": 0.001897431444376707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3023988570002984
        },
        {
          "epoch": 195,
          "updates": 195,
          "samples_seen": 199680,
          "online_loss": 0.02305450662970543,
          "train_loss": 0.022776229307055473,
          "validation_loss": 0.018615521490573883,
          "train_metric": 0.0022939383052289486,
          "validation_metric": 0.0018748871516436338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3103622819999146
        },
        {
          "epoch": 196,
          "updates": 196,
          "samples_seen": 200704,
          "online_loss": 0.022776229307055473,
          "train_loss": 0.022501425817608833,
          "validation_loss": 0.01839490979909897,
          "train_metric": 0.002266261028125882,
          "validation_metric": 0.0018526678904891014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3190971330004686
        },
        {
          "epoch": 197,
          "updates": 197,
          "samples_seen": 201728,
          "online_loss": 0.022501423954963684,
          "train_loss": 0.02223014086484909,
          "validation_loss": 0.01817762851715088,
          "train_metric": 0.002238938584923744,
          "validation_metric": 0.0018307841382920742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.327445783999792
        },
        {
          "epoch": 198,
          "updates": 198,
          "samples_seen": 202752,
          "online_loss": 0.02223013900220394,
          "train_loss": 0.021962448954582214,
          "validation_loss": 0.0179637148976326,
          "train_metric": 0.0022119772620499134,
          "validation_metric": 0.0018092396203428507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.336030448001111
        },
        {
          "epoch": 199,
          "updates": 199,
          "samples_seen": 203776,
          "online_loss": 0.021962447091937065,
          "train_loss": 0.021698372438549995,
          "validation_loss": 0.017753133550286293,
          "train_metric": 0.0021853807847946882,
          "validation_metric": 0.0017880306113511324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.345021724000617
        },
        {
          "epoch": 200,
          "updates": 200,
          "samples_seen": 204800,
          "online_loss": 0.021698370575904846,
          "train_loss": 0.021437974646687508,
          "validation_loss": 0.017545830458402634,
          "train_metric": 0.002159154275432229,
          "validation_metric": 0.0017671518726274371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3540422130008665
        },
        {
          "epoch": 201,
          "updates": 201,
          "samples_seen": 205824,
          "online_loss": 0.02143797278404236,
          "train_loss": 0.02118128538131714,
          "validation_loss": 0.01734180375933647,
          "train_metric": 0.0021333012264221907,
          "validation_metric": 0.0017466029385104775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3634335750011815
        },
        {
          "epoch": 202,
          "updates": 202,
          "samples_seen": 206848,
          "online_loss": 0.02118128351867199,
          "train_loss": 0.02092834748327732,
          "validation_loss": 0.01714104227721691,
          "train_metric": 0.0021078267600387335,
          "validation_metric": 0.0017263829940930009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.37192872200103
        },
        {
          "epoch": 203,
          "updates": 203,
          "samples_seen": 207872,
          "online_loss": 0.02092834934592247,
          "train_loss": 0.020679185166954994,
          "validation_loss": 0.01694352924823761,
          "train_metric": 0.002082731807604432,
          "validation_metric": 0.001706490060314536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.380814313000883
        },
        {
          "epoch": 204,
          "updates": 204,
          "samples_seen": 208896,
          "online_loss": 0.020679183304309845,
          "train_loss": 0.020433826372027397,
          "validation_loss": 0.016749193891882896,
          "train_metric": 0.0020580203272402287,
          "validation_metric": 0.001686917501501739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.389163229001497
        },
        {
          "epoch": 205,
          "updates": 205,
          "samples_seen": 209920,
          "online_loss": 0.020433828234672546,
          "train_loss": 0.020192300900816917,
          "validation_loss": 0.016557956114411354,
          "train_metric": 0.0020336946472525597,
          "validation_metric": 0.0016676567029207945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3981452780008112
        },
        {
          "epoch": 206,
          "updates": 206,
          "samples_seen": 210944,
          "online_loss": 0.020192302763462067,
          "train_loss": 0.019954625517129898,
          "validation_loss": 0.016369763761758804,
          "train_metric": 0.002009756863117218,
          "validation_metric": 0.0016487027751281857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4063144360006845
        },
        {
          "epoch": 207,
          "updates": 207,
          "samples_seen": 211968,
          "online_loss": 0.019954625517129898,
          "train_loss": 0.01972080022096634,
          "validation_loss": 0.016184642910957336,
          "train_metric": 0.00198620674200356,
          "validation_metric": 0.0016300580464303493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.414719074000459
        },
        {
          "epoch": 208,
          "updates": 208,
          "samples_seen": 212992,
          "online_loss": 0.01972080022096634,
          "train_loss": 0.019490854814648628,
          "validation_loss": 0.01600264571607113,
          "train_metric": 0.001963047543540597,
          "validation_metric": 0.0016117278719320893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.423248570999931
        },
        {
          "epoch": 209,
          "updates": 209,
          "samples_seen": 214016,
          "online_loss": 0.019490854814648628,
          "train_loss": 0.019264785572886467,
          "validation_loss": 0.015823811292648315,
          "train_metric": 0.0019402786856517196,
          "validation_metric": 0.0015937162097543478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.432188439001038
        },
        {
          "epoch": 210,
          "updates": 210,
          "samples_seen": 215040,
          "online_loss": 0.019264785572886467,
          "train_loss": 0.019042592495679855,
          "validation_loss": 0.015648137778043747,
          "train_metric": 0.001917900051921606,
          "validation_metric": 0.0015760231763124466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4408681050008454
        },
        {
          "epoch": 211,
          "updates": 211,
          "samples_seen": 216064,
          "online_loss": 0.019042592495679855,
          "train_loss": 0.018824277445673943,
          "validation_loss": 0.015475531108677387,
          "train_metric": 0.0018959123408421874,
          "validation_metric": 0.0015586388763040304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.450276555000528
        },
        {
          "epoch": 212,
          "updates": 212,
          "samples_seen": 217088,
          "online_loss": 0.018824279308319092,
          "train_loss": 0.018609829246997833,
          "validation_loss": 0.01530594751238823,
          "train_metric": 0.0018743140390142798,
          "validation_metric": 0.0015415591187775135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4587446420009655
        },
        {
          "epoch": 213,
          "updates": 213,
          "samples_seen": 218112,
          "online_loss": 0.018609829246997833,
          "train_loss": 0.018399236723780632,
          "validation_loss": 0.015139342285692692,
          "train_metric": 0.0018531037494540215,
          "validation_metric": 0.001524779130704701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.467276102001051
        },
        {
          "epoch": 214,
          "updates": 214,
          "samples_seen": 219136,
          "online_loss": 0.01839923858642578,
          "train_loss": 0.018192484974861145,
          "validation_loss": 0.014975660480558872,
          "train_metric": 0.0018322805408388376,
          "validation_metric": 0.0015082936733961105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.475611645000754
        },
        {
          "epoch": 215,
          "updates": 215,
          "samples_seen": 220160,
          "online_loss": 0.018192484974861145,
          "train_loss": 0.01798955351114273,
          "validation_loss": 0.014814858324825764,
          "train_metric": 0.0018118417356163263,
          "validation_metric": 0.0014920983230695128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.48452677700152
        },
        {
          "epoch": 216,
          "updates": 216,
          "samples_seen": 221184,
          "online_loss": 0.01798955351114273,
          "train_loss": 0.01779041439294815,
          "validation_loss": 0.014656833373010159,
          "train_metric": 0.0017917855875566602,
          "validation_metric": 0.0014761826023459435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4945959169999696
        },
        {
          "epoch": 217,
          "updates": 217,
          "samples_seen": 222208,
          "online_loss": 0.0177904162555933,
          "train_loss": 0.01759505085647106,
          "validation_loss": 0.014501494355499744,
          "train_metric": 0.0017721091862767935,
          "validation_metric": 0.0014605375472456217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5037856700000702
        },
        {
          "epoch": 218,
          "updates": 218,
          "samples_seen": 223232,
          "online_loss": 0.01759505271911621,
          "train_loss": 0.01740342564880848,
          "validation_loss": 0.014348805882036686,
          "train_metric": 0.001752809272147715,
          "validation_metric": 0.0014451593160629272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5125538150005013
        },
        {
          "epoch": 219,
          "updates": 219,
          "samples_seen": 224256,
          "online_loss": 0.01740342564880848,
          "train_loss": 0.01721549779176712,
          "validation_loss": 0.014198750257492065,
          "train_metric": 0.001733881887048483,
          "validation_metric": 0.0014300461625680327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5226997839999967
        },
        {
          "epoch": 220,
          "updates": 220,
          "samples_seen": 225280,
          "online_loss": 0.01721549779176712,
          "train_loss": 0.0170312337577343,
          "validation_loss": 0.014051360078155994,
          "train_metric": 0.0017153233056887984,
          "validation_metric": 0.001415201579220593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5373631900001783
        },
        {
          "epoch": 221,
          "updates": 221,
          "samples_seen": 226304,
          "online_loss": 0.01703123189508915,
          "train_loss": 0.01685059256851673,
          "validation_loss": 0.013906624168157578,
          "train_metric": 0.001697129919193685,
          "validation_metric": 0.0014006244018673897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.547582715000317
        },
        {
          "epoch": 222,
          "updates": 222,
          "samples_seen": 227328,
          "online_loss": 0.016850590705871582,
          "train_loss": 0.016673535108566284,
          "validation_loss": 0.013764486648142338,
          "train_metric": 0.0016792971873655915,
          "validation_metric": 0.0013863085769116879,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5597954400000162
        },
        {
          "epoch": 223,
          "updates": 223,
          "samples_seen": 228352,
          "online_loss": 0.016673533245921135,
          "train_loss": 0.016500014811754227,
          "validation_loss": 0.013624902814626694,
          "train_metric": 0.001661820919252932,
          "validation_metric": 0.001372250379063189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.57324552800128
        },
        {
          "epoch": 224,
          "updates": 224,
          "samples_seen": 229376,
          "online_loss": 0.016500011086463928,
          "train_loss": 0.016329960897564888,
          "validation_loss": 0.01348782517015934,
          "train_metric": 0.0016446937806904316,
          "validation_metric": 0.0013584443368017673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.585779397000806
        },
        {
          "epoch": 225,
          "updates": 225,
          "samples_seen": 230400,
          "online_loss": 0.016329960897564888,
          "train_loss": 0.01616334542632103,
          "validation_loss": 0.013353202491998672,
          "train_metric": 0.0016279129777103662,
          "validation_metric": 0.001344885560683906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5943440120008745
        },
        {
          "epoch": 226,
          "updates": 226,
          "samples_seen": 231424,
          "online_loss": 0.01616334542632103,
          "train_loss": 0.01600009761750698,
          "validation_loss": 0.013221010565757751,
          "train_metric": 0.0016114714089781046,
          "validation_metric": 0.0013315719552338123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.611295560000144
        },
        {
          "epoch": 227,
          "updates": 227,
          "samples_seen": 232448,
          "online_loss": 0.01600009761750698,
          "train_loss": 0.01584017090499401,
          "validation_loss": 0.013091150671243668,
          "train_metric": 0.001595364068634808,
          "validation_metric": 0.0013184929266571999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6252488430000085
        },
        {
          "epoch": 228,
          "updates": 228,
          "samples_seen": 233472,
          "online_loss": 0.01584016904234886,
          "train_loss": 0.015683487057685852,
          "validation_loss": 0.012963546440005302,
          "train_metric": 0.001579583389684558,
          "validation_metric": 0.0013056410243734717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6396607770002447
        },
        {
          "epoch": 229,
          "updates": 229,
          "samples_seen": 234496,
          "online_loss": 0.015683488920331,
          "train_loss": 0.015530006028711796,
          "validation_loss": 0.012838119640946388,
          "train_metric": 0.0015641250647604465,
          "validation_metric": 0.001293008215725422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6531741010003316
        },
        {
          "epoch": 230,
          "updates": 230,
          "samples_seen": 235520,
          "online_loss": 0.015530005097389221,
          "train_loss": 0.015379653312265873,
          "validation_loss": 0.012714877724647522,
          "train_metric": 0.0015489824581891298,
          "validation_metric": 0.0012805958976969123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6628175620007823
        },
        {
          "epoch": 231,
          "updates": 231,
          "samples_seen": 236544,
          "online_loss": 0.015379655174911022,
          "train_loss": 0.015232358127832413,
          "validation_loss": 0.01259379182010889,
          "train_metric": 0.001534147304482758,
          "validation_metric": 0.001268400577828288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6765792920014064
        },
        {
          "epoch": 232,
          "updates": 232,
          "samples_seen": 237568,
          "online_loss": 0.015232358127832413,
          "train_loss": 0.015088064596056938,
          "validation_loss": 0.012474863789975643,
          "train_metric": 0.0015196145977824926,
          "validation_metric": 0.0012564227217808366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6890682820012444
        },
        {
          "epoch": 233,
          "updates": 233,
          "samples_seen": 238592,
          "online_loss": 0.015088065527379513,
          "train_loss": 0.014946703799068928,
          "validation_loss": 0.012358047068119049,
          "train_metric": 0.0015053772367537022,
          "validation_metric": 0.0012446572072803974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8209043060014665
        },
        {
          "epoch": 234,
          "updates": 234,
          "samples_seen": 239616,
          "online_loss": 0.014946704730391502,
          "train_loss": 0.01480820495635271,
          "validation_loss": 0.012243317440152168,
          "train_metric": 0.0014914283528923988,
          "validation_metric": 0.0012331019388511777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8298808910003572
        },
        {
          "epoch": 235,
          "updates": 235,
          "samples_seen": 240640,
          "online_loss": 0.014808205887675285,
          "train_loss": 0.014672495424747467,
          "validation_loss": 0.012130625545978546,
          "train_metric": 0.0014777600299566984,
          "validation_metric": 0.0012217521434649825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8384806030007894
        },
        {
          "epoch": 236,
          "updates": 236,
          "samples_seen": 241664,
          "online_loss": 0.014672495424747467,
          "train_loss": 0.0145395053550601,
          "validation_loss": 0.012019959278404713,
          "train_metric": 0.0014643658651039004,
          "validation_metric": 0.0012106063077226281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8476369400013937
        },
        {
          "epoch": 237,
          "updates": 237,
          "samples_seen": 242688,
          "online_loss": 0.014539506286382675,
          "train_loss": 0.014409175142645836,
          "validation_loss": 0.011911289766430855,
          "train_metric": 0.0014512393390759826,
          "validation_metric": 0.001199661404825747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8560184850011865
        },
        {
          "epoch": 238,
          "updates": 238,
          "samples_seen": 243712,
          "online_loss": 0.014409175142645836,
          "train_loss": 0.014281423762440681,
          "validation_loss": 0.011804533191025257,
          "train_metric": 0.0014383726520463824,
          "validation_metric": 0.0011889094021171331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.864138050001202
        },
        {
          "epoch": 239,
          "updates": 239,
          "samples_seen": 244736,
          "online_loss": 0.014281424693763256,
          "train_loss": 0.014156187884509563,
          "validation_loss": 0.011699625290930271,
          "train_metric": 0.0014257594011723995,
          "validation_metric": 0.0011783433146774769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.873436859001231
        },
        {
          "epoch": 240,
          "updates": 240,
          "samples_seen": 245760,
          "online_loss": 0.014156188815832138,
          "train_loss": 0.01403339859098196,
          "validation_loss": 0.011596514843404293,
          "train_metric": 0.0014133926015347242,
          "validation_metric": 0.0011679584858939052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8820452180007123
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 240
      },
      "best_validation_metric": 0.0011679584858939052,
      "config": {
        "label": "Full batch",
        "batch": 1024,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.03417540761854795
    },
    "6a3fa29d2adc_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.5785819953307509,
          "train_loss": 0.24066151678562164,
          "validation_loss": 0.15968576073646545,
          "train_metric": 0.024238549172878265,
          "validation_metric": 0.016082964837551117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.015164885999183753
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.20281902188435197,
          "train_loss": 0.14825578033924103,
          "validation_loss": 0.12119951099157333,
          "train_metric": 0.014931779354810715,
          "validation_metric": 0.012206770479679108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.028110934999858728
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13218980515375733,
          "train_loss": 0.11559955775737762,
          "validation_loss": 0.09988148510456085,
          "train_metric": 0.011642763391137123,
          "validation_metric": 0.010059698484838009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04075867299980018
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10902172443456948,
          "train_loss": 0.09608157724142075,
          "validation_loss": 0.08105382323265076,
          "train_metric": 0.009676986373960972,
          "validation_metric": 0.008163443766534328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05330574099934893
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08852994372136891,
          "train_loss": 0.07870391011238098,
          "validation_loss": 0.06390558928251266,
          "train_metric": 0.00792676955461502,
          "validation_metric": 0.0064363377168774605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06571307299964246
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.0738776980433613,
          "train_loss": 0.06744672358036041,
          "validation_loss": 0.05580252408981323,
          "train_metric": 0.006792987231165171,
          "validation_metric": 0.005620225332677364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07830333499987319
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.0662799112033099,
          "train_loss": 0.062105804681777954,
          "validation_loss": 0.051005952060222626,
          "train_metric": 0.006255070213228464,
          "validation_metric": 0.0051371329464018345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09047545499925036
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06061952211894095,
          "train_loss": 0.058522891253232956,
          "validation_loss": 0.04748249426484108,
          "train_metric": 0.005894211120903492,
          "validation_metric": 0.0047822631895542145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10299753899926145
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0586743209278211,
          "train_loss": 0.05518710985779762,
          "validation_loss": 0.04696771502494812,
          "train_metric": 0.00555824302136898,
          "validation_metric": 0.0047304159961640835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20262390700008837
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05452180642168969,
          "train_loss": 0.05012526735663414,
          "validation_loss": 0.04216490313410759,
          "train_metric": 0.005048433784395456,
          "validation_metric": 0.004246694967150688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21580182499928924
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0489687561057508,
          "train_loss": 0.0455508716404438,
          "validation_loss": 0.0387922078371048,
          "train_metric": 0.004587717354297638,
          "validation_metric": 0.003907009027898312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22936687199944572
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.045328833162784576,
          "train_loss": 0.04298621788620949,
          "validation_loss": 0.03603599593043327,
          "train_metric": 0.004329414572566748,
          "validation_metric": 0.0036294138990342617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2422471730005782
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.040962925646454096,
          "train_loss": 0.03840823099017143,
          "validation_loss": 0.03274441510438919,
          "train_metric": 0.003868336323648691,
          "validation_metric": 0.003297897521406412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2540327829992748
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.035591799300163984,
          "train_loss": 0.03237031772732735,
          "validation_loss": 0.027760149911046028,
          "train_metric": 0.003260219469666481,
          "validation_metric": 0.0027959002181887627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2666977129993029
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.034175259876064956,
          "train_loss": 0.030200820416212082,
          "validation_loss": 0.02461184561252594,
          "train_metric": 0.0030417158268392086,
          "validation_metric": 0.0024788149166852236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27904925899929367
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.029044358525425196,
          "train_loss": 0.02824326977133751,
          "validation_loss": 0.02484940178692341,
          "train_metric": 0.0028445585630834103,
          "validation_metric": 0.0025027410592883825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2914433210007701
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.027968548878561705,
          "train_loss": 0.026699528098106384,
          "validation_loss": 0.022850092500448227,
          "train_metric": 0.0026890786830335855,
          "validation_metric": 0.0023013774771243334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3042013649992441
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029929061303846538,
          "train_loss": 0.028845902532339096,
          "validation_loss": 0.023522071540355682,
          "train_metric": 0.002905253553763032,
          "validation_metric": 0.002369056921452284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.317609325000376
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.0253515514486935,
          "train_loss": 0.022620845586061478,
          "validation_loss": 0.019433408975601196,
          "train_metric": 0.0022782885935157537,
          "validation_metric": 0.001957261934876442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3307130560006044
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02046690188581124,
          "train_loss": 0.019571326673030853,
          "validation_loss": 0.016824472695589066,
          "train_metric": 0.0019711521454155445,
          "validation_metric": 0.00169449916575104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3433340280007542
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.021430058928672224,
          "train_loss": 0.019809599965810776,
          "validation_loss": 0.01710662618279457,
          "train_metric": 0.001995150465518236,
          "validation_metric": 0.0017229169607162476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3556467969992809
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.019933807488996536,
          "train_loss": 0.018322203308343887,
          "validation_loss": 0.015159186907112598,
          "train_metric": 0.0018453451339155436,
          "validation_metric": 0.0015267777489498258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36864424600025814
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.0184865053743124,
          "train_loss": 0.018727105110883713,
          "validation_loss": 0.016270481050014496,
          "train_metric": 0.0018861254211515188,
          "validation_metric": 0.001638703397475183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38305166700047266
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019489634316414595,
          "train_loss": 0.018848121166229248,
          "validation_loss": 0.015607197768986225,
          "train_metric": 0.0018983136396855116,
          "validation_metric": 0.0015718997456133366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3958769109995046
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.025991787144448608,
          "train_loss": 0.022055301815271378,
          "validation_loss": 0.01751544512808323,
          "train_metric": 0.0022213291376829147,
          "validation_metric": 0.0017640916630625725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40797470600045926
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.020158647326752543,
          "train_loss": 0.019878527149558067,
          "validation_loss": 0.015789136290550232,
          "train_metric": 0.0020020923111587763,
          "validation_metric": 0.0015902240993455052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41956010600006266
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01832235194160603,
          "train_loss": 0.015337939374148846,
          "validation_loss": 0.013187260366976261,
          "train_metric": 0.0015447810292243958,
          "validation_metric": 0.001328172511421144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43243937800070853
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015697864320827648,
          "train_loss": 0.01441369578242302,
          "validation_loss": 0.012476164847612381,
          "train_metric": 0.0014516946393996477,
          "validation_metric": 0.0012565534561872482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44555964800019865
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.01796175292111002,
          "train_loss": 0.015583343803882599,
          "validation_loss": 0.012778116390109062,
          "train_metric": 0.0015694972826167941,
          "validation_metric": 0.0012869652127847075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4576137320000271
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.017315161181613803,
          "train_loss": 0.026303283870220184,
          "validation_loss": 0.022693904116749763,
          "train_metric": 0.0026491701137274504,
          "validation_metric": 0.002285646740347147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46998167099991406
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.018650115351192653,
          "train_loss": 0.015033909119665623,
          "validation_loss": 0.012586742639541626,
          "train_metric": 0.0015141601907089353,
          "validation_metric": 0.0012676906771957874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4817416499990941
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01525422916165553,
          "train_loss": 0.012996863573789597,
          "validation_loss": 0.011027900502085686,
          "train_metric": 0.0013089965796098113,
          "validation_metric": 0.0011106899473816156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49469531400063715
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.017277867038501427,
          "train_loss": 0.018468111753463745,
          "validation_loss": 0.01605597883462906,
          "train_metric": 0.0018600407056510448,
          "validation_metric": 0.001617099391296506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5067682490007428
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.017836909275501966,
          "train_loss": 0.015464769676327705,
          "validation_loss": 0.013479636050760746,
          "train_metric": 0.001557554816827178,
          "validation_metric": 0.0013576196506619453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5186138080007368
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014926508301869035,
          "train_loss": 0.01525525376200676,
          "validation_loss": 0.012201250530779362,
          "train_metric": 0.0015364533755928278,
          "validation_metric": 0.00122886523604393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.530933272000766
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.013606283508124761,
          "train_loss": 0.01189349964261055,
          "validation_loss": 0.010086850263178349,
          "train_metric": 0.0011978696566075087,
          "validation_metric": 0.0010159105295315385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5437800769996102
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013865286426153034,
          "train_loss": 0.01421459298580885,
          "validation_loss": 0.012337259016931057,
          "train_metric": 0.0014316417509689927,
          "validation_metric": 0.00124256347771734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5560948399997869
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.013906719570513815,
          "train_loss": 0.01308034174144268,
          "validation_loss": 0.011206195689737797,
          "train_metric": 0.0013174042105674744,
          "validation_metric": 0.0011286470107734203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.567948398000226
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.015102692559594288,
          "train_loss": 0.012156970798969269,
          "validation_loss": 0.009952702559530735,
          "train_metric": 0.0012244055978953838,
          "validation_metric": 0.00100239971652627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5829965550001361
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.011600760830333456,
          "train_loss": 0.01074554119259119,
          "validation_loss": 0.009170734323561192,
          "train_metric": 0.0010822515469044447,
          "validation_metric": 0.0009236427722498775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.596283326000048
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011461123736808077,
          "train_loss": 0.011186208575963974,
          "validation_loss": 0.008815637789666653,
          "train_metric": 0.001126633957028389,
          "validation_metric": 0.000887878704816103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6097750770004495
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.012731426904792897,
          "train_loss": 0.011028323322534561,
          "validation_loss": 0.008532376028597355,
          "train_metric": 0.0011107323225587606,
          "validation_metric": 0.0008593496750108898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6237899600000674
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011592802344239317,
          "train_loss": 0.01758863590657711,
          "validation_loss": 0.0152521887794137,
          "train_metric": 0.0017714631976559758,
          "validation_metric": 0.0015361446421593428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6400276349995693
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014004348166054115,
          "train_loss": 0.011498655192553997,
          "validation_loss": 0.009649231098592281,
          "train_metric": 0.0011581024155020714,
          "validation_metric": 0.0009718352230265737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6574310430005426
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.011971019470365718,
          "train_loss": 0.010247869417071342,
          "validation_loss": 0.008546342141926289,
          "train_metric": 0.0010321278823539615,
          "validation_metric": 0.0008607562049292028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.671868462999555
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010324565839255229,
          "train_loss": 0.009410005994141102,
          "validation_loss": 0.008049655705690384,
          "train_metric": 0.0009477414423599839,
          "validation_metric": 0.0008107318426482379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6852054610008054
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0099878313485533,
          "train_loss": 0.009574130177497864,
          "validation_loss": 0.008396321907639503,
          "train_metric": 0.0009642713703215122,
          "validation_metric": 0.0008456468349322677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7059081070001412
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010388583701569587,
          "train_loss": 0.010846927762031555,
          "validation_loss": 0.009024742990732193,
          "train_metric": 0.0010924628004431725,
          "validation_metric": 0.0009089390514418483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7174142839994602
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.011729866499081254,
          "train_loss": 0.008839637972414494,
          "validation_loss": 0.007451624143868685,
          "train_metric": 0.0008902959525585175,
          "validation_metric": 0.0007505002431571484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7292529940004897
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009075396141270176,
          "train_loss": 0.008698005229234695,
          "validation_loss": 0.007412043400108814,
          "train_metric": 0.0008760312921367586,
          "validation_metric": 0.0007465138914994895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7418096500005049
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008586802636273205,
          "train_loss": 0.009239615872502327,
          "validation_loss": 0.00846143253147602,
          "train_metric": 0.0009305804269388318,
          "validation_metric": 0.000852204451803118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7537643169998773
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.009399648464750499,
          "train_loss": 0.008207714185118675,
          "validation_loss": 0.007094361819326878,
          "train_metric": 0.0008266508812084794,
          "validation_metric": 0.0007145180716179311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7667185720001726
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.009622707089874893,
          "train_loss": 0.008783034980297089,
          "validation_loss": 0.007847893983125687,
          "train_metric": 0.0008845950360409915,
          "validation_metric": 0.0007904112571850419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7790183569995861
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009374020868563093,
          "train_loss": 0.008376436308026314,
          "validation_loss": 0.007370653096586466,
          "train_metric": 0.0008436439675278962,
          "validation_metric": 0.0007423452916555107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.791290110000773
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008409974529058672,
          "train_loss": 0.008557048626244068,
          "validation_loss": 0.006746719125658274,
          "train_metric": 0.0008618345018476248,
          "validation_metric": 0.0006795047665946186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.804182760999538
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.008126632194034755,
          "train_loss": 0.007593871559947729,
          "validation_loss": 0.006299670785665512,
          "train_metric": 0.000764826952945441,
          "validation_metric": 0.0006344797438941896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8168019820004702
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008103969594230875,
          "train_loss": 0.007627340033650398,
          "validation_loss": 0.006149530876427889,
          "train_metric": 0.0007681977585889399,
          "validation_metric": 0.0006193582667037845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8294154039995192
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007666467514354736,
          "train_loss": 0.007350016385316849,
          "validation_loss": 0.0060503422282636166,
          "train_metric": 0.0007402666378766298,
          "validation_metric": 0.0006093683768995106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8422448959991016
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.008778885996434838,
          "train_loss": 0.007884827442467213,
          "validation_loss": 0.007127765100449324,
          "train_metric": 0.0007941309013403952,
          "validation_metric": 0.0007178823580034077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8545854170006351
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007124831900000572,
          "train_loss": 0.008507896214723587,
          "validation_loss": 0.006483363453298807,
          "train_metric": 0.0008568841149099171,
          "validation_metric": 0.0006529804668389261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8661065889991733
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.00802242913050577,
          "train_loss": 0.0072455033659935,
          "validation_loss": 0.005634790752083063,
          "train_metric": 0.000729740597307682,
          "validation_metric": 0.0005675156135112047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.878004272999533
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007499659288441762,
          "train_loss": 0.006905286107212305,
          "validation_loss": 0.005780530162155628,
          "train_metric": 0.0006954751443117857,
          "validation_metric": 0.0005821938975714147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8902344090001861
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006983088416745886,
          "train_loss": 0.00676384475082159,
          "validation_loss": 0.005742666777223349,
          "train_metric": 0.0006812297506257892,
          "validation_metric": 0.0005783804226666689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9017638130007981
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007186421637015883,
          "train_loss": 0.006360976025462151,
          "validation_loss": 0.0053828442469239235,
          "train_metric": 0.0006406542379409075,
          "validation_metric": 0.000542140391189605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9140058289995068
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008146895095705986,
          "train_loss": 0.00710770720615983,
          "validation_loss": 0.0064037274569272995,
          "train_metric": 0.0007158622611314058,
          "validation_metric": 0.0006449599750339985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9268170300001657
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0070507670170627534,
          "train_loss": 0.006784606724977493,
          "validation_loss": 0.005462770350277424,
          "train_metric": 0.0006833208026364446,
          "validation_metric": 0.0005501902778632939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9385074830006488
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007892993970017415,
          "train_loss": 0.007619489450007677,
          "validation_loss": 0.005710892844945192,
          "train_metric": 0.0007674070075154305,
          "validation_metric": 0.0005751801654696465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9507305480001378
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0075215069082332775,
          "train_loss": 0.008999602869153023,
          "validation_loss": 0.007687784265726805,
          "train_metric": 0.0009064070181921124,
          "validation_metric": 0.0007742854650132358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9626864420006314
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.008329692267579958,
          "train_loss": 0.006699476856738329,
          "validation_loss": 0.005957436747848988,
          "train_metric": 0.000674746697768569,
          "validation_metric": 0.0006000111461617053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.975532424999983
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006823713432822842,
          "train_loss": 0.006718658842146397,
          "validation_loss": 0.005331420339643955,
          "train_metric": 0.0006766787264496088,
          "validation_metric": 0.0005369611317291856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9876782739993359
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0063610283978050575,
          "train_loss": 0.006560879293829203,
          "validation_loss": 0.005013317335397005,
          "train_metric": 0.0006607878021895885,
          "validation_metric": 0.0005049229366704822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0001743619995977
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006628050803556107,
          "train_loss": 0.005550465080887079,
          "validation_loss": 0.00450581219047308,
          "train_metric": 0.0005590225337073207,
          "validation_metric": 0.00045380889787338674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.012686482999925
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006065987341571599,
          "train_loss": 0.006213136482983828,
          "validation_loss": 0.004682971630245447,
          "train_metric": 0.000625764369033277,
          "validation_metric": 0.0004716518160421401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0248585459994501
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0066800750500988215,
          "train_loss": 0.007923943921923637,
          "validation_loss": 0.00671512121334672,
          "train_metric": 0.0007980706868693233,
          "validation_metric": 0.0006763224373571575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.03649960400071
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.008286909156595357,
          "train_loss": 0.005879192613065243,
          "validation_loss": 0.005012104287743568,
          "train_metric": 0.0005921307601965964,
          "validation_metric": 0.0005048008169978857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0482806420004636
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.008240004448452964,
          "train_loss": 0.006389837712049484,
          "validation_loss": 0.005709755700081587,
          "train_metric": 0.0006435610703192651,
          "validation_metric": 0.0005750657292082906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0604206280004291
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007031992237898521,
          "train_loss": 0.0061054527759552,
          "validation_loss": 0.005275707691907883,
          "train_metric": 0.000614918943028897,
          "validation_metric": 0.0005313499132171273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0723332109992043
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006232404222828336,
          "train_loss": 0.005883560515940189,
          "validation_loss": 0.004626329056918621,
          "train_metric": 0.0005925706936977804,
          "validation_metric": 0.00046594697050750256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0841412650006532
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006698730117932428,
          "train_loss": 0.00830888468772173,
          "validation_loss": 0.005257878918200731,
          "train_metric": 0.0008368405397050083,
          "validation_metric": 0.0005295543232932687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0962097770006949
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.007175867009209469,
          "train_loss": 0.005164492409676313,
          "validation_loss": 0.0041005248203873634,
          "train_metric": 0.0005201487801969051,
          "validation_metric": 0.00041298987343907356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.108934046000286
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005640656236209907,
          "train_loss": 0.005428474862128496,
          "validation_loss": 0.004622603766620159,
          "train_metric": 0.000546736060641706,
          "validation_metric": 0.00046557170571759343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.121310442000322
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.007278883465914987,
          "train_loss": 0.007279138546437025,
          "validation_loss": 0.005336048547178507,
          "train_metric": 0.0007331281085498631,
          "validation_metric": 0.0005374272586777806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.133177108999007
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006428858388972003,
          "train_loss": 0.005414268001914024,
          "validation_loss": 0.004378072917461395,
          "train_metric": 0.000545305258128792,
          "validation_metric": 0.0004409434332046658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1449509960002615
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.006930987336090766,
          "train_loss": 0.004939272999763489,
          "validation_loss": 0.004174148663878441,
          "train_metric": 0.0004974654875695705,
          "validation_metric": 0.0004204050055705011,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.15696688800017
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005648306665534619,
          "train_loss": 0.004681389778852463,
          "validation_loss": 0.004043219611048698,
          "train_metric": 0.00047149244346655905,
          "validation_metric": 0.0004072182346135378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.16894381199927
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005824902589665726,
          "train_loss": 0.005570245441049337,
          "validation_loss": 0.004077776335179806,
          "train_metric": 0.0005610146909020841,
          "validation_metric": 0.00041069870349019766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.180864580999696
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005732758989324793,
          "train_loss": 0.006048447918146849,
          "validation_loss": 0.0051933652721345425,
          "train_metric": 0.0006091775139793754,
          "validation_metric": 0.0005230566603131592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1933542949991534
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.00487511732353596,
          "train_loss": 0.004281728528439999,
          "validation_loss": 0.003500000573694706,
          "train_metric": 0.00043124001240357757,
          "validation_metric": 0.00035250725341029465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2060548520003067
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005490752446348779,
          "train_loss": 0.0075278449803590775,
          "validation_loss": 0.005237332079559565,
          "train_metric": 0.000758176960516721,
          "validation_metric": 0.0005274848663248122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2178867279999395
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.006032784658600576,
          "train_loss": 0.004147201310843229,
          "validation_loss": 0.003196843434125185,
          "train_metric": 0.00041769095696508884,
          "validation_metric": 0.00032197439577430487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2307063179996476
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0045539015118265525,
          "train_loss": 0.003808197332546115,
          "validation_loss": 0.0031398420687764883,
          "train_metric": 0.0003835476818494499,
          "validation_metric": 0.00031623340328224003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.243711782999526
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004446769387868699,
          "train_loss": 0.003838970558717847,
          "validation_loss": 0.00305731687694788,
          "train_metric": 0.00038664706517010927,
          "validation_metric": 0.0003079217567574233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2565293589996145
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004661549588490743,
          "train_loss": 0.003889218671247363,
          "validation_loss": 0.0030916959512978792,
          "train_metric": 0.0003917078720405698,
          "validation_metric": 0.0003113843558821827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.269117040999845
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004343237917055376,
          "train_loss": 0.003794271731749177,
          "validation_loss": 0.0030378333758562803,
          "train_metric": 0.0003821451682597399,
          "validation_metric": 0.000305959430988878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.281432092000614
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004010706965345889,
          "train_loss": 0.004695076495409012,
          "validation_loss": 0.0032074693590402603,
          "train_metric": 0.0004728709172923118,
          "validation_metric": 0.0003230445727240294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.29379417599921
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0048800584190757945,
          "train_loss": 0.004135572351515293,
          "validation_loss": 0.003357178997248411,
          "train_metric": 0.00041651964420452714,
          "validation_metric": 0.00033812274341471493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3071879920007632
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.004035678837681189,
          "train_loss": 0.003541690995916724,
          "validation_loss": 0.0029904022812843323,
          "train_metric": 0.00035670612123794854,
          "validation_metric": 0.0003011823573615402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3219475770001736
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004549244105874095,
          "train_loss": 0.007052480708807707,
          "validation_loss": 0.005183737725019455,
          "train_metric": 0.0007103000534698367,
          "validation_metric": 0.0005220869788900018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3453818369998771
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.008762747966102324,
          "train_loss": 0.004422488622367382,
          "validation_loss": 0.003422736655920744,
          "train_metric": 0.00044541689567267895,
          "validation_metric": 0.00034472550032660365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3650768799998332
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004970894042344298,
          "train_loss": 0.007190850097686052,
          "validation_loss": 0.004736239090561867,
          "train_metric": 0.0007242360734380782,
          "validation_metric": 0.00047701667062938213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.382957529000123
        }
      ],
      "best": {
        "epoch": 97,
        "updates": 1552
      },
      "best_validation_metric": 0.0003011823573615402,
      "config": {
        "label": "Batch 64",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01735460622893934
    },
    "014fce8be51b_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 64,
          "samples_seen": 1024,
          "online_loss": 0.30405234018689953,
          "train_loss": 0.09432892501354218,
          "validation_loss": 0.07780647277832031,
          "train_metric": 0.009500463493168354,
          "validation_metric": 0.00783638097345829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26459588699981396
        },
        {
          "epoch": 2,
          "updates": 128,
          "samples_seen": 2048,
          "online_loss": 0.07553436110902112,
          "train_loss": 0.06811955571174622,
          "validation_loss": 0.056377023458480835,
          "train_metric": 0.006860751658678055,
          "validation_metric": 0.005678087472915649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5602700930012361
        },
        {
          "epoch": 3,
          "updates": 192,
          "samples_seen": 3072,
          "online_loss": 0.058350919804070145,
          "train_loss": 0.05071042850613594,
          "validation_loss": 0.04290137439966202,
          "train_metric": 0.005107368342578411,
          "validation_metric": 0.004320869222283363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7894224709998525
        },
        {
          "epoch": 4,
          "updates": 256,
          "samples_seen": 4096,
          "online_loss": 0.05183539605786791,
          "train_loss": 0.04143005609512329,
          "validation_loss": 0.03292068466544151,
          "train_metric": 0.004172683227807283,
          "validation_metric": 0.003315650625154376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.917456619999939
        },
        {
          "epoch": 5,
          "updates": 320,
          "samples_seen": 5120,
          "online_loss": 0.041577326410333626,
          "train_loss": 0.0330607108771801,
          "validation_loss": 0.025740105658769608,
          "train_metric": 0.0033297536429017782,
          "validation_metric": 0.0025924488436430693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0375486740013002
        },
        {
          "epoch": 6,
          "updates": 384,
          "samples_seen": 6144,
          "online_loss": 0.03692157486148062,
          "train_loss": 0.028630802407860756,
          "validation_loss": 0.02386183850467205,
          "train_metric": 0.0028835893608629704,
          "validation_metric": 0.0024032772053033113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0833128860012948
        },
        {
          "epoch": 7,
          "updates": 448,
          "samples_seen": 7168,
          "online_loss": 0.029337134736124426,
          "train_loss": 0.021653776988387108,
          "validation_loss": 0.018267223611474037,
          "train_metric": 0.0021808892488479614,
          "validation_metric": 0.0018398077227175236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1306692720008868
        },
        {
          "epoch": 8,
          "updates": 512,
          "samples_seen": 8192,
          "online_loss": 0.028584459039848298,
          "train_loss": 0.0523504801094532,
          "validation_loss": 0.036134831607341766,
          "train_metric": 0.005272548645734787,
          "validation_metric": 0.0036393676418811083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.177010933000929
        },
        {
          "epoch": 9,
          "updates": 576,
          "samples_seen": 9216,
          "online_loss": 0.034820369830413256,
          "train_loss": 0.027965635061264038,
          "validation_loss": 0.02425423450767994,
          "train_metric": 0.0028165963012725115,
          "validation_metric": 0.002442797878757119,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3611489209997671
        },
        {
          "epoch": 10,
          "updates": 640,
          "samples_seen": 10240,
          "online_loss": 0.022823747025540797,
          "train_loss": 0.02376246452331543,
          "validation_loss": 0.021696094423532486,
          "train_metric": 0.002393268281593919,
          "validation_metric": 0.0021851512137800455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4030861030005326
        },
        {
          "epoch": 11,
          "updates": 704,
          "samples_seen": 11264,
          "online_loss": 0.021480128412804333,
          "train_loss": 0.01560952141880989,
          "validation_loss": 0.012483786791563034,
          "train_metric": 0.0015721336239948869,
          "validation_metric": 0.0012573213316500187,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4408212670005014
        },
        {
          "epoch": 12,
          "updates": 768,
          "samples_seen": 12288,
          "online_loss": 0.016016627560020424,
          "train_loss": 0.015190854668617249,
          "validation_loss": 0.013768654316663742,
          "train_metric": 0.0015299671795219183,
          "validation_metric": 0.0013867283705621958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4773308430012548
        },
        {
          "epoch": 13,
          "updates": 832,
          "samples_seen": 13312,
          "online_loss": 0.013792937425932905,
          "train_loss": 0.013874605298042297,
          "validation_loss": 0.01084647886455059,
          "train_metric": 0.00139739946462214,
          "validation_metric": 0.0010924176312983036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5128958719997172
        },
        {
          "epoch": 14,
          "updates": 896,
          "samples_seen": 14336,
          "online_loss": 0.015341836156949284,
          "train_loss": 0.019977113232016563,
          "validation_loss": 0.018726734444499016,
          "train_metric": 0.002012021606788039,
          "validation_metric": 0.0018860880518332124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5474732310012769
        },
        {
          "epoch": 15,
          "updates": 960,
          "samples_seen": 15360,
          "online_loss": 0.01804348247424059,
          "train_loss": 0.026645472273230553,
          "validation_loss": 0.021020296961069107,
          "train_metric": 0.002683634404093027,
          "validation_metric": 0.0021170873660594225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5822362350008916
        },
        {
          "epoch": 16,
          "updates": 1024,
          "samples_seen": 16384,
          "online_loss": 0.015056374375490122,
          "train_loss": 0.011738265864551067,
          "validation_loss": 0.010271968320012093,
          "train_metric": 0.0011822350788861513,
          "validation_metric": 0.0010345550253987312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6168132229995535
        },
        {
          "epoch": 17,
          "updates": 1088,
          "samples_seen": 17408,
          "online_loss": 0.013324697471034597,
          "train_loss": 0.01271234080195427,
          "validation_loss": 0.011665533296763897,
          "train_metric": 0.0012803403660655022,
          "validation_metric": 0.0011749097611755133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6485113480011933
        },
        {
          "epoch": 18,
          "updates": 1152,
          "samples_seen": 18432,
          "online_loss": 0.01210375746632053,
          "train_loss": 0.014222957193851471,
          "validation_loss": 0.00854982528835535,
          "train_metric": 0.0014324841322377324,
          "validation_metric": 0.0008611070807091892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6819739290003781
        },
        {
          "epoch": 19,
          "updates": 1216,
          "samples_seen": 19456,
          "online_loss": 0.016799078177427873,
          "train_loss": 0.012104406021535397,
          "validation_loss": 0.01112705934792757,
          "train_metric": 0.0012191114947199821,
          "validation_metric": 0.0011206766357645392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7143580480005767
        },
        {
          "epoch": 20,
          "updates": 1280,
          "samples_seen": 20480,
          "online_loss": 0.012688635461017839,
          "train_loss": 0.010795699432492256,
          "validation_loss": 0.00687992200255394,
          "train_metric": 0.0010873033897951245,
          "validation_metric": 0.0006929205846972764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7484933320010896
        },
        {
          "epoch": 21,
          "updates": 1344,
          "samples_seen": 21504,
          "online_loss": 0.010700059143346152,
          "train_loss": 0.00927622802555561,
          "validation_loss": 0.006946666166186333,
          "train_metric": 0.0009342675912193954,
          "validation_metric": 0.000699642754625529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8384668090002378
        },
        {
          "epoch": 22,
          "updates": 1408,
          "samples_seen": 22528,
          "online_loss": 0.010594090073936968,
          "train_loss": 0.007905232720077038,
          "validation_loss": 0.005931931547820568,
          "train_metric": 0.0007961862138472497,
          "validation_metric": 0.0005974425002932549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9107421030003024
        },
        {
          "epoch": 23,
          "updates": 1472,
          "samples_seen": 23552,
          "online_loss": 0.01006094761987697,
          "train_loss": 0.00798938050866127,
          "validation_loss": 0.005514916963875294,
          "train_metric": 0.0008046611328609288,
          "validation_metric": 0.0005554422968998551,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9474955440000485
        },
        {
          "epoch": 24,
          "updates": 1536,
          "samples_seen": 24576,
          "online_loss": 0.011372572630534705,
          "train_loss": 0.007471478544175625,
          "validation_loss": 0.006100369617342949,
          "train_metric": 0.0007525000255554914,
          "validation_metric": 0.0006144069484435022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9825885280006332
        },
        {
          "epoch": 25,
          "updates": 1600,
          "samples_seen": 25600,
          "online_loss": 0.014801576882746303,
          "train_loss": 0.00864070188254118,
          "validation_loss": 0.007496920879930258,
          "train_metric": 0.0008702597697265446,
          "validation_metric": 0.0007550624432042241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.017170065000755
        }
      ],
      "best": {
        "epoch": 23,
        "updates": 1472
      },
      "best_validation_metric": 0.0005554422968998551,
      "config": {
        "label": "Batch 16",
        "batch": 16,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.0235678233381841
    },
    "1fb401058891_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 1,
          "samples_seen": 1024,
          "online_loss": 1.0258803367614746,
          "train_loss": 0.9930850267410278,
          "validation_loss": 0.7092403173446655,
          "train_metric": 0.10001988708972931,
          "validation_metric": 0.07143208384513855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.00815888199940673
        },
        {
          "epoch": 2,
          "updates": 2,
          "samples_seen": 2048,
          "online_loss": 0.9930850267410278,
          "train_loss": 0.9605929851531982,
          "validation_loss": 0.6880066990852356,
          "train_metric": 0.09674739837646484,
          "validation_metric": 0.06929351389408112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.014033245999598876
        },
        {
          "epoch": 3,
          "updates": 3,
          "samples_seen": 3072,
          "online_loss": 0.9605928659439087,
          "train_loss": 0.9267858266830444,
          "validation_loss": 0.6659043431282043,
          "train_metric": 0.09334246814250946,
          "validation_metric": 0.06706744432449341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01985638899896003
        },
        {
          "epoch": 4,
          "updates": 4,
          "samples_seen": 4096,
          "online_loss": 0.9267858266830444,
          "train_loss": 0.8905212879180908,
          "validation_loss": 0.6421554684638977,
          "train_metric": 0.08969003707170486,
          "validation_metric": 0.06467554718255997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11112558700006048
        },
        {
          "epoch": 5,
          "updates": 5,
          "samples_seen": 5120,
          "online_loss": 0.8905212879180908,
          "train_loss": 0.8509505987167358,
          "validation_loss": 0.6161958575248718,
          "train_metric": 0.08570463955402374,
          "validation_metric": 0.062060996890068054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11712545799855434
        },
        {
          "epoch": 6,
          "updates": 6,
          "samples_seen": 6144,
          "online_loss": 0.8509507179260254,
          "train_loss": 0.8074913620948792,
          "validation_loss": 0.587623655796051,
          "train_metric": 0.08132757246494293,
          "validation_metric": 0.05918330326676369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12294379400009348
        },
        {
          "epoch": 7,
          "updates": 7,
          "samples_seen": 7168,
          "online_loss": 0.8074913620948792,
          "train_loss": 0.7598076462745667,
          "validation_loss": 0.5561920404434204,
          "train_metric": 0.07652503252029419,
          "validation_metric": 0.05601762980222702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12850124599935953
        },
        {
          "epoch": 8,
          "updates": 8,
          "samples_seen": 8192,
          "online_loss": 0.7598075866699219,
          "train_loss": 0.7078191637992859,
          "validation_loss": 0.5218236446380615,
          "train_metric": 0.07128895819187164,
          "validation_metric": 0.05255616456270218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1340533000002324
        },
        {
          "epoch": 9,
          "updates": 9,
          "samples_seen": 9216,
          "online_loss": 0.7078192234039307,
          "train_loss": 0.6517389416694641,
          "validation_loss": 0.4846305251121521,
          "train_metric": 0.06564076244831085,
          "validation_metric": 0.048810213804244995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14023597599953064
        },
        {
          "epoch": 10,
          "updates": 10,
          "samples_seen": 10240,
          "online_loss": 0.6517389416694641,
          "train_loss": 0.5921094417572021,
          "validation_loss": 0.4449300765991211,
          "train_metric": 0.05963509902358055,
          "validation_metric": 0.04481172934174538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14587566499903915
        },
        {
          "epoch": 11,
          "updates": 11,
          "samples_seen": 11264,
          "online_loss": 0.5921094417572021,
          "train_loss": 0.5298457741737366,
          "validation_loss": 0.40326598286628723,
          "train_metric": 0.05336412787437439,
          "validation_metric": 0.04061547666788101,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15149202899920056
        },
        {
          "epoch": 12,
          "updates": 12,
          "samples_seen": 12288,
          "online_loss": 0.5298458337783813,
          "train_loss": 0.4663173258304596,
          "validation_loss": 0.36046072840690613,
          "train_metric": 0.04696577414870262,
          "validation_metric": 0.0363042838871479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1572370880003291
        },
        {
          "epoch": 13,
          "updates": 13,
          "samples_seen": 13312,
          "online_loss": 0.4663173258304596,
          "train_loss": 0.4034489691257477,
          "validation_loss": 0.31767868995666504,
          "train_metric": 0.0406339056789875,
          "validation_metric": 0.03199543431401253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1627545339997596
        },
        {
          "epoch": 14,
          "updates": 14,
          "samples_seen": 14336,
          "online_loss": 0.40344899892807007,
          "train_loss": 0.343758225440979,
          "validation_loss": 0.27644097805023193,
          "train_metric": 0.03462206944823265,
          "validation_metric": 0.027842121198773384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16830288999881304
        },
        {
          "epoch": 15,
          "updates": 15,
          "samples_seen": 15360,
          "online_loss": 0.343758225440979,
          "train_loss": 0.2902417480945587,
          "validation_loss": 0.23856022953987122,
          "train_metric": 0.0292320866137743,
          "validation_metric": 0.024026911705732346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17439051200017275
        },
        {
          "epoch": 16,
          "updates": 16,
          "samples_seen": 16384,
          "online_loss": 0.2902417480945587,
          "train_loss": 0.2460261732339859,
          "validation_loss": 0.20596328377723694,
          "train_metric": 0.024778854101896286,
          "validation_metric": 0.02074386551976204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18032013599986385
        },
        {
          "epoch": 17,
          "updates": 17,
          "samples_seen": 17408,
          "online_loss": 0.2460261732339859,
          "train_loss": 0.21371406316757202,
          "validation_loss": 0.18037617206573486,
          "train_metric": 0.021524500101804733,
          "validation_metric": 0.01816682517528534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18612554199899023
        },
        {
          "epoch": 18,
          "updates": 18,
          "samples_seen": 18432,
          "online_loss": 0.21371407806873322,
          "train_loss": 0.19444365799427032,
          "validation_loss": 0.16286668181419373,
          "train_metric": 0.019583653658628464,
          "validation_metric": 0.016403336077928543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19255885199891054
        },
        {
          "epoch": 19,
          "updates": 19,
          "samples_seen": 19456,
          "online_loss": 0.19444367289543152,
          "train_loss": 0.1870141327381134,
          "validation_loss": 0.15339042246341705,
          "train_metric": 0.01883537508547306,
          "validation_metric": 0.0154489204287529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19832862399925943
        },
        {
          "epoch": 20,
          "updates": 20,
          "samples_seen": 20480,
          "online_loss": 0.18701410293579102,
          "train_loss": 0.1878947913646698,
          "validation_loss": 0.15067341923713684,
          "train_metric": 0.01892407238483429,
          "validation_metric": 0.015175273641943932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20383214899993618
        },
        {
          "epoch": 21,
          "updates": 21,
          "samples_seen": 21504,
          "online_loss": 0.1878947764635086,
          "train_loss": 0.19239771366119385,
          "validation_loss": 0.15249933302402496,
          "train_metric": 0.01937759481370449,
          "validation_metric": 0.015359176322817802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20901978499932738
        },
        {
          "epoch": 22,
          "updates": 22,
          "samples_seen": 22528,
          "online_loss": 0.19239771366119385,
          "train_loss": 0.19602666795253754,
          "validation_loss": 0.1560187041759491,
          "train_metric": 0.019743088632822037,
          "validation_metric": 0.015713632106781006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21409547999974166
        },
        {
          "epoch": 23,
          "updates": 23,
          "samples_seen": 23552,
          "online_loss": 0.19602665305137634,
          "train_loss": 0.19552338123321533,
          "validation_loss": 0.15826033055782318,
          "train_metric": 0.019692396745085716,
          "validation_metric": 0.01593940332531929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.219091438999385
        },
        {
          "epoch": 24,
          "updates": 24,
          "samples_seen": 24576,
          "online_loss": 0.19552338123321533,
          "train_loss": 0.18979987502098083,
          "validation_loss": 0.15719343721866608,
          "train_metric": 0.01911594718694687,
          "validation_metric": 0.015831949189305305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22438145599880954
        },
        {
          "epoch": 25,
          "updates": 25,
          "samples_seen": 25600,
          "online_loss": 0.18979986011981964,
          "train_loss": 0.17997907102108002,
          "validation_loss": 0.15244463086128235,
          "train_metric": 0.0181268323212862,
          "validation_metric": 0.015353664755821228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22938880800029438
        },
        {
          "epoch": 26,
          "updates": 26,
          "samples_seen": 26624,
          "online_loss": 0.17997907102108002,
          "train_loss": 0.168293759226799,
          "validation_loss": 0.14496231079101562,
          "train_metric": 0.016949929296970367,
          "validation_metric": 0.014600072987377644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23488257399912982
        },
        {
          "epoch": 27,
          "updates": 27,
          "samples_seen": 27648,
          "online_loss": 0.168293759226799,
          "train_loss": 0.1568431854248047,
          "validation_loss": 0.13618676364421844,
          "train_metric": 0.01579667069017887,
          "validation_metric": 0.013716231100261211,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24082424699918192
        },
        {
          "epoch": 28,
          "updates": 28,
          "samples_seen": 28672,
          "online_loss": 0.1568431854248047,
          "train_loss": 0.1469542235136032,
          "validation_loss": 0.12741924822330475,
          "train_metric": 0.014800691977143288,
          "validation_metric": 0.01283320039510727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2466043059994263
        },
        {
          "epoch": 29,
          "updates": 29,
          "samples_seen": 29696,
          "online_loss": 0.1469542235136032,
          "train_loss": 0.13911910355091095,
          "validation_loss": 0.11954168230295181,
          "train_metric": 0.014011567458510399,
          "validation_metric": 0.01203980017453432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2523330209987762
        },
        {
          "epoch": 30,
          "updates": 30,
          "samples_seen": 30720,
          "online_loss": 0.13911908864974976,
          "train_loss": 0.13322202861309052,
          "validation_loss": 0.11298336833715439,
          "train_metric": 0.013417634181678295,
          "validation_metric": 0.011379270814359188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2588271099994017
        },
        {
          "epoch": 31,
          "updates": 31,
          "samples_seen": 31744,
          "online_loss": 0.13322202861309052,
          "train_loss": 0.12883873283863068,
          "validation_loss": 0.10782220959663391,
          "train_metric": 0.012976166792213917,
          "validation_metric": 0.010859457775950432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26455191399873
        },
        {
          "epoch": 32,
          "updates": 32,
          "samples_seen": 32768,
          "online_loss": 0.12883874773979187,
          "train_loss": 0.12548963725566864,
          "validation_loss": 0.10392986983060837,
          "train_metric": 0.01263885572552681,
          "validation_metric": 0.010467435233294964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2704175529997883
        },
        {
          "epoch": 33,
          "updates": 33,
          "samples_seen": 33792,
          "online_loss": 0.12548963725566864,
          "train_loss": 0.12279529124498367,
          "validation_loss": 0.10109636932611465,
          "train_metric": 0.012367491610348225,
          "validation_metric": 0.01018205564469099,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27638818899868056
        },
        {
          "epoch": 34,
          "updates": 34,
          "samples_seen": 34816,
          "online_loss": 0.12279529869556427,
          "train_loss": 0.12052901089191437,
          "validation_loss": 0.0991029366850853,
          "train_metric": 0.012139239348471165,
          "validation_metric": 0.00998128391802311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28211846799968043
        },
        {
          "epoch": 35,
          "updates": 35,
          "samples_seen": 35840,
          "online_loss": 0.12052901834249496,
          "train_loss": 0.11858896911144257,
          "validation_loss": 0.09774117916822433,
          "train_metric": 0.011943846940994263,
          "validation_metric": 0.00984413456171751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2880772299995442
        },
        {
          "epoch": 36,
          "updates": 36,
          "samples_seen": 36864,
          "online_loss": 0.11858897656202316,
          "train_loss": 0.11692400276660919,
          "validation_loss": 0.09679676592350006,
          "train_metric": 0.011776156723499298,
          "validation_metric": 0.009749014861881733,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2946806899999501
        },
        {
          "epoch": 37,
          "updates": 37,
          "samples_seen": 37888,
          "online_loss": 0.1169239953160286,
          "train_loss": 0.11545105278491974,
          "validation_loss": 0.0960247591137886,
          "train_metric": 0.011627807281911373,
          "validation_metric": 0.009671262465417385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3005961419985397
        },
        {
          "epoch": 38,
          "updates": 38,
          "samples_seen": 38912,
          "online_loss": 0.11545105278491974,
          "train_loss": 0.11400263011455536,
          "validation_loss": 0.09514527022838593,
          "train_metric": 0.011481929570436478,
          "validation_metric": 0.009582682512700558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30671746999905736
        },
        {
          "epoch": 39,
          "updates": 39,
          "samples_seen": 39936,
          "online_loss": 0.11400264501571655,
          "train_loss": 0.11233367770910263,
          "validation_loss": 0.093878373503685,
          "train_metric": 0.011313837952911854,
          "validation_metric": 0.009455086663365364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3129695959996752
        },
        {
          "epoch": 40,
          "updates": 40,
          "samples_seen": 40960,
          "online_loss": 0.11233370006084442,
          "train_loss": 0.11019480228424072,
          "validation_loss": 0.09201671183109283,
          "train_metric": 0.011098417453467846,
          "validation_metric": 0.009267585352063179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3188776679999137
        },
        {
          "epoch": 41,
          "updates": 41,
          "samples_seen": 41984,
          "online_loss": 0.11019480228424072,
          "train_loss": 0.10744266957044601,
          "validation_loss": 0.08950686454772949,
          "train_metric": 0.010821232572197914,
          "validation_metric": 0.009014803916215897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4124951880003209
        },
        {
          "epoch": 42,
          "updates": 42,
          "samples_seen": 43008,
          "online_loss": 0.10744266211986542,
          "train_loss": 0.10413007438182831,
          "validation_loss": 0.08649515360593796,
          "train_metric": 0.010487599298357964,
          "validation_metric": 0.008711474947631359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41841796199878445
        },
        {
          "epoch": 43,
          "updates": 43,
          "samples_seen": 44032,
          "online_loss": 0.1041300818324089,
          "train_loss": 0.10052191466093063,
          "validation_loss": 0.08330486714839935,
          "train_metric": 0.010124199092388153,
          "validation_metric": 0.008390161208808422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42436439099947165
        },
        {
          "epoch": 44,
          "updates": 44,
          "samples_seen": 45056,
          "online_loss": 0.10052193701267242,
          "train_loss": 0.09702043980360031,
          "validation_loss": 0.08034425973892212,
          "train_metric": 0.009771543554961681,
          "validation_metric": 0.008091979660093784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4299851279993163
        },
        {
          "epoch": 45,
          "updates": 45,
          "samples_seen": 46080,
          "online_loss": 0.09702044725418091,
          "train_loss": 0.09403005242347717,
          "validation_loss": 0.07797850668430328,
          "train_metric": 0.009470362216234207,
          "validation_metric": 0.007853709161281586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43565762499929406
        },
        {
          "epoch": 46,
          "updates": 46,
          "samples_seen": 47104,
          "online_loss": 0.09403005242347717,
          "train_loss": 0.09181907027959824,
          "validation_loss": 0.07641331851482391,
          "train_metric": 0.009247681125998497,
          "validation_metric": 0.007696069777011871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4417599419994076
        },
        {
          "epoch": 47,
          "updates": 47,
          "samples_seen": 48128,
          "online_loss": 0.09181907773017883,
          "train_loss": 0.09043284505605698,
          "validation_loss": 0.07563193887472153,
          "train_metric": 0.009108063764870167,
          "validation_metric": 0.0076173716224730015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44773392400020384
        },
        {
          "epoch": 48,
          "updates": 48,
          "samples_seen": 49152,
          "online_loss": 0.09043283760547638,
          "train_loss": 0.08968853205442429,
          "validation_loss": 0.07540559768676758,
          "train_metric": 0.009033100679516792,
          "validation_metric": 0.007594576105475426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4534675889990467
        },
        {
          "epoch": 49,
          "updates": 49,
          "samples_seen": 50176,
          "online_loss": 0.08968853950500488,
          "train_loss": 0.08924926072359085,
          "validation_loss": 0.07537256926298141,
          "train_metric": 0.008988859131932259,
          "validation_metric": 0.007591248955577612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45930745700025
        },
        {
          "epoch": 50,
          "updates": 50,
          "samples_seen": 51200,
          "online_loss": 0.08924926072359085,
          "train_loss": 0.08874537795782089,
          "validation_loss": 0.07515601813793182,
          "train_metric": 0.008938110433518887,
          "validation_metric": 0.007569437846541405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4650225429995771
        },
        {
          "epoch": 51,
          "updates": 51,
          "samples_seen": 52224,
          "online_loss": 0.08874538540840149,
          "train_loss": 0.08789462596178055,
          "validation_loss": 0.07447745651006699,
          "train_metric": 0.008852425031363964,
          "validation_metric": 0.007501095999032259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4710804449987336
        },
        {
          "epoch": 52,
          "updates": 52,
          "samples_seen": 53248,
          "online_loss": 0.08789463341236115,
          "train_loss": 0.08657757937908173,
          "validation_loss": 0.07322690635919571,
          "train_metric": 0.008719777688384056,
          "validation_metric": 0.0073751467280089855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4774414289986453
        },
        {
          "epoch": 53,
          "updates": 53,
          "samples_seen": 54272,
          "online_loss": 0.08657759428024292,
          "train_loss": 0.08484683930873871,
          "validation_loss": 0.0714704766869545,
          "train_metric": 0.008545463904738426,
          "validation_metric": 0.007198244798928499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4831234189987299
        },
        {
          "epoch": 54,
          "updates": 54,
          "samples_seen": 55296,
          "online_loss": 0.08484683930873871,
          "train_loss": 0.08287624269723892,
          "validation_loss": 0.06940144300460815,
          "train_metric": 0.008346992544829845,
          "validation_metric": 0.006989859510213137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4889771429989196
        },
        {
          "epoch": 55,
          "updates": 55,
          "samples_seen": 56320,
          "online_loss": 0.08287625759840012,
          "train_loss": 0.0808776393532753,
          "validation_loss": 0.06725942343473434,
          "train_metric": 0.00814569927752018,
          "validation_metric": 0.006774122826755047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.494664443000147
        },
        {
          "epoch": 56,
          "updates": 56,
          "samples_seen": 57344,
          "online_loss": 0.0808776393532753,
          "train_loss": 0.07902024686336517,
          "validation_loss": 0.065249502658844,
          "train_metric": 0.007958630099892616,
          "validation_metric": 0.0065716905519366264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5005851639998582
        },
        {
          "epoch": 57,
          "updates": 57,
          "samples_seen": 58368,
          "online_loss": 0.07902024686336517,
          "train_loss": 0.07738320529460907,
          "validation_loss": 0.06348907947540283,
          "train_metric": 0.007793753407895565,
          "validation_metric": 0.006394387222826481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5066033279999829
        },
        {
          "epoch": 58,
          "updates": 58,
          "samples_seen": 59392,
          "online_loss": 0.07738321274518967,
          "train_loss": 0.07595473527908325,
          "validation_loss": 0.061997175216674805,
          "train_metric": 0.007649882696568966,
          "validation_metric": 0.006244128104299307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5123707279999508
        },
        {
          "epoch": 59,
          "updates": 59,
          "samples_seen": 60416,
          "online_loss": 0.07595473527908325,
          "train_loss": 0.07466956228017807,
          "validation_loss": 0.06072187051177025,
          "train_metric": 0.007520444691181183,
          "validation_metric": 0.006115684751421213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5182594679990871
        },
        {
          "epoch": 60,
          "updates": 60,
          "samples_seen": 61440,
          "online_loss": 0.07466956973075867,
          "train_loss": 0.07345877587795258,
          "validation_loss": 0.05958591774106026,
          "train_metric": 0.007398498244583607,
          "validation_metric": 0.006001275032758713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5243325689989433
        },
        {
          "epoch": 61,
          "updates": 61,
          "samples_seen": 62464,
          "online_loss": 0.07345877587795258,
          "train_loss": 0.0722837820649147,
          "validation_loss": 0.05852556973695755,
          "train_metric": 0.007280157878994942,
          "validation_metric": 0.005894480738788843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.529958582999825
        },
        {
          "epoch": 62,
          "updates": 62,
          "samples_seen": 63488,
          "online_loss": 0.0722837895154953,
          "train_loss": 0.07114094495773315,
          "validation_loss": 0.057507358491420746,
          "train_metric": 0.007165056187659502,
          "validation_metric": 0.005791930481791496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.535611692999737
        },
        {
          "epoch": 63,
          "updates": 63,
          "samples_seen": 64512,
          "online_loss": 0.07114094495773315,
          "train_loss": 0.0700426921248436,
          "validation_loss": 0.056522827595472336,
          "train_metric": 0.007054443471133709,
          "validation_metric": 0.005692771635949612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5416296719995444
        },
        {
          "epoch": 64,
          "updates": 64,
          "samples_seen": 65536,
          "online_loss": 0.0700426921248436,
          "train_loss": 0.06899376958608627,
          "validation_loss": 0.055573031306266785,
          "train_metric": 0.006948800757527351,
          "validation_metric": 0.005597111769020557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5473189479998837
        },
        {
          "epoch": 65,
          "updates": 65,
          "samples_seen": 66560,
          "online_loss": 0.06899377703666687,
          "train_loss": 0.0679788887500763,
          "validation_loss": 0.05465605482459068,
          "train_metric": 0.0068465848453342915,
          "validation_metric": 0.005504758562892675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5529987949994393
        },
        {
          "epoch": 66,
          "updates": 66,
          "samples_seen": 67584,
          "online_loss": 0.06797889620065689,
          "train_loss": 0.06696764379739761,
          "validation_loss": 0.053764164447784424,
          "train_metric": 0.00674473587423563,
          "validation_metric": 0.005414930172264576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5590423329995247
        },
        {
          "epoch": 67,
          "updates": 67,
          "samples_seen": 68608,
          "online_loss": 0.06696764379739761,
          "train_loss": 0.06593137234449387,
          "validation_loss": 0.05288952589035034,
          "train_metric": 0.006640366278588772,
          "validation_metric": 0.005326839629560709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5646809259997099
        },
        {
          "epoch": 68,
          "updates": 68,
          "samples_seen": 69632,
          "online_loss": 0.06593137234449387,
          "train_loss": 0.06486069411039352,
          "validation_loss": 0.05203245207667351,
          "train_metric": 0.006532531231641769,
          "validation_metric": 0.005240518134087324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5705604489994585
        },
        {
          "epoch": 69,
          "updates": 69,
          "samples_seen": 70656,
          "online_loss": 0.06486069411039352,
          "train_loss": 0.06377396732568741,
          "validation_loss": 0.05120578035712242,
          "train_metric": 0.0064230808056890965,
          "validation_metric": 0.005157258361577988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5764881550003338
        },
        {
          "epoch": 70,
          "updates": 70,
          "samples_seen": 71680,
          "online_loss": 0.06377396732568741,
          "train_loss": 0.0627121701836586,
          "validation_loss": 0.05043190345168114,
          "train_metric": 0.006316140294075012,
          "validation_metric": 0.005079315975308418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5822406859988405
        },
        {
          "epoch": 71,
          "updates": 71,
          "samples_seen": 72704,
          "online_loss": 0.062712162733078,
          "train_loss": 0.06172265112400055,
          "validation_loss": 0.049733664840459824,
          "train_metric": 0.006216479931026697,
          "validation_metric": 0.005008992273360491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5880368660000386
        },
        {
          "epoch": 72,
          "updates": 72,
          "samples_seen": 73728,
          "online_loss": 0.06172265112400055,
          "train_loss": 0.06083950400352478,
          "validation_loss": 0.04912300035357475,
          "train_metric": 0.006127532105892897,
          "validation_metric": 0.0049474891275167465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.593675505999272
        },
        {
          "epoch": 73,
          "updates": 73,
          "samples_seen": 74752,
          "online_loss": 0.06083950772881508,
          "train_loss": 0.060069505125284195,
          "validation_loss": 0.048593468964099884,
          "train_metric": 0.006049980875104666,
          "validation_metric": 0.004894156940281391,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.599358689998553
        },
        {
          "epoch": 74,
          "updates": 74,
          "samples_seen": 75776,
          "online_loss": 0.060069505125284195,
          "train_loss": 0.05939045548439026,
          "validation_loss": 0.04812036082148552,
          "train_metric": 0.005981588736176491,
          "validation_metric": 0.004846506752073765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6052724639994267
        },
        {
          "epoch": 75,
          "updates": 75,
          "samples_seen": 76800,
          "online_loss": 0.05939045175909996,
          "train_loss": 0.0587625652551651,
          "validation_loss": 0.04766906425356865,
          "train_metric": 0.005918350536376238,
          "validation_metric": 0.004801054019480944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.611143758000253
        },
        {
          "epoch": 76,
          "updates": 76,
          "samples_seen": 77824,
          "online_loss": 0.0587625652551651,
          "train_loss": 0.05814670771360397,
          "validation_loss": 0.0472075492143631,
          "train_metric": 0.005856323521584272,
          "validation_metric": 0.004754571244120598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6167024049991596
        },
        {
          "epoch": 77,
          "updates": 77,
          "samples_seen": 78848,
          "online_loss": 0.05814670771360397,
          "train_loss": 0.05751989036798477,
          "validation_loss": 0.04671688377857208,
          "train_metric": 0.0057931928895413876,
          "validation_metric": 0.004705153405666351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.622477192999213
        },
        {
          "epoch": 78,
          "updates": 78,
          "samples_seen": 79872,
          "online_loss": 0.05751989036798477,
          "train_loss": 0.056880563497543335,
          "validation_loss": 0.046195000410079956,
          "train_metric": 0.005728801712393761,
          "validation_metric": 0.004652590956538916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6281433069998457
        },
        {
          "epoch": 79,
          "updates": 79,
          "samples_seen": 80896,
          "online_loss": 0.05688056722283363,
          "train_loss": 0.056242652237415314,
          "validation_loss": 0.045653123408555984,
          "train_metric": 0.005664553493261337,
          "validation_metric": 0.004598014988005161,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6337977029998001
        },
        {
          "epoch": 80,
          "updates": 80,
          "samples_seen": 81920,
          "online_loss": 0.056242648512125015,
          "train_loss": 0.05562387406826019,
          "validation_loss": 0.04510842263698578,
          "train_metric": 0.005602233111858368,
          "validation_metric": 0.004543154500424862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6397202850002941
        },
        {
          "epoch": 81,
          "updates": 81,
          "samples_seen": 82944,
          "online_loss": 0.055623870342969894,
          "train_loss": 0.055036187171936035,
          "validation_loss": 0.044577740132808685,
          "train_metric": 0.005543044302612543,
          "validation_metric": 0.004489706829190254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7244925509985478
        },
        {
          "epoch": 82,
          "updates": 82,
          "samples_seen": 83968,
          "online_loss": 0.055036190897226334,
          "train_loss": 0.054483119398355484,
          "validation_loss": 0.04407498240470886,
          "train_metric": 0.005487340502440929,
          "validation_metric": 0.004439070820808411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7310567479999008
        },
        {
          "epoch": 83,
          "updates": 83,
          "samples_seen": 84992,
          "online_loss": 0.054483115673065186,
          "train_loss": 0.05396325886249542,
          "validation_loss": 0.04361151158809662,
          "train_metric": 0.005434981547296047,
          "validation_metric": 0.0043923924677073956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7373568629991496
        },
        {
          "epoch": 84,
          "updates": 84,
          "samples_seen": 86016,
          "online_loss": 0.053963251411914825,
          "train_loss": 0.05347517877817154,
          "validation_loss": 0.04319652542471886,
          "train_metric": 0.005385824479162693,
          "validation_metric": 0.004350596107542515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7435698760000378
        },
        {
          "epoch": 85,
          "updates": 85,
          "samples_seen": 87040,
          "online_loss": 0.05347517877817154,
          "train_loss": 0.05301891639828682,
          "validation_loss": 0.04283514991402626,
          "train_metric": 0.005339870695024729,
          "validation_metric": 0.004314199555665255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7497537629988074
        },
        {
          "epoch": 86,
          "updates": 86,
          "samples_seen": 88064,
          "online_loss": 0.05301891267299652,
          "train_loss": 0.05259322375059128,
          "validation_loss": 0.04252493381500244,
          "train_metric": 0.005296996794641018,
          "validation_metric": 0.004282955080270767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7562364240002353
        },
        {
          "epoch": 87,
          "updates": 87,
          "samples_seen": 89088,
          "online_loss": 0.05259322747588158,
          "train_loss": 0.05219176411628723,
          "validation_loss": 0.04225369915366173,
          "train_metric": 0.005256563425064087,
          "validation_metric": 0.004255637992173433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7626481409988628
        },
        {
          "epoch": 88,
          "updates": 88,
          "samples_seen": 90112,
          "online_loss": 0.05219176411628723,
          "train_loss": 0.05180253088474274,
          "validation_loss": 0.04200197011232376,
          "train_metric": 0.00521736079826951,
          "validation_metric": 0.004230284132063389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7686107190002076
        },
        {
          "epoch": 89,
          "updates": 89,
          "samples_seen": 91136,
          "online_loss": 0.05180253088474274,
          "train_loss": 0.05141232907772064,
          "validation_loss": 0.04175002872943878,
          "train_metric": 0.005178062245249748,
          "validation_metric": 0.004204909782856703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.775259254998673
        },
        {
          "epoch": 90,
          "updates": 90,
          "samples_seen": 92160,
          "online_loss": 0.05141233652830124,
          "train_loss": 0.05101370811462402,
          "validation_loss": 0.0414862260222435,
          "train_metric": 0.00513791386038065,
          "validation_metric": 0.004178340546786785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7810482049990242
        },
        {
          "epoch": 91,
          "updates": 91,
          "samples_seen": 93184,
          "online_loss": 0.05101371183991432,
          "train_loss": 0.05060913413763046,
          "validation_loss": 0.04121150076389313,
          "train_metric": 0.005097166169434786,
          "validation_metric": 0.004150671418756247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7874537349998718
        },
        {
          "epoch": 92,
          "updates": 92,
          "samples_seen": 94208,
          "online_loss": 0.05060913786292076,
          "train_loss": 0.05020901560783386,
          "validation_loss": 0.04093712940812111,
          "train_metric": 0.005056868307292461,
          "validation_metric": 0.004123037680983543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7932957439988968
        },
        {
          "epoch": 93,
          "updates": 93,
          "samples_seen": 95232,
          "online_loss": 0.05020901560783386,
          "train_loss": 0.04982450604438782,
          "validation_loss": 0.04067692905664444,
          "train_metric": 0.005018142051994801,
          "validation_metric": 0.0040968311950564384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.799327701999573
        },
        {
          "epoch": 94,
          "updates": 94,
          "samples_seen": 96256,
          "online_loss": 0.049824509769678116,
          "train_loss": 0.04946020990610123,
          "validation_loss": 0.04043892025947571,
          "train_metric": 0.004981451202183962,
          "validation_metric": 0.004072860348969698,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8061388959995384
        },
        {
          "epoch": 95,
          "updates": 95,
          "samples_seen": 97280,
          "online_loss": 0.049460213631391525,
          "train_loss": 0.04911188781261444,
          "validation_loss": 0.04022151231765747,
          "train_metric": 0.004946369677782059,
          "validation_metric": 0.004050963558256626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8121612929990079
        },
        {
          "epoch": 96,
          "updates": 96,
          "samples_seen": 98304,
          "online_loss": 0.04911188781261444,
          "train_loss": 0.04877028986811638,
          "validation_loss": 0.04001576825976372,
          "train_metric": 0.004911964759230614,
          "validation_metric": 0.004030242096632719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.817991568999787
        },
        {
          "epoch": 97,
          "updates": 97,
          "samples_seen": 99328,
          "online_loss": 0.04877028986811638,
          "train_loss": 0.0484275296330452,
          "validation_loss": 0.03981102257966995,
          "train_metric": 0.004877443425357342,
          "validation_metric": 0.004009620286524296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.82401724599913
        },
        {
          "epoch": 98,
          "updates": 98,
          "samples_seen": 100352,
          "online_loss": 0.0484275296330452,
          "train_loss": 0.04808063805103302,
          "validation_loss": 0.03959909453988075,
          "train_metric": 0.004842505790293217,
          "validation_metric": 0.003988275304436684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8298583139985567
        },
        {
          "epoch": 99,
          "updates": 99,
          "samples_seen": 101376,
          "online_loss": 0.04808063805103302,
          "train_loss": 0.04773023724555969,
          "validation_loss": 0.03937488794326782,
          "train_metric": 0.004807214252650738,
          "validation_metric": 0.003965694457292557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8356950449997385
        },
        {
          "epoch": 100,
          "updates": 100,
          "samples_seen": 102400,
          "online_loss": 0.047730229794979095,
          "train_loss": 0.047376878559589386,
          "validation_loss": 0.03913471847772598,
          "train_metric": 0.0047716256231069565,
          "validation_metric": 0.003941505216062069,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8419564139985596
        },
        {
          "epoch": 101,
          "updates": 101,
          "samples_seen": 103424,
          "online_loss": 0.047376878559589386,
          "train_loss": 0.04701949283480644,
          "validation_loss": 0.038875531405210495,
          "train_metric": 0.004735630936920643,
          "validation_metric": 0.003915401175618172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8475929079995694
        },
        {
          "epoch": 102,
          "updates": 102,
          "samples_seen": 104448,
          "online_loss": 0.04701949283480644,
          "train_loss": 0.04665706306695938,
          "validation_loss": 0.038596559315919876,
          "train_metric": 0.004699129145592451,
          "validation_metric": 0.0038873038720339537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8530508169988025
        },
        {
          "epoch": 103,
          "updates": 103,
          "samples_seen": 105472,
          "online_loss": 0.04665707051753998,
          "train_loss": 0.04629116132855415,
          "validation_loss": 0.038301218301057816,
          "train_metric": 0.004662275779992342,
          "validation_metric": 0.003857559058815241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.859021070999006
        },
        {
          "epoch": 104,
          "updates": 104,
          "samples_seen": 106496,
          "online_loss": 0.046291157603263855,
          "train_loss": 0.04592559486627579,
          "validation_loss": 0.037996985018253326,
          "train_metric": 0.004625457338988781,
          "validation_metric": 0.0038269171491265297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8646850299992366
        },
        {
          "epoch": 105,
          "updates": 105,
          "samples_seen": 107520,
          "online_loss": 0.04592559114098549,
          "train_loss": 0.045563213527202606,
          "validation_loss": 0.03769233822822571,
          "train_metric": 0.00458896066993475,
          "validation_metric": 0.0037962342612445354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8702041929991537
        },
        {
          "epoch": 106,
          "updates": 106,
          "samples_seen": 108544,
          "online_loss": 0.045563217252492905,
          "train_loss": 0.04520297423005104,
          "validation_loss": 0.03739339858293533,
          "train_metric": 0.004552678670734167,
          "validation_metric": 0.003766125999391079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8758984609994513
        },
        {
          "epoch": 107,
          "updates": 107,
          "samples_seen": 109568,
          "online_loss": 0.04520297795534134,
          "train_loss": 0.04484056681394577,
          "validation_loss": 0.0371025875210762,
          "train_metric": 0.004516178276389837,
          "validation_metric": 0.0037368368357419968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8816038129989465
        },
        {
          "epoch": 108,
          "updates": 108,
          "samples_seen": 110592,
          "online_loss": 0.04484056681394577,
          "train_loss": 0.04447215795516968,
          "validation_loss": 0.036819763481616974,
          "train_metric": 0.004479072522372007,
          "validation_metric": 0.003708351869136095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.887065912998878
        },
        {
          "epoch": 109,
          "updates": 109,
          "samples_seen": 111616,
          "online_loss": 0.04447215795516968,
          "train_loss": 0.04409727081656456,
          "validation_loss": 0.036543626338243484,
          "train_metric": 0.004441315773874521,
          "validation_metric": 0.0036805400159209967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.893135276999601
        },
        {
          "epoch": 110,
          "updates": 110,
          "samples_seen": 112640,
          "online_loss": 0.04409727454185486,
          "train_loss": 0.04371815547347069,
          "validation_loss": 0.0362718440592289,
          "train_metric": 0.004403132479637861,
          "validation_metric": 0.003653166815638542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.898919074999867
        },
        {
          "epoch": 111,
          "updates": 111,
          "samples_seen": 113664,
          "online_loss": 0.04371815174818039,
          "train_loss": 0.043336883187294006,
          "validation_loss": 0.03600064292550087,
          "train_metric": 0.004364732187241316,
          "validation_metric": 0.0036258529871702194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9047310999994806
        },
        {
          "epoch": 112,
          "updates": 112,
          "samples_seen": 114688,
          "online_loss": 0.043336883187294006,
          "train_loss": 0.04295378923416138,
          "validation_loss": 0.03572562709450722,
          "train_metric": 0.004326148424297571,
          "validation_metric": 0.003598154289647937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9110159499996371
        },
        {
          "epoch": 113,
          "updates": 113,
          "samples_seen": 115712,
          "online_loss": 0.042953792959451675,
          "train_loss": 0.04256860166788101,
          "validation_loss": 0.035443976521492004,
          "train_metric": 0.004287353716790676,
          "validation_metric": 0.0035697873681783676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.916623418999734
        },
        {
          "epoch": 114,
          "updates": 114,
          "samples_seen": 116736,
          "online_loss": 0.042568594217300415,
          "train_loss": 0.04218215495347977,
          "validation_loss": 0.03515595942735672,
          "train_metric": 0.004248432349413633,
          "validation_metric": 0.0035407794639468193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9227160859991272
        },
        {
          "epoch": 115,
          "updates": 115,
          "samples_seen": 117760,
          "online_loss": 0.04218215122818947,
          "train_loss": 0.04179615527391434,
          "validation_loss": 0.03486410155892372,
          "train_metric": 0.004209556151181459,
          "validation_metric": 0.003511384828016162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9284932409991598
        },
        {
          "epoch": 116,
          "updates": 116,
          "samples_seen": 118784,
          "online_loss": 0.041796162724494934,
          "train_loss": 0.04141122102737427,
          "validation_loss": 0.03457057476043701,
          "train_metric": 0.004170786589384079,
          "validation_metric": 0.003481821622699499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9341863099998591
        },
        {
          "epoch": 117,
          "updates": 117,
          "samples_seen": 119808,
          "online_loss": 0.04141122102737427,
          "train_loss": 0.04102613404393196,
          "validation_loss": 0.034275852143764496,
          "train_metric": 0.004132001660764217,
          "validation_metric": 0.0034521385096013546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9404038809989288
        },
        {
          "epoch": 118,
          "updates": 118,
          "samples_seen": 120832,
          "online_loss": 0.04102613776922226,
          "train_loss": 0.04063945636153221,
          "validation_loss": 0.0339793860912323,
          "train_metric": 0.004093057010322809,
          "validation_metric": 0.003422279842197895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.02414036799928
        },
        {
          "epoch": 119,
          "updates": 119,
          "samples_seen": 121856,
          "online_loss": 0.04063945636153221,
          "train_loss": 0.04025142639875412,
          "validation_loss": 0.03368090093135834,
          "train_metric": 0.004053976386785507,
          "validation_metric": 0.0033922172151505947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0301272249998874
        },
        {
          "epoch": 120,
          "updates": 120,
          "samples_seen": 122880,
          "online_loss": 0.04025142639875412,
          "train_loss": 0.03986368328332901,
          "validation_loss": 0.03338056802749634,
          "train_metric": 0.004014924168586731,
          "validation_metric": 0.0033619687892496586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0357021849995363
        },
        {
          "epoch": 121,
          "updates": 121,
          "samples_seen": 123904,
          "online_loss": 0.03986368328332901,
          "train_loss": 0.039477504789829254,
          "validation_loss": 0.03307804837822914,
          "train_metric": 0.003976029343903065,
          "validation_metric": 0.0033314996398985386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0415812609990098
        },
        {
          "epoch": 122,
          "updates": 122,
          "samples_seen": 124928,
          "online_loss": 0.039477501064538956,
          "train_loss": 0.039093002676963806,
          "validation_loss": 0.03277259320020676,
          "train_metric": 0.003937304485589266,
          "validation_metric": 0.003300735494121909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0473579709996557
        },
        {
          "epoch": 123,
          "updates": 123,
          "samples_seen": 125952,
          "online_loss": 0.039093006402254105,
          "train_loss": 0.038710277527570724,
          "validation_loss": 0.032464414834976196,
          "train_metric": 0.0038987575098872185,
          "validation_metric": 0.0032696968410164118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0529667669998162
        },
        {
          "epoch": 124,
          "updates": 124,
          "samples_seen": 126976,
          "online_loss": 0.03871028125286102,
          "train_loss": 0.038330137729644775,
          "validation_loss": 0.03215558081865311,
          "train_metric": 0.00386047107167542,
          "validation_metric": 0.003238592529669404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.058956312999726
        },
        {
          "epoch": 125,
          "updates": 125,
          "samples_seen": 128000,
          "online_loss": 0.038330137729644775,
          "train_loss": 0.0379534550011158,
          "validation_loss": 0.031849201768636703,
          "train_metric": 0.0038225334137678146,
          "validation_metric": 0.0032077347859740257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.064731201999166
        },
        {
          "epoch": 126,
          "updates": 126,
          "samples_seen": 129024,
          "online_loss": 0.0379534587264061,
          "train_loss": 0.03758011758327484,
          "validation_loss": 0.03154782950878143,
          "train_metric": 0.0037849314976483583,
          "validation_metric": 0.0031773820519447327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0703296679985215
        },
        {
          "epoch": 127,
          "updates": 127,
          "samples_seen": 130048,
          "online_loss": 0.03758011758327484,
          "train_loss": 0.037209488451480865,
          "validation_loss": 0.03125258535146713,
          "train_metric": 0.0037476036231964827,
          "validation_metric": 0.003147645853459835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.07641290599895
        },
        {
          "epoch": 128,
          "updates": 128,
          "samples_seen": 131072,
          "online_loss": 0.037209488451480865,
          "train_loss": 0.03684167191386223,
          "validation_loss": 0.030963294208049774,
          "train_metric": 0.0037105584051460028,
          "validation_metric": 0.00311850942671299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0821382510002877
        },
        {
          "epoch": 129,
          "updates": 129,
          "samples_seen": 132096,
          "online_loss": 0.03684167563915253,
          "train_loss": 0.036477573215961456,
          "validation_loss": 0.03067859075963497,
          "train_metric": 0.0036738873459398746,
          "validation_metric": 0.003089835401624441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0878094459985732
        },
        {
          "epoch": 130,
          "updates": 130,
          "samples_seen": 133120,
          "online_loss": 0.036477576941251755,
          "train_loss": 0.036117807030677795,
          "validation_loss": 0.030396156013011932,
          "train_metric": 0.003637653309851885,
          "validation_metric": 0.0030613893177360296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0937309989985806
        },
        {
          "epoch": 131,
          "updates": 131,
          "samples_seen": 134144,
          "online_loss": 0.036117807030677795,
          "train_loss": 0.03576233983039856,
          "validation_loss": 0.03011367656290531,
          "train_metric": 0.0036018523387610912,
          "validation_metric": 0.003032939275726676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0994930119995843
        },
        {
          "epoch": 132,
          "updates": 132,
          "samples_seen": 135168,
          "online_loss": 0.03576234355568886,
          "train_loss": 0.03541123867034912,
          "validation_loss": 0.029830317944288254,
          "train_metric": 0.003566490253433585,
          "validation_metric": 0.0030044002924114466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1053945099993143
        },
        {
          "epoch": 133,
          "updates": 133,
          "samples_seen": 136192,
          "online_loss": 0.03541124239563942,
          "train_loss": 0.03506504371762276,
          "validation_loss": 0.029547058045864105,
          "train_metric": 0.0035316227003932,
          "validation_metric": 0.0029758713208138943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1111206089990446
        },
        {
          "epoch": 134,
          "updates": 134,
          "samples_seen": 137216,
          "online_loss": 0.035065047442913055,
          "train_loss": 0.03472407907247543,
          "validation_loss": 0.029265543445944786,
          "train_metric": 0.0034972820430994034,
          "validation_metric": 0.0029475183691829443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1167059109993716
        },
        {
          "epoch": 135,
          "updates": 135,
          "samples_seen": 138240,
          "online_loss": 0.03472407907247543,
          "train_loss": 0.034388091415166855,
          "validation_loss": 0.028986889868974686,
          "train_metric": 0.0034634429030120373,
          "validation_metric": 0.002919453661888838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1223940049985686
        },
        {
          "epoch": 136,
          "updates": 136,
          "samples_seen": 139264,
          "online_loss": 0.034388087689876556,
          "train_loss": 0.03405686095356941,
          "validation_loss": 0.028711378574371338,
          "train_metric": 0.0034300819970667362,
          "validation_metric": 0.002891704672947526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1280674849986099
        },
        {
          "epoch": 137,
          "updates": 137,
          "samples_seen": 140288,
          "online_loss": 0.034056857228279114,
          "train_loss": 0.033730652183294296,
          "validation_loss": 0.028438625857234,
          "train_metric": 0.0033972279634326696,
          "validation_metric": 0.0028642339166253805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.133708362998732
        },
        {
          "epoch": 138,
          "updates": 138,
          "samples_seen": 141312,
          "online_loss": 0.033730655908584595,
          "train_loss": 0.03340977430343628,
          "validation_loss": 0.028167687356472015,
          "train_metric": 0.0033649103716015816,
          "validation_metric": 0.0028369459323585033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1396654660002241
        },
        {
          "epoch": 139,
          "updates": 139,
          "samples_seen": 142336,
          "online_loss": 0.03340977802872658,
          "train_loss": 0.03309409320354462,
          "validation_loss": 0.027897637337446213,
          "train_metric": 0.0033331154845654964,
          "validation_metric": 0.0028097478207200766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1453326020000532
        },
        {
          "epoch": 140,
          "updates": 140,
          "samples_seen": 143360,
          "online_loss": 0.03309409320354462,
          "train_loss": 0.03278348967432976,
          "validation_loss": 0.027628464624285698,
          "train_metric": 0.00330183282494545,
          "validation_metric": 0.0027826379518955946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1510767999989184
        },
        {
          "epoch": 141,
          "updates": 141,
          "samples_seen": 144384,
          "online_loss": 0.03278348594903946,
          "train_loss": 0.03247813135385513,
          "validation_loss": 0.027361474931240082,
          "train_metric": 0.00327107822522521,
          "validation_metric": 0.0027557476423680782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1571077829994465
        },
        {
          "epoch": 142,
          "updates": 142,
          "samples_seen": 145408,
          "online_loss": 0.03247813135385513,
          "train_loss": 0.03217807412147522,
          "validation_loss": 0.02709847502410412,
          "train_metric": 0.003240857506170869,
          "validation_metric": 0.002729259431362152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1627952309991088
        },
        {
          "epoch": 143,
          "updates": 143,
          "samples_seen": 146432,
          "online_loss": 0.03217807412147522,
          "train_loss": 0.03188301622867584,
          "validation_loss": 0.02684067189693451,
          "train_metric": 0.003211140399798751,
          "validation_metric": 0.002703294390812516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1683668049990956
        },
        {
          "epoch": 144,
          "updates": 144,
          "samples_seen": 147456,
          "online_loss": 0.031883012503385544,
          "train_loss": 0.03159274160861969,
          "validation_loss": 0.026588251814246178,
          "train_metric": 0.0031819052528589964,
          "validation_metric": 0.0026778713800013065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1739536770000996
        },
        {
          "epoch": 145,
          "updates": 145,
          "samples_seen": 148480,
          "online_loss": 0.03159274160861969,
          "train_loss": 0.03130728751420975,
          "validation_loss": 0.026340413838624954,
          "train_metric": 0.0031531555578112602,
          "validation_metric": 0.002652909839525819,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1797060369990504
        },
        {
          "epoch": 146,
          "updates": 146,
          "samples_seen": 149504,
          "online_loss": 0.031307291239500046,
          "train_loss": 0.031026573851704597,
          "validation_loss": 0.026095816865563393,
          "train_metric": 0.003124882932752371,
          "validation_metric": 0.002628275426104665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.185425239998949
        },
        {
          "epoch": 147,
          "updates": 147,
          "samples_seen": 150528,
          "online_loss": 0.031026573851704597,
          "train_loss": 0.030750317499041557,
          "validation_loss": 0.025853363797068596,
          "train_metric": 0.003097059205174446,
          "validation_metric": 0.0026038559153676033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1913592660002905
        },
        {
          "epoch": 148,
          "updates": 148,
          "samples_seen": 151552,
          "online_loss": 0.030750319361686707,
          "train_loss": 0.03047836199402809,
          "validation_loss": 0.02561296336352825,
          "train_metric": 0.003069669473916292,
          "validation_metric": 0.0025796438567340374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1968049779989087
        },
        {
          "epoch": 149,
          "updates": 149,
          "samples_seen": 152576,
          "online_loss": 0.03047836199402809,
          "train_loss": 0.03021063841879368,
          "validation_loss": 0.025375360623002052,
          "train_metric": 0.0030427046585828066,
          "validation_metric": 0.0025557130575180054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2024284009985422
        },
        {
          "epoch": 150,
          "updates": 150,
          "samples_seen": 153600,
          "online_loss": 0.030210640281438828,
          "train_loss": 0.029946928843855858,
          "validation_loss": 0.02514127641916275,
          "train_metric": 0.0030161449685692787,
          "validation_metric": 0.0025321373250335455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.208804252999471
        },
        {
          "epoch": 151,
          "updates": 151,
          "samples_seen": 154624,
          "online_loss": 0.029946930706501007,
          "train_loss": 0.02968693897128105,
          "validation_loss": 0.02491086721420288,
          "train_metric": 0.00298995990306139,
          "validation_metric": 0.0025089315604418516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.214667180998731
        },
        {
          "epoch": 152,
          "updates": 152,
          "samples_seen": 155648,
          "online_loss": 0.029686937108635902,
          "train_loss": 0.029430503025650978,
          "validation_loss": 0.024683713912963867,
          "train_metric": 0.002964132698252797,
          "validation_metric": 0.0024860536213964224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2203949699996883
        },
        {
          "epoch": 153,
          "updates": 153,
          "samples_seen": 156672,
          "online_loss": 0.029430504888296127,
          "train_loss": 0.029177503660321236,
          "validation_loss": 0.02445906586945057,
          "train_metric": 0.002938651479780674,
          "validation_metric": 0.002463427372276783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2262162609986262
        },
        {
          "epoch": 154,
          "updates": 154,
          "samples_seen": 157696,
          "online_loss": 0.029177501797676086,
          "train_loss": 0.028927674517035484,
          "validation_loss": 0.024236397817730904,
          "train_metric": 0.0029134894721210003,
          "validation_metric": 0.0024410011246800423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2319338469987997
        },
        {
          "epoch": 155,
          "updates": 155,
          "samples_seen": 158720,
          "online_loss": 0.028927674517035484,
          "train_loss": 0.028680797666311264,
          "validation_loss": 0.024015776813030243,
          "train_metric": 0.0028886250220239162,
          "validation_metric": 0.0024187813978642225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.237479944998995
        },
        {
          "epoch": 156,
          "updates": 156,
          "samples_seen": 159744,
          "online_loss": 0.028680801391601562,
          "train_loss": 0.02843674272298813,
          "validation_loss": 0.02379787340760231,
          "train_metric": 0.0028640448581427336,
          "validation_metric": 0.0023968347813934088,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3214972610003315
        },
        {
          "epoch": 157,
          "updates": 157,
          "samples_seen": 160768,
          "online_loss": 0.028436744585633278,
          "train_loss": 0.028195317834615707,
          "validation_loss": 0.023583384230732918,
          "train_metric": 0.0028397291898727417,
          "validation_metric": 0.002375232055783272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.327662314999543
        },
        {
          "epoch": 158,
          "updates": 158,
          "samples_seen": 161792,
          "online_loss": 0.028195319697260857,
          "train_loss": 0.027956288307905197,
          "validation_loss": 0.023372555151581764,
          "train_metric": 0.002815654966980219,
          "validation_metric": 0.0023539981339126825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3335844739995082
        },
        {
          "epoch": 159,
          "updates": 159,
          "samples_seen": 162816,
          "online_loss": 0.027956292033195496,
          "train_loss": 0.027719510719180107,
          "validation_loss": 0.02316519059240818,
          "train_metric": 0.0027918072883039713,
          "validation_metric": 0.0023331132251769304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3395163469995168
        },
        {
          "epoch": 160,
          "updates": 160,
          "samples_seen": 163840,
          "online_loss": 0.027719508856534958,
          "train_loss": 0.02748485654592514,
          "validation_loss": 0.022960782051086426,
          "train_metric": 0.002768174046650529,
          "validation_metric": 0.002312525874003768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.345235374999902
        },
        {
          "epoch": 161,
          "updates": 161,
          "samples_seen": 164864,
          "online_loss": 0.02748485654592514,
          "train_loss": 0.027252160012722015,
          "validation_loss": 0.022758863866329193,
          "train_metric": 0.0027447377797216177,
          "validation_metric": 0.0022921895142644644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.350968799999464
        },
        {
          "epoch": 162,
          "updates": 162,
          "samples_seen": 165888,
          "online_loss": 0.027252160012722015,
          "train_loss": 0.02702128142118454,
          "validation_loss": 0.02255934290587902,
          "train_metric": 0.0027214845176786184,
          "validation_metric": 0.00227209459990263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3570449059989187
        },
        {
          "epoch": 163,
          "updates": 163,
          "samples_seen": 166912,
          "online_loss": 0.02702128142118454,
          "train_loss": 0.026792125776410103,
          "validation_loss": 0.022362379357218742,
          "train_metric": 0.0026984047144651413,
          "validation_metric": 0.002252256963402033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3628449619991443
        },
        {
          "epoch": 164,
          "updates": 164,
          "samples_seen": 167936,
          "online_loss": 0.026792127639055252,
          "train_loss": 0.026564572006464005,
          "validation_loss": 0.022168081253767014,
          "train_metric": 0.0026754867285490036,
          "validation_metric": 0.0022326880134642124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3686605749990122
        },
        {
          "epoch": 165,
          "updates": 165,
          "samples_seen": 168960,
          "online_loss": 0.026564572006464005,
          "train_loss": 0.02633850835263729,
          "validation_loss": 0.021976284682750702,
          "train_metric": 0.0026527182199060917,
          "validation_metric": 0.0022133709862828255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3746247950002726
        },
        {
          "epoch": 166,
          "updates": 166,
          "samples_seen": 169984,
          "online_loss": 0.02633850835263729,
          "train_loss": 0.026113850995898247,
          "validation_loss": 0.021786633878946304,
          "train_metric": 0.0026300917379558086,
          "validation_metric": 0.0021942697931081057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.380483369999638
        },
        {
          "epoch": 167,
          "updates": 167,
          "samples_seen": 171008,
          "online_loss": 0.026113852858543396,
          "train_loss": 0.025890545919537544,
          "validation_loss": 0.021598778665065765,
          "train_metric": 0.0026076012291014194,
          "validation_metric": 0.002175349975004792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3863779559997056
        },
        {
          "epoch": 168,
          "updates": 168,
          "samples_seen": 172032,
          "online_loss": 0.025890545919537544,
          "train_loss": 0.02566850185394287,
          "validation_loss": 0.021412523463368416,
          "train_metric": 0.0025852371472865343,
          "validation_metric": 0.0021565908100456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3922739009994984
        },
        {
          "epoch": 169,
          "updates": 169,
          "samples_seen": 173056,
          "online_loss": 0.025668499991297722,
          "train_loss": 0.02544766664505005,
          "validation_loss": 0.021227912977337837,
          "train_metric": 0.0025629953015595675,
          "validation_metric": 0.0021379978861659765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3981870709994837
        },
        {
          "epoch": 170,
          "updates": 170,
          "samples_seen": 174080,
          "online_loss": 0.02544766291975975,
          "train_loss": 0.025228001177310944,
          "validation_loss": 0.02104506827890873,
          "train_metric": 0.0025408719666302204,
          "validation_metric": 0.0021195823792368174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4040571019995696
        },
        {
          "epoch": 171,
          "updates": 171,
          "samples_seen": 175104,
          "online_loss": 0.025228003039956093,
          "train_loss": 0.02500946819782257,
          "validation_loss": 0.020864052698016167,
          "train_metric": 0.002518862020224333,
          "validation_metric": 0.0021013508085161448,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4099557749996166
        },
        {
          "epoch": 172,
          "updates": 172,
          "samples_seen": 176128,
          "online_loss": 0.02500947006046772,
          "train_loss": 0.02479201927781105,
          "validation_loss": 0.020684804767370224,
          "train_metric": 0.002496961737051606,
          "validation_metric": 0.0020832980517297983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4155867760000547
        },
        {
          "epoch": 173,
          "updates": 173,
          "samples_seen": 177152,
          "online_loss": 0.02479202300310135,
          "train_loss": 0.024575665593147278,
          "validation_loss": 0.02050723321735859,
          "train_metric": 0.0024751711171120405,
          "validation_metric": 0.0020654136314988136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4214197379988036
        },
        {
          "epoch": 174,
          "updates": 174,
          "samples_seen": 178176,
          "online_loss": 0.024575665593147278,
          "train_loss": 0.024360371753573418,
          "validation_loss": 0.020331302657723427,
          "train_metric": 0.002453487366437912,
          "validation_metric": 0.0020476942881941795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4270444199992198
        },
        {
          "epoch": 175,
          "updates": 175,
          "samples_seen": 179200,
          "online_loss": 0.024360371753573418,
          "train_loss": 0.02414610981941223,
          "validation_loss": 0.020156988874077797,
          "train_metric": 0.0024319079238921404,
          "validation_metric": 0.002030138159170747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4327644519999012
        },
        {
          "epoch": 176,
          "updates": 176,
          "samples_seen": 180224,
          "online_loss": 0.02414610981941223,
          "train_loss": 0.023932896554470062,
          "validation_loss": 0.019984275102615356,
          "train_metric": 0.0024104334879666567,
          "validation_metric": 0.002012742916122079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.438598655999158
        },
        {
          "epoch": 177,
          "updates": 177,
          "samples_seen": 181248,
          "online_loss": 0.023932896554470062,
          "train_loss": 0.023720718920230865,
          "validation_loss": 0.019813038408756256,
          "train_metric": 0.0023890638258308172,
          "validation_metric": 0.001995496451854706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4444181409999146
        },
        {
          "epoch": 178,
          "updates": 178,
          "samples_seen": 182272,
          "online_loss": 0.023720718920230865,
          "train_loss": 0.023509567603468895,
          "validation_loss": 0.019643103703856468,
          "train_metric": 0.0023677973076701164,
          "validation_metric": 0.001978381536900997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4500303939985315
        },
        {
          "epoch": 179,
          "updates": 179,
          "samples_seen": 183296,
          "online_loss": 0.023509565740823746,
          "train_loss": 0.023299453780055046,
          "validation_loss": 0.01947432942688465,
          "train_metric": 0.0023466357961297035,
          "validation_metric": 0.0019613830372691154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4561827429988625
        },
        {
          "epoch": 180,
          "updates": 180,
          "samples_seen": 184320,
          "online_loss": 0.023299451917409897,
          "train_loss": 0.023090394213795662,
          "validation_loss": 0.019306605681777,
          "train_metric": 0.0023255799897015095,
          "validation_metric": 0.00194449070841074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4619339759992727
        },
        {
          "epoch": 181,
          "updates": 181,
          "samples_seen": 185344,
          "online_loss": 0.023090394213795662,
          "train_loss": 0.022882379591464996,
          "validation_loss": 0.019139841198921204,
          "train_metric": 0.0023046296555548906,
          "validation_metric": 0.0019276948878541589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4677299710001535
        },
        {
          "epoch": 182,
          "updates": 182,
          "samples_seen": 186368,
          "online_loss": 0.022882383316755295,
          "train_loss": 0.022675421088933945,
          "validation_loss": 0.0189739391207695,
          "train_metric": 0.002283785492181778,
          "validation_metric": 0.001910985796712339,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.473840217999168
        },
        {
          "epoch": 183,
          "updates": 183,
          "samples_seen": 187392,
          "online_loss": 0.022675422951579094,
          "train_loss": 0.02246953547000885,
          "validation_loss": 0.018808746710419655,
          "train_metric": 0.0022630493622273207,
          "validation_metric": 0.0018943483009934425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4794834599997557
        },
        {
          "epoch": 184,
          "updates": 184,
          "samples_seen": 188416,
          "online_loss": 0.02246953547000885,
          "train_loss": 0.02226472645998001,
          "validation_loss": 0.01864410936832428,
          "train_metric": 0.002242421731352806,
          "validation_metric": 0.0018777662189677358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4851152249993902
        },
        {
          "epoch": 185,
          "updates": 185,
          "samples_seen": 189440,
          "online_loss": 0.022264724597334862,
          "train_loss": 0.022061005234718323,
          "validation_loss": 0.018479984253644943,
          "train_metric": 0.0022219037637114525,
          "validation_metric": 0.0018612362910062075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4909702779987128
        },
        {
          "epoch": 186,
          "updates": 186,
          "samples_seen": 190464,
          "online_loss": 0.022061007097363472,
          "train_loss": 0.021858401596546173,
          "validation_loss": 0.01831640861928463,
          "train_metric": 0.0022014982532709837,
          "validation_metric": 0.0018447617767378688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4965821569985565
        },
        {
          "epoch": 187,
          "updates": 187,
          "samples_seen": 191488,
          "online_loss": 0.021858401596546173,
          "train_loss": 0.02165691927075386,
          "validation_loss": 0.018153509125113487,
          "train_metric": 0.0021812058985233307,
          "validation_metric": 0.0018283551326021552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.502259849999973
        },
        {
          "epoch": 188,
          "updates": 188,
          "samples_seen": 192512,
          "online_loss": 0.02165692113339901,
          "train_loss": 0.021456578746438026,
          "validation_loss": 0.01799130067229271,
          "train_metric": 0.0021610280964523554,
          "validation_metric": 0.0018120178719982505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.508425537000221
        },
        {
          "epoch": 189,
          "updates": 189,
          "samples_seen": 193536,
          "online_loss": 0.021456578746438026,
          "train_loss": 0.02125740237534046,
          "validation_loss": 0.017829734832048416,
          "train_metric": 0.002140967408195138,
          "validation_metric": 0.0017957454547286034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5141117719995236
        },
        {
          "epoch": 190,
          "updates": 190,
          "samples_seen": 194560,
          "online_loss": 0.021257400512695312,
          "train_loss": 0.021059410646557808,
          "validation_loss": 0.017668716609477997,
          "train_metric": 0.0021210266277194023,
          "validation_metric": 0.0017795284511521459,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.519751459998588
        },
        {
          "epoch": 191,
          "updates": 191,
          "samples_seen": 195584,
          "online_loss": 0.021059412509202957,
          "train_loss": 0.020862611010670662,
          "validation_loss": 0.017508231103420258,
          "train_metric": 0.0021012055221945047,
          "validation_metric": 0.001763364765793085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5258103859996481
        },
        {
          "epoch": 192,
          "updates": 192,
          "samples_seen": 196608,
          "online_loss": 0.020862609148025513,
          "train_loss": 0.020667031407356262,
          "validation_loss": 0.01734834909439087,
          "train_metric": 0.0020815078169107437,
          "validation_metric": 0.0017472621984779835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5315646979997837
        },
        {
          "epoch": 193,
          "updates": 193,
          "samples_seen": 197632,
          "online_loss": 0.02066703327000141,
          "train_loss": 0.020472707226872444,
          "validation_loss": 0.01718917302787304,
          "train_metric": 0.0020619358401745558,
          "validation_metric": 0.0017312304116785526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6201355319990398
        },
        {
          "epoch": 194,
          "updates": 194,
          "samples_seen": 198656,
          "online_loss": 0.020472705364227295,
          "train_loss": 0.020279642194509506,
          "validation_loss": 0.01703072339296341,
          "train_metric": 0.002042490988969803,
          "validation_metric": 0.0017152719665318727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6264122630000202
        },
        {
          "epoch": 195,
          "updates": 195,
          "samples_seen": 199680,
          "online_loss": 0.020279640331864357,
          "train_loss": 0.02008787728846073,
          "validation_loss": 0.016872890293598175,
          "train_metric": 0.002023177221417427,
          "validation_metric": 0.0016993754543364048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6319413319997693
        },
        {
          "epoch": 196,
          "updates": 196,
          "samples_seen": 200704,
          "online_loss": 0.020087875425815582,
          "train_loss": 0.019897427409887314,
          "validation_loss": 0.016715584322810173,
          "train_metric": 0.0020039959345012903,
          "validation_metric": 0.0016835323767736554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6375917379991733
        },
        {
          "epoch": 197,
          "updates": 197,
          "samples_seen": 201728,
          "online_loss": 0.019897425547242165,
          "train_loss": 0.01970832422375679,
          "validation_loss": 0.01655876822769642,
          "train_metric": 0.001984950387850404,
          "validation_metric": 0.001667738426476717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6434823399995366
        },
        {
          "epoch": 198,
          "updates": 198,
          "samples_seen": 202752,
          "online_loss": 0.01970832608640194,
          "train_loss": 0.0195205919444561,
          "validation_loss": 0.01640251837670803,
          "train_metric": 0.0019660424441099167,
          "validation_metric": 0.001652001403272152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.649131738999131
        },
        {
          "epoch": 199,
          "updates": 199,
          "samples_seen": 203776,
          "online_loss": 0.01952059008181095,
          "train_loss": 0.019334254786372185,
          "validation_loss": 0.016246991232037544,
          "train_metric": 0.0019472752464935184,
          "validation_metric": 0.0016363373724743724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.654747057998975
        },
        {
          "epoch": 200,
          "updates": 200,
          "samples_seen": 204800,
          "online_loss": 0.019334256649017334,
          "train_loss": 0.019149335101246834,
          "validation_loss": 0.01609225571155548,
          "train_metric": 0.0019286508904770017,
          "validation_metric": 0.001620753202587366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.660601396999482
        },
        {
          "epoch": 201,
          "updates": 201,
          "samples_seen": 205824,
          "online_loss": 0.019149336963891983,
          "train_loss": 0.01896587200462818,
          "validation_loss": 0.015938322991132736,
          "train_metric": 0.0019101729849353433,
          "validation_metric": 0.0016052494756877422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6662340639995818
        },
        {
          "epoch": 202,
          "updates": 202,
          "samples_seen": 206848,
          "online_loss": 0.018965870141983032,
          "train_loss": 0.018783878535032272,
          "validation_loss": 0.015785153955221176,
          "train_metric": 0.0018918435089290142,
          "validation_metric": 0.0015898229321464896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6721075099994778
        },
        {
          "epoch": 203,
          "updates": 203,
          "samples_seen": 207872,
          "online_loss": 0.018783878535032272,
          "train_loss": 0.018603375181555748,
          "validation_loss": 0.015632793307304382,
          "train_metric": 0.0018736638594418764,
          "validation_metric": 0.0015744775300845504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6778917369992996
        },
        {
          "epoch": 204,
          "updates": 204,
          "samples_seen": 208896,
          "online_loss": 0.018603377044200897,
          "train_loss": 0.018424399197101593,
          "validation_loss": 0.015481379814445972,
          "train_metric": 0.001855637994594872,
          "validation_metric": 0.001559227705001831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6836601569993945
        },
        {
          "epoch": 205,
          "updates": 205,
          "samples_seen": 209920,
          "online_loss": 0.018424399197101593,
          "train_loss": 0.018246965482831,
          "validation_loss": 0.015331093221902847,
          "train_metric": 0.0018377677770331502,
          "validation_metric": 0.0015440917341038585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6899618429997645
        },
        {
          "epoch": 206,
          "updates": 206,
          "samples_seen": 210944,
          "online_loss": 0.01824696734547615,
          "train_loss": 0.018071098253130913,
          "validation_loss": 0.015181987546384335,
          "train_metric": 0.0018200549529865384,
          "validation_metric": 0.0015290741575881839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.695684894999431
        },
        {
          "epoch": 207,
          "updates": 207,
          "samples_seen": 211968,
          "online_loss": 0.018071098253130913,
          "train_loss": 0.017896821722388268,
          "validation_loss": 0.015034045092761517,
          "train_metric": 0.0018025024328380823,
          "validation_metric": 0.0015141740441322327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7013347729989619
        },
        {
          "epoch": 208,
          "updates": 208,
          "samples_seen": 212992,
          "online_loss": 0.017896821722388268,
          "train_loss": 0.01772415079176426,
          "validation_loss": 0.014887221157550812,
          "train_metric": 0.0017851116135716438,
          "validation_metric": 0.0014993863878771663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7074119560002146
        },
        {
          "epoch": 209,
          "updates": 209,
          "samples_seen": 214016,
          "online_loss": 0.01772415079176426,
          "train_loss": 0.01755310781300068,
          "validation_loss": 0.014741534367203712,
          "train_metric": 0.0017678847070783377,
          "validation_metric": 0.0014847132842987776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7132892800000263
        },
        {
          "epoch": 210,
          "updates": 210,
          "samples_seen": 215040,
          "online_loss": 0.017553109675645828,
          "train_loss": 0.017383717000484467,
          "validation_loss": 0.014597069472074509,
          "train_metric": 0.0017508243909105659,
          "validation_metric": 0.0014701634645462036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7190734359992348
        },
        {
          "epoch": 211,
          "updates": 211,
          "samples_seen": 216064,
          "online_loss": 0.017383717000484467,
          "train_loss": 0.017215993255376816,
          "validation_loss": 0.01445392332971096,
          "train_metric": 0.0017339318292215466,
          "validation_metric": 0.0014557462418451905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7253029419989616
        },
        {
          "epoch": 212,
          "updates": 212,
          "samples_seen": 217088,
          "online_loss": 0.017215993255376816,
          "train_loss": 0.017049945890903473,
          "validation_loss": 0.014312129467725754,
          "train_metric": 0.0017172080697491765,
          "validation_metric": 0.0014414653414860368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7309741259996372
        },
        {
          "epoch": 213,
          "updates": 213,
          "samples_seen": 218112,
          "online_loss": 0.017049945890903473,
          "train_loss": 0.016885602846741676,
          "validation_loss": 0.014171640388667583,
          "train_metric": 0.001700656022876501,
          "validation_metric": 0.001427315641194582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7364971529987088
        },
        {
          "epoch": 214,
          "updates": 214,
          "samples_seen": 219136,
          "online_loss": 0.016885604709386826,
          "train_loss": 0.01672298088669777,
          "validation_loss": 0.014032428152859211,
          "train_metric": 0.0016842774348333478,
          "validation_metric": 0.0014132946962490678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7428462759999093
        },
        {
          "epoch": 215,
          "updates": 215,
          "samples_seen": 220160,
          "online_loss": 0.01672298088669777,
          "train_loss": 0.01656208373606205,
          "validation_loss": 0.013894537463784218,
          "train_metric": 0.0016680725384503603,
          "validation_metric": 0.0013994069304317236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7485523559989815
        },
        {
          "epoch": 216,
          "updates": 216,
          "samples_seen": 221184,
          "online_loss": 0.01656208373606205,
          "train_loss": 0.01640293374657631,
          "validation_loss": 0.013758068904280663,
          "train_metric": 0.0016520433127880096,
          "validation_metric": 0.0013856622390449047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7541200449995813
        },
        {
          "epoch": 217,
          "updates": 217,
          "samples_seen": 222208,
          "online_loss": 0.01640293188393116,
          "train_loss": 0.016245543956756592,
          "validation_loss": 0.013623112812638283,
          "train_metric": 0.0016361918533220887,
          "validation_metric": 0.001372070051729679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7601589280002372
        },
        {
          "epoch": 218,
          "updates": 218,
          "samples_seen": 223232,
          "online_loss": 0.01624554581940174,
          "train_loss": 0.016089918091893196,
          "validation_loss": 0.013489669188857079,
          "train_metric": 0.001620517810806632,
          "validation_metric": 0.0013586300192400813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.76584211699992
        },
        {
          "epoch": 219,
          "updates": 219,
          "samples_seen": 224256,
          "online_loss": 0.016089918091893196,
          "train_loss": 0.01593608781695366,
          "validation_loss": 0.013357715681195259,
          "train_metric": 0.0016050240956246853,
          "validation_metric": 0.001345340278930962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7714361299986194
        },
        {
          "epoch": 220,
          "updates": 220,
          "samples_seen": 225280,
          "online_loss": 0.01593608595430851,
          "train_loss": 0.015784041956067085,
          "validation_loss": 0.013227248564362526,
          "train_metric": 0.0015897109406068921,
          "validation_metric": 0.0013322001323103905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7772479559989733
        },
        {
          "epoch": 221,
          "updates": 221,
          "samples_seen": 226304,
          "online_loss": 0.015784041956067085,
          "train_loss": 0.015633797273039818,
          "validation_loss": 0.01309829019010067,
          "train_metric": 0.0015745788114145398,
          "validation_metric": 0.0013192116748541594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7830385079996631
        },
        {
          "epoch": 222,
          "updates": 222,
          "samples_seen": 227328,
          "online_loss": 0.015633797273039818,
          "train_loss": 0.015485364943742752,
          "validation_loss": 0.012970906682312489,
          "train_metric": 0.0015596292214468122,
          "validation_metric": 0.0013063822407275438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7887817609989725
        },
        {
          "epoch": 223,
          "updates": 223,
          "samples_seen": 228352,
          "online_loss": 0.015485365875065327,
          "train_loss": 0.015338747762143612,
          "validation_loss": 0.01284513995051384,
          "train_metric": 0.0015448624035343528,
          "validation_metric": 0.0012937153223901987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7945522999998502
        },
        {
          "epoch": 224,
          "updates": 224,
          "samples_seen": 229376,
          "online_loss": 0.015338745899498463,
          "train_loss": 0.01519395224750042,
          "validation_loss": 0.01272093877196312,
          "train_metric": 0.0015302790561690927,
          "validation_metric": 0.001281206263229251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8002889919989684
        },
        {
          "epoch": 225,
          "updates": 225,
          "samples_seen": 230400,
          "online_loss": 0.01519395224750042,
          "train_loss": 0.015050990507006645,
          "validation_loss": 0.012598264031112194,
          "train_metric": 0.0015158806927502155,
          "validation_metric": 0.0012688509887084365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.806154380999942
        },
        {
          "epoch": 226,
          "updates": 226,
          "samples_seen": 231424,
          "online_loss": 0.01505099143832922,
          "train_loss": 0.014909853227436543,
          "validation_loss": 0.012477121315896511,
          "train_metric": 0.0015016660327091813,
          "validation_metric": 0.0012566500809043646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8118033399987326
        },
        {
          "epoch": 227,
          "updates": 227,
          "samples_seen": 232448,
          "online_loss": 0.014909852296113968,
          "train_loss": 0.014770550653338432,
          "validation_loss": 0.012357545085251331,
          "train_metric": 0.0014876356581225991,
          "validation_metric": 0.0012446066830307245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.817506172999856
        },
        {
          "epoch": 228,
          "updates": 228,
          "samples_seen": 233472,
          "online_loss": 0.014770549722015858,
          "train_loss": 0.014633073471486568,
          "validation_loss": 0.012239589355885983,
          "train_metric": 0.0014737894525751472,
          "validation_metric": 0.0012327264994382858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8235539809993497
        },
        {
          "epoch": 229,
          "updates": 229,
          "samples_seen": 234496,
          "online_loss": 0.014633072540163994,
          "train_loss": 0.01449743565171957,
          "validation_loss": 0.01212325133383274,
          "train_metric": 0.0014601288130506873,
          "validation_metric": 0.0012210095301270485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8294506399997772
        },
        {
          "epoch": 230,
          "updates": 230,
          "samples_seen": 235520,
          "online_loss": 0.014497434720396996,
          "train_loss": 0.014363635331392288,
          "validation_loss": 0.012008504010736942,
          "train_metric": 0.0014466525753960013,
          "validation_metric": 0.0012094525154680014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8350840080001944
        },
        {
          "epoch": 231,
          "updates": 231,
          "samples_seen": 236544,
          "online_loss": 0.014363635331392288,
          "train_loss": 0.014231644570827484,
          "validation_loss": 0.011895321309566498,
          "train_metric": 0.0014333591097965837,
          "validation_metric": 0.0011980532435700297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8408502219990623
        },
        {
          "epoch": 232,
          "updates": 232,
          "samples_seen": 237568,
          "online_loss": 0.014231645502150059,
          "train_loss": 0.014101490378379822,
          "validation_loss": 0.011783742345869541,
          "train_metric": 0.0014202503953129053,
          "validation_metric": 0.0011868153233081102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8500856909995491
        },
        {
          "epoch": 233,
          "updates": 233,
          "samples_seen": 238592,
          "online_loss": 0.014101488515734673,
          "train_loss": 0.01397313829511404,
          "validation_loss": 0.011673802509903908,
          "train_metric": 0.0014073234051465988,
          "validation_metric": 0.0011757425963878632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9427511769990815
        },
        {
          "epoch": 234,
          "updates": 234,
          "samples_seen": 239616,
          "online_loss": 0.01397313829511404,
          "train_loss": 0.013846595771610737,
          "validation_loss": 0.011565502732992172,
          "train_metric": 0.0013945784885436296,
          "validation_metric": 0.0011648350628092885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9485344309996435
        },
        {
          "epoch": 235,
          "updates": 235,
          "samples_seen": 240640,
          "online_loss": 0.013846594840288162,
          "train_loss": 0.013721850700676441,
          "validation_loss": 0.01145881973206997,
          "train_metric": 0.0013820144813507795,
          "validation_metric": 0.001154090161435306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9542172689998552
        },
        {
          "epoch": 236,
          "updates": 236,
          "samples_seen": 241664,
          "online_loss": 0.013721850700676441,
          "train_loss": 0.013598894700407982,
          "validation_loss": 0.01135371346026659,
          "train_metric": 0.0013696306850761175,
          "validation_metric": 0.0011435045162215829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9603903489987715
        },
        {
          "epoch": 237,
          "updates": 237,
          "samples_seen": 242688,
          "online_loss": 0.013598893769085407,
          "train_loss": 0.013477708213031292,
          "validation_loss": 0.011250190436840057,
          "train_metric": 0.001357425469905138,
          "validation_metric": 0.0011330778943374753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9661479520000285
        },
        {
          "epoch": 238,
          "updates": 238,
          "samples_seen": 243712,
          "online_loss": 0.013477709144353867,
          "train_loss": 0.013358290307223797,
          "validation_loss": 0.011148265562951565,
          "train_metric": 0.0013453979045152664,
          "validation_metric": 0.0011228123912587762,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9719194849985797
        },
        {
          "epoch": 239,
          "updates": 239,
          "samples_seen": 244736,
          "online_loss": 0.013358289375901222,
          "train_loss": 0.013240613043308258,
          "validation_loss": 0.011047950945794582,
          "train_metric": 0.0013335458934307098,
          "validation_metric": 0.0011127092875540257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9777143549999892
        },
        {
          "epoch": 240,
          "updates": 240,
          "samples_seen": 245760,
          "online_loss": 0.013240613043308258,
          "train_loss": 0.013124671764671803,
          "validation_loss": 0.010949228890240192,
          "train_metric": 0.0013218688545748591,
          "validation_metric": 0.0011027661385014653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9834698049999133
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 240
      },
      "best_validation_metric": 0.0011027661385014653,
      "config": {
        "label": "Full batch",
        "batch": 1024,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.033207922827263155
    },
    "1150d4df4a82_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.5785819953307509,
          "train_loss": 0.24066151678562164,
          "validation_loss": 0.15968576073646545,
          "train_metric": 0.024238549172878265,
          "validation_metric": 0.016082964837551117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.021245440999337006
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.20281902188435197,
          "train_loss": 0.14825578033924103,
          "validation_loss": 0.12119951099157333,
          "train_metric": 0.014931779354810715,
          "validation_metric": 0.012206770479679108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03849066600014339
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13218980515375733,
          "train_loss": 0.11559955775737762,
          "validation_loss": 0.09988148510456085,
          "train_metric": 0.011642763391137123,
          "validation_metric": 0.010059698484838009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05538341300052707
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10902172443456948,
          "train_loss": 0.09608157724142075,
          "validation_loss": 0.08105382323265076,
          "train_metric": 0.009676986373960972,
          "validation_metric": 0.008163443766534328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07194715200057544
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08852994372136891,
          "train_loss": 0.07870391011238098,
          "validation_loss": 0.06390558928251266,
          "train_metric": 0.00792676955461502,
          "validation_metric": 0.0064363377168774605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08881906299939146
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.0738776980433613,
          "train_loss": 0.06744672358036041,
          "validation_loss": 0.05580252408981323,
          "train_metric": 0.006792987231165171,
          "validation_metric": 0.005620225332677364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13589236100051494
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.0662799112033099,
          "train_loss": 0.062105804681777954,
          "validation_loss": 0.051005952060222626,
          "train_metric": 0.006255070213228464,
          "validation_metric": 0.0051371329464018345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16529159200035792
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06061952211894095,
          "train_loss": 0.058522891253232956,
          "validation_loss": 0.04748249426484108,
          "train_metric": 0.005894211120903492,
          "validation_metric": 0.0047822631895542145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19284062000042468
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0586743209278211,
          "train_loss": 0.05518710985779762,
          "validation_loss": 0.04696771502494812,
          "train_metric": 0.00555824302136898,
          "validation_metric": 0.0047304159961640835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3325209359991277
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05452180642168969,
          "train_loss": 0.05012526735663414,
          "validation_loss": 0.04216490313410759,
          "train_metric": 0.005048433784395456,
          "validation_metric": 0.004246694967150688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34432423900034337
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0489687561057508,
          "train_loss": 0.0455508716404438,
          "validation_loss": 0.0387922078371048,
          "train_metric": 0.004587717354297638,
          "validation_metric": 0.003907009027898312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3561684790001891
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.045328833162784576,
          "train_loss": 0.04298621788620949,
          "validation_loss": 0.03603599593043327,
          "train_metric": 0.004329414572566748,
          "validation_metric": 0.0036294138990342617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3677858640003251
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.040962925646454096,
          "train_loss": 0.03840823099017143,
          "validation_loss": 0.03274441510438919,
          "train_metric": 0.003868336323648691,
          "validation_metric": 0.003297897521406412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37925838899900555
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.035591799300163984,
          "train_loss": 0.03237031772732735,
          "validation_loss": 0.027760149911046028,
          "train_metric": 0.003260219469666481,
          "validation_metric": 0.0027959002181887627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39094049999948766
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.034175259876064956,
          "train_loss": 0.030200820416212082,
          "validation_loss": 0.02461184561252594,
          "train_metric": 0.0030417158268392086,
          "validation_metric": 0.0024788149166852236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4024609459993371
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.029044358525425196,
          "train_loss": 0.02824326977133751,
          "validation_loss": 0.02484940178692341,
          "train_metric": 0.0028445585630834103,
          "validation_metric": 0.0025027410592883825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41356743699907383
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.027968548878561705,
          "train_loss": 0.026699528098106384,
          "validation_loss": 0.022850092500448227,
          "train_metric": 0.0026890786830335855,
          "validation_metric": 0.0023013774771243334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4250626070006547
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029929061303846538,
          "train_loss": 0.028845902532339096,
          "validation_loss": 0.023522071540355682,
          "train_metric": 0.002905253553763032,
          "validation_metric": 0.002369056921452284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4357555610004056
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.0253515514486935,
          "train_loss": 0.022620845586061478,
          "validation_loss": 0.019433408975601196,
          "train_metric": 0.0022782885935157537,
          "validation_metric": 0.001957261934876442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44704702499984705
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02046690188581124,
          "train_loss": 0.019571326673030853,
          "validation_loss": 0.016824472695589066,
          "train_metric": 0.0019711521454155445,
          "validation_metric": 0.00169449916575104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4587405700003728
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.021430058928672224,
          "train_loss": 0.019809599965810776,
          "validation_loss": 0.01710662618279457,
          "train_metric": 0.001995150465518236,
          "validation_metric": 0.0017229169607162476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47017164200042316
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.019933807488996536,
          "train_loss": 0.018322203308343887,
          "validation_loss": 0.015159186907112598,
          "train_metric": 0.0018453451339155436,
          "validation_metric": 0.0015267777489498258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4953879439999582
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.0184865053743124,
          "train_loss": 0.018727105110883713,
          "validation_loss": 0.016270481050014496,
          "train_metric": 0.0018861254211515188,
          "validation_metric": 0.001638703397475183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5120316440006718
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019489634316414595,
          "train_loss": 0.018848121166229248,
          "validation_loss": 0.015607197768986225,
          "train_metric": 0.0018983136396855116,
          "validation_metric": 0.0015718997456133366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5233046609992016
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.025991787144448608,
          "train_loss": 0.022055301815271378,
          "validation_loss": 0.01751544512808323,
          "train_metric": 0.0022213291376829147,
          "validation_metric": 0.0017640916630625725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5353992189993733
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.020158647326752543,
          "train_loss": 0.019878527149558067,
          "validation_loss": 0.015789136290550232,
          "train_metric": 0.0020020923111587763,
          "validation_metric": 0.0015902240993455052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5471312179997767
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01832235194160603,
          "train_loss": 0.015337939374148846,
          "validation_loss": 0.013187260366976261,
          "train_metric": 0.0015447810292243958,
          "validation_metric": 0.001328172511421144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.563337228999444
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015697864320827648,
          "train_loss": 0.01441369578242302,
          "validation_loss": 0.012476164847612381,
          "train_metric": 0.0014516946393996477,
          "validation_metric": 0.0012565534561872482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5756641759999184
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.01796175292111002,
          "train_loss": 0.015583343803882599,
          "validation_loss": 0.012778116390109062,
          "train_metric": 0.0015694972826167941,
          "validation_metric": 0.0012869652127847075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.592702592999558
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.017315161181613803,
          "train_loss": 0.026303283870220184,
          "validation_loss": 0.022693904116749763,
          "train_metric": 0.0026491701137274504,
          "validation_metric": 0.002285646740347147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6102209370001219
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.018650115351192653,
          "train_loss": 0.015033909119665623,
          "validation_loss": 0.012586742639541626,
          "train_metric": 0.0015141601907089353,
          "validation_metric": 0.0012676906771957874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6218287270003202
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01525422916165553,
          "train_loss": 0.012996863573789597,
          "validation_loss": 0.011027900502085686,
          "train_metric": 0.0013089965796098113,
          "validation_metric": 0.0011106899473816156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6345068350001384
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.017277867038501427,
          "train_loss": 0.018468111753463745,
          "validation_loss": 0.01605597883462906,
          "train_metric": 0.0018600407056510448,
          "validation_metric": 0.001617099391296506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6465626520002843
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.017836909275501966,
          "train_loss": 0.015464769676327705,
          "validation_loss": 0.013479636050760746,
          "train_metric": 0.001557554816827178,
          "validation_metric": 0.0013576196506619453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6578930639989267
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014926508301869035,
          "train_loss": 0.01525525376200676,
          "validation_loss": 0.012201250530779362,
          "train_metric": 0.0015364533755928278,
          "validation_metric": 0.00122886523604393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6729799469994759
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.013606283508124761,
          "train_loss": 0.01189349964261055,
          "validation_loss": 0.010086850263178349,
          "train_metric": 0.0011978696566075087,
          "validation_metric": 0.0010159105295315385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6918827690005855
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013865286426153034,
          "train_loss": 0.01421459298580885,
          "validation_loss": 0.012337259016931057,
          "train_metric": 0.0014316417509689927,
          "validation_metric": 0.00124256347771734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.710786552999707
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.013906719570513815,
          "train_loss": 0.01308034174144268,
          "validation_loss": 0.011206195689737797,
          "train_metric": 0.0013174042105674744,
          "validation_metric": 0.0011286470107734203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7296411299994361
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.015102692559594288,
          "train_loss": 0.012156970798969269,
          "validation_loss": 0.009952702559530735,
          "train_metric": 0.0012244055978953838,
          "validation_metric": 0.00100239971652627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7486161830001947
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.011600760830333456,
          "train_loss": 0.01074554119259119,
          "validation_loss": 0.009170734323561192,
          "train_metric": 0.0010822515469044447,
          "validation_metric": 0.0009236427722498775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7672800719992665
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011461123736808077,
          "train_loss": 0.011186208575963974,
          "validation_loss": 0.008815637789666653,
          "train_metric": 0.001126633957028389,
          "validation_metric": 0.000887878704816103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7867815190002148
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.012731426904792897,
          "train_loss": 0.011028323322534561,
          "validation_loss": 0.008532376028597355,
          "train_metric": 0.0011107323225587606,
          "validation_metric": 0.0008593496750108898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8057653899995785
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011592802344239317,
          "train_loss": 0.01758863590657711,
          "validation_loss": 0.0152521887794137,
          "train_metric": 0.0017714631976559758,
          "validation_metric": 0.0015361446421593428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8237847380005405
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014004348166054115,
          "train_loss": 0.011498655192553997,
          "validation_loss": 0.009649231098592281,
          "train_metric": 0.0011581024155020714,
          "validation_metric": 0.0009718352230265737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8412866460002988
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.011971019470365718,
          "train_loss": 0.010247869417071342,
          "validation_loss": 0.008546342141926289,
          "train_metric": 0.0010321278823539615,
          "validation_metric": 0.0008607562049292028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8586193609990005
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010324565839255229,
          "train_loss": 0.009410005994141102,
          "validation_loss": 0.008049655705690384,
          "train_metric": 0.0009477414423599839,
          "validation_metric": 0.0008107318426482379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8719080649989337
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0099878313485533,
          "train_loss": 0.009574130177497864,
          "validation_loss": 0.008396321907639503,
          "train_metric": 0.0009642713703215122,
          "validation_metric": 0.0008456468349322677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8895444629997655
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010388583701569587,
          "train_loss": 0.010846927762031555,
          "validation_loss": 0.009024742990732193,
          "train_metric": 0.0010924628004431725,
          "validation_metric": 0.0009089390514418483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9078683219995582
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.011729866499081254,
          "train_loss": 0.008839637972414494,
          "validation_loss": 0.007451624143868685,
          "train_metric": 0.0008902959525585175,
          "validation_metric": 0.0007505002431571484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9238897140003246
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009075396141270176,
          "train_loss": 0.008698005229234695,
          "validation_loss": 0.007412043400108814,
          "train_metric": 0.0008760312921367586,
          "validation_metric": 0.0007465138914994895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9381962139996176
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008586802636273205,
          "train_loss": 0.009239615872502327,
          "validation_loss": 0.00846143253147602,
          "train_metric": 0.0009305804269388318,
          "validation_metric": 0.000852204451803118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.95426081199912
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.009399648464750499,
          "train_loss": 0.008207714185118675,
          "validation_loss": 0.007094361819326878,
          "train_metric": 0.0008266508812084794,
          "validation_metric": 0.0007145180716179311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9807512309998856
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.009622707089874893,
          "train_loss": 0.008783034980297089,
          "validation_loss": 0.007847893983125687,
          "train_metric": 0.0008845950360409915,
          "validation_metric": 0.0007904112571850419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.996637386000657
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009374020868563093,
          "train_loss": 0.008376436308026314,
          "validation_loss": 0.007370653096586466,
          "train_metric": 0.0008436439675278962,
          "validation_metric": 0.0007423452916555107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0126210960006574
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008409974529058672,
          "train_loss": 0.008557048626244068,
          "validation_loss": 0.006746719125658274,
          "train_metric": 0.0008618345018476248,
          "validation_metric": 0.0006795047665946186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0294730009991326
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.008126632194034755,
          "train_loss": 0.007593871559947729,
          "validation_loss": 0.006299670785665512,
          "train_metric": 0.000764826952945441,
          "validation_metric": 0.0006344797438941896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0465842350004095
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008103969594230875,
          "train_loss": 0.007627340033650398,
          "validation_loss": 0.006149530876427889,
          "train_metric": 0.0007681977585889399,
          "validation_metric": 0.0006193582667037845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.064146319999054
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007666467514354736,
          "train_loss": 0.007350016385316849,
          "validation_loss": 0.0060503422282636166,
          "train_metric": 0.0007402666378766298,
          "validation_metric": 0.0006093683768995106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.081532418000279
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.008778885996434838,
          "train_loss": 0.007884827442467213,
          "validation_loss": 0.007127765100449324,
          "train_metric": 0.0007941309013403952,
          "validation_metric": 0.0007178823580034077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0982661569996708
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007124831900000572,
          "train_loss": 0.008507896214723587,
          "validation_loss": 0.006483363453298807,
          "train_metric": 0.0008568841149099171,
          "validation_metric": 0.0006529804668389261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1145541349997075
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.00802242913050577,
          "train_loss": 0.0072455033659935,
          "validation_loss": 0.005634790752083063,
          "train_metric": 0.000729740597307682,
          "validation_metric": 0.0005675156135112047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1314246669990098
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007499659288441762,
          "train_loss": 0.006905286107212305,
          "validation_loss": 0.005780530162155628,
          "train_metric": 0.0006954751443117857,
          "validation_metric": 0.0005821938975714147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.147982745000263
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006983088416745886,
          "train_loss": 0.00676384475082159,
          "validation_loss": 0.005742666777223349,
          "train_metric": 0.0006812297506257892,
          "validation_metric": 0.0005783804226666689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1643748459991912
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007186421637015883,
          "train_loss": 0.006360976025462151,
          "validation_loss": 0.0053828442469239235,
          "train_metric": 0.0006406542379409075,
          "validation_metric": 0.000542140391189605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1816414999993867
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008146895095705986,
          "train_loss": 0.00710770720615983,
          "validation_loss": 0.0064037274569272995,
          "train_metric": 0.0007158622611314058,
          "validation_metric": 0.0006449599750339985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.198178774000553
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0070507670170627534,
          "train_loss": 0.006784606724977493,
          "validation_loss": 0.005462770350277424,
          "train_metric": 0.0006833208026364446,
          "validation_metric": 0.0005501902778632939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2145999040003517
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007892993970017415,
          "train_loss": 0.007619489450007677,
          "validation_loss": 0.005710892844945192,
          "train_metric": 0.0007674070075154305,
          "validation_metric": 0.0005751801654696465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2308899330000713
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0075215069082332775,
          "train_loss": 0.008999602869153023,
          "validation_loss": 0.007687784265726805,
          "train_metric": 0.0009064070181921124,
          "validation_metric": 0.0007742854650132358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2471576040006767
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.008329692267579958,
          "train_loss": 0.006699476856738329,
          "validation_loss": 0.005957436747848988,
          "train_metric": 0.000674746697768569,
          "validation_metric": 0.0006000111461617053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2645645039992814
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006823713432822842,
          "train_loss": 0.006718658842146397,
          "validation_loss": 0.005331420339643955,
          "train_metric": 0.0006766787264496088,
          "validation_metric": 0.0005369611317291856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.282320090000212
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0063610283978050575,
          "train_loss": 0.006560879293829203,
          "validation_loss": 0.005013317335397005,
          "train_metric": 0.0006607878021895885,
          "validation_metric": 0.0005049229366704822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3107265910002752
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006628050803556107,
          "train_loss": 0.005550465080887079,
          "validation_loss": 0.00450581219047308,
          "train_metric": 0.0005590225337073207,
          "validation_metric": 0.00045380889787338674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3277268809997622
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006065987341571599,
          "train_loss": 0.006213136482983828,
          "validation_loss": 0.004682971630245447,
          "train_metric": 0.000625764369033277,
          "validation_metric": 0.0004716518160421401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3444636900003388
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0066800750500988215,
          "train_loss": 0.007923943921923637,
          "validation_loss": 0.00671512121334672,
          "train_metric": 0.0007980706868693233,
          "validation_metric": 0.0006763224373571575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.361010818000068
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.008286909156595357,
          "train_loss": 0.005879192613065243,
          "validation_loss": 0.005012104287743568,
          "train_metric": 0.0005921307601965964,
          "validation_metric": 0.0005048008169978857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3767763599989848
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.008240004448452964,
          "train_loss": 0.006389837712049484,
          "validation_loss": 0.005709755700081587,
          "train_metric": 0.0006435610703192651,
          "validation_metric": 0.0005750657292082906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3928331909992266
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007031992237898521,
          "train_loss": 0.0061054527759552,
          "validation_loss": 0.005275707691907883,
          "train_metric": 0.000614918943028897,
          "validation_metric": 0.0005313499132171273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4087770149999415
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006232404222828336,
          "train_loss": 0.005883560515940189,
          "validation_loss": 0.004626329056918621,
          "train_metric": 0.0005925706936977804,
          "validation_metric": 0.00046594697050750256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4248944259998098
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006698730117932428,
          "train_loss": 0.00830888468772173,
          "validation_loss": 0.005257878918200731,
          "train_metric": 0.0008368405397050083,
          "validation_metric": 0.0005295543232932687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.440855606999321
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.007175867009209469,
          "train_loss": 0.005164492409676313,
          "validation_loss": 0.0041005248203873634,
          "train_metric": 0.0005201487801969051,
          "validation_metric": 0.00041298987343907356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4575346349993197
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005640656236209907,
          "train_loss": 0.005428474862128496,
          "validation_loss": 0.004622603766620159,
          "train_metric": 0.000546736060641706,
          "validation_metric": 0.00046557170571759343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4738841469988984
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.007278883465914987,
          "train_loss": 0.007279138546437025,
          "validation_loss": 0.005336048547178507,
          "train_metric": 0.0007331281085498631,
          "validation_metric": 0.0005374272586777806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4897334639990731
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006428858388972003,
          "train_loss": 0.005414268001914024,
          "validation_loss": 0.004378072917461395,
          "train_metric": 0.000545305258128792,
          "validation_metric": 0.0004409434332046658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5059498930004338
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.006930987336090766,
          "train_loss": 0.004939272999763489,
          "validation_loss": 0.004174148663878441,
          "train_metric": 0.0004974654875695705,
          "validation_metric": 0.0004204050055705011,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5216513599989412
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005648306665534619,
          "train_loss": 0.004681389778852463,
          "validation_loss": 0.004043219611048698,
          "train_metric": 0.00047149244346655905,
          "validation_metric": 0.0004072182346135378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5387032779999572
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005824902589665726,
          "train_loss": 0.005570245441049337,
          "validation_loss": 0.004077776335179806,
          "train_metric": 0.0005610146909020841,
          "validation_metric": 0.00041069870349019766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5548484099999769
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005732758989324793,
          "train_loss": 0.006048447918146849,
          "validation_loss": 0.0051933652721345425,
          "train_metric": 0.0006091775139793754,
          "validation_metric": 0.0005230566603131592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5708143619995099
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.00487511732353596,
          "train_loss": 0.004281728528439999,
          "validation_loss": 0.003500000573694706,
          "train_metric": 0.00043124001240357757,
          "validation_metric": 0.00035250725341029465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.587878145999639
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005490752446348779,
          "train_loss": 0.0075278449803590775,
          "validation_loss": 0.005237332079559565,
          "train_metric": 0.000758176960516721,
          "validation_metric": 0.0005274848663248122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6044097399990278
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.006032784658600576,
          "train_loss": 0.004147201310843229,
          "validation_loss": 0.003196843434125185,
          "train_metric": 0.00041769095696508884,
          "validation_metric": 0.00032197439577430487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6214374829996814
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0045539015118265525,
          "train_loss": 0.003808197332546115,
          "validation_loss": 0.0031398420687764883,
          "train_metric": 0.0003835476818494499,
          "validation_metric": 0.00031623340328224003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6384987519995775
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004446769387868699,
          "train_loss": 0.003838970558717847,
          "validation_loss": 0.00305731687694788,
          "train_metric": 0.00038664706517010927,
          "validation_metric": 0.0003079217567574233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6555330119990685
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004661549588490743,
          "train_loss": 0.003889218671247363,
          "validation_loss": 0.0030916959512978792,
          "train_metric": 0.0003917078720405698,
          "validation_metric": 0.0003113843558821827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.671494697999151
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004343237917055376,
          "train_loss": 0.003794271731749177,
          "validation_loss": 0.0030378333758562803,
          "train_metric": 0.0003821451682597399,
          "validation_metric": 0.000305959430988878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6889780090004933
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004010706965345889,
          "train_loss": 0.004695076495409012,
          "validation_loss": 0.0032074693590402603,
          "train_metric": 0.0004728709172923118,
          "validation_metric": 0.0003230445727240294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7059788509995997
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0048800584190757945,
          "train_loss": 0.004135572351515293,
          "validation_loss": 0.003357178997248411,
          "train_metric": 0.00041651964420452714,
          "validation_metric": 0.00033812274341471493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7221261260001484
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.004035678837681189,
          "train_loss": 0.003541690995916724,
          "validation_loss": 0.0029904022812843323,
          "train_metric": 0.00035670612123794854,
          "validation_metric": 0.0003011823573615402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7388967890001368
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004549244105874095,
          "train_loss": 0.007052480708807707,
          "validation_loss": 0.005183737725019455,
          "train_metric": 0.0007103000534698367,
          "validation_metric": 0.0005220869788900018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7551238699998066
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.008762747966102324,
          "train_loss": 0.004422488622367382,
          "validation_loss": 0.003422736655920744,
          "train_metric": 0.00044541689567267895,
          "validation_metric": 0.00034472550032660365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7710629199991672
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004970894042344298,
          "train_loss": 0.007190850097686052,
          "validation_loss": 0.004736239090561867,
          "train_metric": 0.0007242360734380782,
          "validation_metric": 0.00047701667062938213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.787357744999099
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.006154970556963235,
          "train_loss": 0.006614692974835634,
          "validation_loss": 0.006176328286528587,
          "train_metric": 0.0006662076339125633,
          "validation_metric": 0.0006220571231096983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8036176539990265
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.004848615717492066,
          "train_loss": 0.0038373428396880627,
          "validation_loss": 0.0031854603439569473,
          "train_metric": 0.0003864831232931465,
          "validation_metric": 0.00032082790858112276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.819599816999471
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0043659103830577806,
          "train_loss": 0.0037215915508568287,
          "validation_loss": 0.0033144704066216946,
          "train_metric": 0.00037482508923858404,
          "validation_metric": 0.0003338213136885315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.835250484999051
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0037703865782532375,
          "train_loss": 0.0033768422435969114,
          "validation_loss": 0.0028291926719248295,
          "train_metric": 0.00034010314266197383,
          "validation_metric": 0.00028494588332250714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.851985049999712
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003850408087600954,
          "train_loss": 0.0038566645234823227,
          "validation_loss": 0.002810301724821329,
          "train_metric": 0.0003884290927089751,
          "validation_metric": 0.00028304324951022863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8690316299998813
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0036429797328310087,
          "train_loss": 0.003893908578902483,
          "validation_loss": 0.0028391662053763866,
          "train_metric": 0.00039218022720888257,
          "validation_metric": 0.0002859504020307213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8854719909995765
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.003939842183172004,
          "train_loss": 0.004020857624709606,
          "validation_loss": 0.00285907881334424,
          "train_metric": 0.0004049660637974739,
          "validation_metric": 0.0002879559760913253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9016862169992237
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.006216302259417716,
          "train_loss": 0.004236565437167883,
          "validation_loss": 0.003314891830086708,
          "train_metric": 0.0004266913456376642,
          "validation_metric": 0.00033386374707333744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9176526219998777
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.004633348784409463,
          "train_loss": 0.0040680477395653725,
          "validation_loss": 0.0030227238312363625,
          "train_metric": 0.0004097189230378717,
          "validation_metric": 0.0003044376499019563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9337373319995095
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.00391630687227007,
          "train_loss": 0.006275310181081295,
          "validation_loss": 0.004126521293073893,
          "train_metric": 0.000632026232779026,
          "validation_metric": 0.00041560811223462224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9505196270001761
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004763353004818782,
          "train_loss": 0.0034501319751143456,
          "validation_loss": 0.002509691519662738,
          "train_metric": 0.0003474846016615629,
          "validation_metric": 0.00025276694213971496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.967727871999159
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.00421532570180716,
          "train_loss": 0.0038624031003564596,
          "validation_loss": 0.0031479850877076387,
          "train_metric": 0.00038900706567801535,
          "validation_metric": 0.00031705352012068033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9841534770002909
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0037217386852717027,
          "train_loss": 0.006393822841346264,
          "validation_loss": 0.005712313111871481,
          "train_metric": 0.0006439624121412635,
          "validation_metric": 0.0005753233563154936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0006491990006907
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.006156473507871851,
          "train_loss": 0.003115376690402627,
          "validation_loss": 0.0025999839417636395,
          "train_metric": 0.0003137692983727902,
          "validation_metric": 0.00026186087052337825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0172234999990906
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0036181848408887163,
          "train_loss": 0.0032176231034100056,
          "validation_loss": 0.0024235867895185947,
          "train_metric": 0.0003240671940147877,
          "validation_metric": 0.00024409478646703064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.034741094999845
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0032519254091312177,
          "train_loss": 0.0034723232965916395,
          "validation_loss": 0.003006942104548216,
          "train_metric": 0.00034971965942531824,
          "validation_metric": 0.00030284817330539227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0516424130000814
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0029091055839671753,
          "train_loss": 0.0030959751456975937,
          "validation_loss": 0.0021414398215711117,
          "train_metric": 0.0003118152671959251,
          "validation_metric": 0.0002156779810320586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0697091029996955
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0030300220460048877,
          "train_loss": 0.002905888482928276,
          "validation_loss": 0.0024696867913007736,
          "train_metric": 0.0002926704764831811,
          "validation_metric": 0.00024873780785128474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.087023534999389
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0030614418465120252,
          "train_loss": 0.003689597826451063,
          "validation_loss": 0.0024924471508711576,
          "train_metric": 0.00037160280044190586,
          "validation_metric": 0.0002510301419533789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.103880014999959
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0033793316761148162,
          "train_loss": 0.0029266411438584328,
          "validation_loss": 0.002395999850705266,
          "train_metric": 0.000294760538963601,
          "validation_metric": 0.00024131633108481765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1203889190001064
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0032884067622944713,
          "train_loss": 0.0027470323257148266,
          "validation_loss": 0.0022338894195854664,
          "train_metric": 0.0002766710240393877,
          "validation_metric": 0.00022498915495816618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.137026942000375
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0035691659795702435,
          "train_loss": 0.0031002990435808897,
          "validation_loss": 0.002106620464473963,
          "train_metric": 0.00031225071870721877,
          "validation_metric": 0.000212171085877344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1544564989999344
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.003656093656900339,
          "train_loss": 0.003222736297175288,
          "validation_loss": 0.0027210693806409836,
          "train_metric": 0.0003245821862947196,
          "validation_metric": 0.00027405619039200246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1708697180001764
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0036966746047255583,
          "train_loss": 0.002750445157289505,
          "validation_loss": 0.002295314334332943,
          "train_metric": 0.00027701474027708173,
          "validation_metric": 0.00023117563978303224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1872713259999728
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0033217461896128953,
          "train_loss": 0.0028786661569029093,
          "validation_loss": 0.0021965266205370426,
          "train_metric": 0.00028992866282351315,
          "validation_metric": 0.00022122611699160188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.203802550000546
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.00400566911775968,
          "train_loss": 0.0028464628849178553,
          "validation_loss": 0.0023582750000059605,
          "train_metric": 0.0002866853028535843,
          "validation_metric": 0.00023751685512252152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2205754309998156
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0029259902075864375,
          "train_loss": 0.002619984559714794,
          "validation_loss": 0.0021287843119353056,
          "train_metric": 0.0002638752630446106,
          "validation_metric": 0.00021440337877720594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2369849770002475
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0029659844767593313,
          "train_loss": 0.003342970972880721,
          "validation_loss": 0.0023593404330313206,
          "train_metric": 0.00033669182448647916,
          "validation_metric": 0.00023762413184158504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.253124970000499
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0025056919257622212,
          "train_loss": 0.0035709175281226635,
          "validation_loss": 0.0029357061721384525,
          "train_metric": 0.00035964971175417304,
          "validation_metric": 0.0002956735552288592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2697092019989213
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.004325161571614444,
          "train_loss": 0.004080065526068211,
          "validation_loss": 0.0034732180647552013,
          "train_metric": 0.0004109292640350759,
          "validation_metric": 0.0003498097648844123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2864126989989018
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0034356102732999716,
          "train_loss": 0.003065694123506546,
          "validation_loss": 0.002774572465568781,
          "train_metric": 0.00030876544769853354,
          "validation_metric": 0.0002794447646010667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.303350431999206
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0026468074865988456,
          "train_loss": 0.002433589193969965,
          "validation_loss": 0.00198219227604568,
          "train_metric": 0.00024510218645446,
          "validation_metric": 0.00019963915110565722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.321315240000331
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0026325581020500977,
          "train_loss": 0.0026633315719664097,
          "validation_loss": 0.0019586319103837013,
          "train_metric": 0.00026824098313227296,
          "validation_metric": 0.00019726621394511312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.339878138000131
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0032648108062858228,
          "train_loss": 0.005058703012764454,
          "validation_loss": 0.003942716401070356,
          "train_metric": 0.0005094940424896777,
          "validation_metric": 0.0003970960096921772,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3571775860000344
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.003347333404235542,
          "train_loss": 0.0037611848674714565,
          "validation_loss": 0.0030503952875733376,
          "train_metric": 0.00037881272146478295,
          "validation_metric": 0.0003072246618103236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3743814209992706
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004248867589922156,
          "train_loss": 0.004442738834768534,
          "validation_loss": 0.0032175909727811813,
          "train_metric": 0.00044745643390342593,
          "validation_metric": 0.0003240640216972679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.391501316000358
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.005000377132091671,
          "train_loss": 0.0031094548758119345,
          "validation_loss": 0.0022825312335044146,
          "train_metric": 0.0003131728444714099,
          "validation_metric": 0.00022988820273894817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4079617299994425
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.003343742628203472,
          "train_loss": 0.0022487544920295477,
          "validation_loss": 0.0018114466220140457,
          "train_metric": 0.00022648631420452148,
          "validation_metric": 0.00018244225066155195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4247682269997313
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0035823472535412293,
          "train_loss": 0.002789368387311697,
          "validation_loss": 0.0023474714253097773,
          "train_metric": 0.0002809349971357733,
          "validation_metric": 0.00023642870655748993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4414091919989005
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0032741659415478352,
          "train_loss": 0.0030390822794288397,
          "validation_loss": 0.002136779949069023,
          "train_metric": 0.000306085217744112,
          "validation_metric": 0.0002152086526621133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4580953149998095
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.004115030576940626,
          "train_loss": 0.0036924120504409075,
          "validation_loss": 0.0027729016728699207,
          "train_metric": 0.00037188621354289353,
          "validation_metric": 0.0002792764862533659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4749960680001095
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.00393645279109478,
          "train_loss": 0.008459090255200863,
          "validation_loss": 0.00747752794995904,
          "train_metric": 0.0008519685943610966,
          "validation_metric": 0.0007531092269346118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.492408637999688
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.006228268466657028,
          "train_loss": 0.009348686784505844,
          "validation_loss": 0.006419853772968054,
          "train_metric": 0.000941565609537065,
          "validation_metric": 0.000646584143396467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5089957149994007
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.004103624189156108,
          "train_loss": 0.003576316637918353,
          "validation_loss": 0.00284695066511631,
          "train_metric": 0.0003601935168262571,
          "validation_metric": 0.0002867344592232257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.525767242999791
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0034283755740034394,
          "train_loss": 0.005887740757316351,
          "validation_loss": 0.004913967568427324,
          "train_metric": 0.0005929917679168284,
          "validation_metric": 0.0004949168069288135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.542586764000589
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0034671610919758677,
          "train_loss": 0.002304855501279235,
          "validation_loss": 0.0016873340355232358,
          "train_metric": 0.0002321366046089679,
          "validation_metric": 0.0001699420972727239,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.560341739999785
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.002704069975152379,
          "train_loss": 0.004082697443664074,
          "validation_loss": 0.003579318756237626,
          "train_metric": 0.00041119434172287583,
          "validation_metric": 0.0003604958765208721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5775043009998626
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0031789580134500284,
          "train_loss": 0.0023760306648910046,
          "validation_loss": 0.0018197472672909498,
          "train_metric": 0.0002393051254330203,
          "validation_metric": 0.00018327828729525208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5946356989989
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.003971660691604484,
          "train_loss": 0.0027882917784154415,
          "validation_loss": 0.0023017728235572577,
          "train_metric": 0.0002808265562634915,
          "validation_metric": 0.00023182611039374024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6121714359996986
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.003291083972726483,
          "train_loss": 0.0025270399637520313,
          "validation_loss": 0.0020886908750981092,
          "train_metric": 0.0002545142197050154,
          "validation_metric": 0.00021036529506091028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6294551360006153
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0025449169079365674,
          "train_loss": 0.0022497933823615313,
          "validation_loss": 0.001876316498965025,
          "train_metric": 0.0002265909279230982,
          "validation_metric": 0.00018897574045695364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.646637407999151
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0024972992650873493,
          "train_loss": 0.002153246197849512,
          "validation_loss": 0.001594153931364417,
          "train_metric": 0.0002168670907849446,
          "validation_metric": 0.00016055736341513693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.664402061000146
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.002246584597742185,
          "train_loss": 0.002830921672284603,
          "validation_loss": 0.002287196461111307,
          "train_metric": 0.0002851200697477907,
          "validation_metric": 0.00023035804042592645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6815752450002037
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.002593605038782698,
          "train_loss": 0.0021118426229804754,
          "validation_loss": 0.0016635372303426266,
          "train_metric": 0.0002126970503013581,
          "validation_metric": 0.00016754538228269666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6994097599999805
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0031510186217929004,
          "train_loss": 0.002624044194817543,
          "validation_loss": 0.0018387632444500923,
          "train_metric": 0.00026428408455103636,
          "validation_metric": 0.00018519350851420313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7158727690002706
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0027625510047073476,
          "train_loss": 0.0022697411477565765,
          "validation_loss": 0.001722378539852798,
          "train_metric": 0.00022860003809910268,
          "validation_metric": 0.0001734716643113643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7326208369995584
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0033160917519126087,
          "train_loss": 0.002653650473803282,
          "validation_loss": 0.002215032232925296,
          "train_metric": 0.000267265917500481,
          "validation_metric": 0.00022308994084596634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.749385965998954
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.002511357110051904,
          "train_loss": 0.0023278361186385155,
          "validation_loss": 0.0018823802238330245,
          "train_metric": 0.00023445113038178533,
          "validation_metric": 0.00018958644068334252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.766984109999612
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.003506247459881706,
          "train_loss": 0.0033895859960466623,
          "validation_loss": 0.0022423863410949707,
          "train_metric": 0.00034138665068894625,
          "validation_metric": 0.00022584492398891598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7841291900003853
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0029674919205717742,
          "train_loss": 0.008144916966557503,
          "validation_loss": 0.0066173565573990345,
          "train_metric": 0.0008203262113966048,
          "validation_metric": 0.0006664759130217135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.801073444999929
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0031998440390452743,
          "train_loss": 0.001968646189197898,
          "validation_loss": 0.0015453215455636382,
          "train_metric": 0.00019827483629342169,
          "validation_metric": 0.00015563909255433828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.818682761000673
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0021865172784600873,
          "train_loss": 0.0018129736417904496,
          "validation_loss": 0.0013912522699683905,
          "train_metric": 0.0001825960644055158,
          "validation_metric": 0.0001401218178216368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.837450410999736
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0024513482239854056,
          "train_loss": 0.002262311987578869,
          "validation_loss": 0.0019994315225631,
          "train_metric": 0.00022785179316997528,
          "validation_metric": 0.00020137542742304504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8547742139999173
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0024888959342206363,
          "train_loss": 0.0025285331066697836,
          "validation_loss": 0.0019790949299931526,
          "train_metric": 0.00025466459919698536,
          "validation_metric": 0.00019932720169890672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.871483929000533
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0022398457185772713,
          "train_loss": 0.002626333385705948,
          "validation_loss": 0.0022655182983726263,
          "train_metric": 0.0002645147033035755,
          "validation_metric": 0.0002281746856169775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8879212850006297
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.002452572782203788,
          "train_loss": 0.001836198614910245,
          "validation_loss": 0.0014690352836623788,
          "train_metric": 0.00018493521201889962,
          "validation_metric": 0.0001479558413848281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9045015100000455
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0022283794605755247,
          "train_loss": 0.002300068037584424,
          "validation_loss": 0.0016054699663072824,
          "train_metric": 0.00023165441234596074,
          "validation_metric": 0.00016169706941582263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9209236949991464
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.002103007213008823,
          "train_loss": 0.0028266573790460825,
          "validation_loss": 0.00233782222494483,
          "train_metric": 0.0002846905554179102,
          "validation_metric": 0.0002354569296585396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.942509147000237
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0024775326819508336,
          "train_loss": 0.0017721385229378939,
          "validation_loss": 0.0013533185701817274,
          "train_metric": 0.00017848331481218338,
          "validation_metric": 0.00013630131434183568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9600550390005083
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0019697380921570584,
          "train_loss": 0.0019243928836658597,
          "validation_loss": 0.0013920315541327,
          "train_metric": 0.00019381780293770134,
          "validation_metric": 0.0001402003108523786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9764827830003924
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0019699413060152438,
          "train_loss": 0.002462118398398161,
          "validation_loss": 0.0021635456942021847,
          "train_metric": 0.00024797554942779243,
          "validation_metric": 0.0002179044095100835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9932058550002694
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0026716859501902945,
          "train_loss": 0.002299377229064703,
          "validation_loss": 0.0017158512491732836,
          "train_metric": 0.0002315848832949996,
          "validation_metric": 0.00017281425243709236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0099546700002975
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0037436901147884782,
          "train_loss": 0.0029854823369532824,
          "validation_loss": 0.002388652181252837,
          "train_metric": 0.0003006868646480143,
          "validation_metric": 0.00024057630798779428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.02633998499914
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.006622522880206816,
          "train_loss": 0.00392886670306325,
          "validation_loss": 0.00383438216522336,
          "train_metric": 0.00039570106309838593,
          "validation_metric": 0.00038618495455011725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.042198750999887
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.003209460621292237,
          "train_loss": 0.003360713366419077,
          "validation_loss": 0.0022624803241342306,
          "train_metric": 0.0003384787414688617,
          "validation_metric": 0.0002278687315993011,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.058961097000065
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0035642458533402532,
          "train_loss": 0.005642406642436981,
          "validation_loss": 0.004157574847340584,
          "train_metric": 0.0005682825576514006,
          "validation_metric": 0.00041873575537465513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0755908390001423
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.002530596659198636,
          "train_loss": 0.002013087971135974,
          "validation_loss": 0.0017329694237560034,
          "train_metric": 0.00020275085989851505,
          "validation_metric": 0.0001745383342495188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0931002069992246
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0032096905233629514,
          "train_loss": 0.002447322476655245,
          "validation_loss": 0.0016677271341904998,
          "train_metric": 0.0002464853459969163,
          "validation_metric": 0.00016796737327240407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1104766720000043
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0020797251927433535,
          "train_loss": 0.0021595750004053116,
          "validation_loss": 0.0015930652152746916,
          "train_metric": 0.00021750450832769275,
          "validation_metric": 0.0001604477147338912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1270807000000787
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0021043486885901075,
          "train_loss": 0.0026218239217996597,
          "validation_loss": 0.0023277190048247576,
          "train_metric": 0.00026406050892546773,
          "validation_metric": 0.00023443932877853513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.143720903999565
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002858354757336201,
          "train_loss": 0.0035401168279349804,
          "validation_loss": 0.0028613973408937454,
          "train_metric": 0.00035654762177728117,
          "validation_metric": 0.0002881894470192492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.159809729000699
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0031979974555724766,
          "train_loss": 0.0018082292517647147,
          "validation_loss": 0.0013486590469256043,
          "train_metric": 0.00018211820861324668,
          "validation_metric": 0.00013583200052380562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1774406599997747
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0017721012354741106,
          "train_loss": 0.002631650771945715,
          "validation_loss": 0.001973874168470502,
          "train_metric": 0.00026505024288780987,
          "validation_metric": 0.00019880139734596014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.194280780000554
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0034224333430756815,
          "train_loss": 0.004744780715554953,
          "validation_loss": 0.003866693004965782,
          "train_metric": 0.0004778769507538527,
          "validation_metric": 0.0003894392284564674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2110756719994242
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.00375221781541768,
          "train_loss": 0.006881888955831528,
          "validation_loss": 0.005438690539449453,
          "train_metric": 0.0006931186653673649,
          "validation_metric": 0.0005477649974636734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.227637912999853
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0046817812326480635,
          "train_loss": 0.002332257339730859,
          "validation_loss": 0.0017482249531894922,
          "train_metric": 0.00023489640443585813,
          "validation_metric": 0.00017607481277082115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2443885960001353
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.004030474647152005,
          "train_loss": 0.003308525076135993,
          "validation_loss": 0.0022299669217318296,
          "train_metric": 0.0003332225314807147,
          "validation_metric": 0.0002245941141154617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2605508840006223
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0024181636872526724,
          "train_loss": 0.0036386882420629263,
          "validation_loss": 0.002591398311778903,
          "train_metric": 0.0003664753749035299,
          "validation_metric": 0.0002609961375128478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2769961899994087
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0047239330742741,
          "train_loss": 0.0081107709556818,
          "validation_loss": 0.005827875807881355,
          "train_metric": 0.0008168871281668544,
          "validation_metric": 0.0005869623273611069,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2936289320005017
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.005417836306151003,
          "train_loss": 0.0037393802776932716,
          "validation_loss": 0.0031499015167355537,
          "train_metric": 0.0003766166919376701,
          "validation_metric": 0.0003172465949319303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3097906849998253
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.002834818500559777,
          "train_loss": 0.002181644318625331,
          "validation_loss": 0.0014685823116451502,
          "train_metric": 0.00021972724061924964,
          "validation_metric": 0.00014791025023441762,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3260077369995997
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002087031982227927,
          "train_loss": 0.001975832972675562,
          "validation_loss": 0.0015644789673388004,
          "train_metric": 0.00019899864855688065,
          "validation_metric": 0.00015756858920212835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.341814430999875
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.001834216502174968,
          "train_loss": 0.0017870580777525902,
          "validation_loss": 0.001545511418953538,
          "train_metric": 0.000179985974682495,
          "validation_metric": 0.00015565825742669404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.358172370999455
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0020187367845210247,
          "train_loss": 0.0015709433937445283,
          "validation_loss": 0.0011798959458246827,
          "train_metric": 0.0001582196564413607,
          "validation_metric": 0.00011883481056429446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3758011240006454
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0020420363853190793,
          "train_loss": 0.0016756566474214196,
          "validation_loss": 0.0011260653845965862,
          "train_metric": 0.0001687660114839673,
          "validation_metric": 0.00011341318895574659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3935345390000293
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.003535700117936358,
          "train_loss": 0.002009452786296606,
          "validation_loss": 0.0014825722901150584,
          "train_metric": 0.00020238474826328456,
          "validation_metric": 0.00014931925397831947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.410378327000217
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0030268881891970523,
          "train_loss": 0.002030134666711092,
          "validation_loss": 0.0013721196446567774,
          "train_metric": 0.00020446772396098822,
          "validation_metric": 0.00013819486775901169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4272558169996046
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0017708564228087198,
          "train_loss": 0.0018044731114059687,
          "validation_loss": 0.0013847390655428171,
          "train_metric": 0.00018173991702497005,
          "validation_metric": 0.00013946584658697248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.444069903000127
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.001806859243515646,
          "train_loss": 0.0013815542915835977,
          "validation_loss": 0.001103834481909871,
          "train_metric": 0.00013914510782342404,
          "validation_metric": 0.00011117416579509154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4616347609990044
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0020158611114311498,
          "train_loss": 0.0015769845340400934,
          "validation_loss": 0.0012437945697456598,
          "train_metric": 0.00015882811567280442,
          "validation_metric": 0.00012527043872978538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4786216119991877
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0021543320144701283,
          "train_loss": 0.0015498747816309333,
          "validation_loss": 0.001152456970885396,
          "train_metric": 0.00015609772526659071,
          "validation_metric": 0.00011607125634327531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4954884399994626
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0024223790314863436,
          "train_loss": 0.004564328119158745,
          "validation_loss": 0.003460861509665847,
          "train_metric": 0.00045970239443704486,
          "validation_metric": 0.0003485652559902519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.512435311000445
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0035447640184429474,
          "train_loss": 0.0032158696558326483,
          "validation_loss": 0.0025482182390987873,
          "train_metric": 0.00032389062107540667,
          "validation_metric": 0.000256647210335359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5290816359993187
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0025967639085138217,
          "train_loss": 0.0015891505172476172,
          "validation_loss": 0.001182785490527749,
          "train_metric": 0.00016005341603886336,
          "validation_metric": 0.00011912581248907372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5458653690002393
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0034798845463228645,
          "train_loss": 0.003918919712305069,
          "validation_loss": 0.0036831495817750692,
          "train_metric": 0.0003946992219425738,
          "validation_metric": 0.00037095329025760293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.562995587999467
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.002709730433707591,
          "train_loss": 0.0019259070977568626,
          "validation_loss": 0.0013645259896293283,
          "train_metric": 0.00019397032156120986,
          "validation_metric": 0.00013743007730226964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5792228400005115
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.002660632897459436,
          "train_loss": 0.0019747840706259012,
          "validation_loss": 0.0017572782235220075,
          "train_metric": 0.00019889301620423794,
          "validation_metric": 0.00017698665033094585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.691419673999917
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0016582533025939483,
          "train_loss": 0.0019085097592324018,
          "validation_loss": 0.0015391088090837002,
          "train_metric": 0.00019221811089664698,
          "validation_metric": 0.0001550134038552642,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.708357093999439
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0015420671461470192,
          "train_loss": 0.0019329902715981007,
          "validation_loss": 0.0015712087042629719,
          "train_metric": 0.00019468370010145009,
          "validation_metric": 0.00015824638830963522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.724794316000043
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0022422104921133723,
          "train_loss": 0.001971279736608267,
          "validation_loss": 0.0016593592008575797,
          "train_metric": 0.00019854010315611959,
          "validation_metric": 0.00016712455544620752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.741703077999773
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0032978640447254293,
          "train_loss": 0.0035676260013133287,
          "validation_loss": 0.0025643666740506887,
          "train_metric": 0.00035931821912527084,
          "validation_metric": 0.0002582736196927726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7580681939998613
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.003959945446695201,
          "train_loss": 0.001483625266700983,
          "validation_loss": 0.0010815274436026812,
          "train_metric": 0.0001494253083365038,
          "validation_metric": 0.00010892748832702637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.775395140999535
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0020421499903022777,
          "train_loss": 0.0013110657455399632,
          "validation_loss": 0.0009854331146925688,
          "train_metric": 0.00013204575225245208,
          "validation_metric": 9.924921323545277e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7928623299994797
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.001717893552267924,
          "train_loss": 0.0020734064746648073,
          "validation_loss": 0.0013545278925448656,
          "train_metric": 0.0002088258770527318,
          "validation_metric": 0.00013642308476846665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.809655892999217
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.005369386104575824,
          "train_loss": 0.0035192465875297785,
          "validation_loss": 0.002790774218738079,
          "train_metric": 0.0003544455976225436,
          "validation_metric": 0.0002810765290632844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8363655409993953
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0036638795645558275,
          "train_loss": 0.002217271365225315,
          "validation_loss": 0.001830258290283382,
          "train_metric": 0.00022331545187626034,
          "validation_metric": 0.00018433692457620054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8720844319996104
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0018136844228138216,
          "train_loss": 0.0014203599421307445,
          "validation_loss": 0.0011738934554159641,
          "train_metric": 0.00014305344666354358,
          "validation_metric": 0.00011823025852208957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8886413109994464
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.001522813065093942,
          "train_loss": 0.0013901959173381329,
          "validation_loss": 0.0010826600482687354,
          "train_metric": 0.00014001545787323266,
          "validation_metric": 0.00010904156806645915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.905051481000555
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0016834964735608082,
          "train_loss": 0.001982133137062192,
          "validation_loss": 0.0013388218358159065,
          "train_metric": 0.00019963315571658313,
          "validation_metric": 0.00013484121882356703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.92149124900061
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0016522817040822702,
          "train_loss": 0.0020323272328823805,
          "validation_loss": 0.0018227396067231894,
          "train_metric": 0.00020468856382649392,
          "validation_metric": 0.00018357965745963156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9377751179999905
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.002382092676270986,
          "train_loss": 0.0051719737239181995,
          "validation_loss": 0.003412492573261261,
          "train_metric": 0.0005209022201597691,
          "validation_metric": 0.0003436937986407429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9539121109992266
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0029842510921298526,
          "train_loss": 0.002651164773851633,
          "validation_loss": 0.0023358447942882776,
          "train_metric": 0.0002670156245585531,
          "validation_metric": 0.00023525772849097848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9707083989997045
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.002159293890144909,
          "train_loss": 0.001817540731281042,
          "validation_loss": 0.0012760567478835583,
          "train_metric": 0.00018305605044588447,
          "validation_metric": 0.00012851975043304265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.987598693000109
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0016644451843603747,
          "train_loss": 0.001340570510365069,
          "validation_loss": 0.0010391264222562313,
          "train_metric": 0.00013501734065357596,
          "validation_metric": 0.00010465701052453369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.004500015998929
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0017403272959199967,
          "train_loss": 0.0018654080340638757,
          "validation_loss": 0.0012096527498215437,
          "train_metric": 0.00018787705630529672,
          "validation_metric": 0.00012183179205749184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.021108812999955
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0025487070060989936,
          "train_loss": 0.00934937596321106,
          "validation_loss": 0.0072653889656066895,
          "train_metric": 0.0009416348184458911,
          "validation_metric": 0.0007317432900890708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.037799469000674
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.005651771673001349,
          "train_loss": 0.006730910390615463,
          "validation_loss": 0.005228749010711908,
          "train_metric": 0.0006779127288609743,
          "validation_metric": 0.0005266204243525863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0538904660006665
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0033529339925735258,
          "train_loss": 0.005118172615766525,
          "validation_loss": 0.00416925922036171,
          "train_metric": 0.0005154836107976735,
          "validation_metric": 0.0004199125396553427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.070406008999271
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.0032726377903600223,
          "train_loss": 0.003911915700882673,
          "validation_loss": 0.0023869199212640524,
          "train_metric": 0.0003939938615076244,
          "validation_metric": 0.00024040183052420616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.086944188000416
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0023398790071951225,
          "train_loss": 0.0016701372805982828,
          "validation_loss": 0.001103052170947194,
          "train_metric": 0.00016821011377032846,
          "validation_metric": 0.00011109538172604516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.103566200999921
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.002554344984673662,
          "train_loss": 0.0014720947947353125,
          "validation_loss": 0.0011697988957166672,
          "train_metric": 0.00014826400729361922,
          "validation_metric": 0.00011781786452047527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.119966951999231
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0014726116169185843,
          "train_loss": 0.0018055712571367621,
          "validation_loss": 0.0012845877790823579,
          "train_metric": 0.0001818505406845361,
          "validation_metric": 0.00012937896826770157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.13659680300043
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.0023534670708613703,
          "train_loss": 0.001699766842648387,
          "validation_loss": 0.0012851071078330278,
          "train_metric": 0.00017119428957812488,
          "validation_metric": 0.00012943128240294755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.153240714000276
        }
      ],
      "best": {
        "epoch": 213,
        "updates": 3408
      },
      "best_validation_metric": 9.924921323545277e-05,
      "config": {
        "label": "Batch 64",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009962389935926658
    },
    "be8cbd994d03_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3570044133812189,
          "train_loss": 0.1583603173494339,
          "validation_loss": 0.13871486485004425,
          "train_metric": 0.015949469059705734,
          "validation_metric": 0.01397085189819336,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.017929618999914965
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.10737723065540195,
          "train_loss": 0.10755489766597748,
          "validation_loss": 0.0826863944530487,
          "train_metric": 0.010832535102963448,
          "validation_metric": 0.008327869698405266,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.0358268029995088
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.07111676270142198,
          "train_loss": 0.06138412654399872,
          "validation_loss": 0.057298578321933746,
          "train_metric": 0.006182384677231312,
          "validation_metric": 0.005770903080701828,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.05356908099929569
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.05721504054963589,
          "train_loss": 0.03993847966194153,
          "validation_loss": 0.03575877100229263,
          "train_metric": 0.004022457171231508,
          "validation_metric": 0.003601492615416646,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.07213000299998384
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.04426180652808398,
          "train_loss": 0.04908743500709534,
          "validation_loss": 0.038820404559373856,
          "train_metric": 0.004943906795233488,
          "validation_metric": 0.0039098490960896015,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.08854335999967589
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.03523908933857456,
          "train_loss": 0.035935550928115845,
          "validation_loss": 0.03204776719212532,
          "train_metric": 0.0036192969419062138,
          "validation_metric": 0.003227733541280031,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.10550518600030045
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.026958750444464386,
          "train_loss": 0.017882298678159714,
          "validation_loss": 0.013910208828747272,
          "train_metric": 0.0018010396743193269,
          "validation_metric": 0.0014009852893650532,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.12283023599957232
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.024503783381078392,
          "train_loss": 0.025684261694550514,
          "validation_loss": 0.022527214139699936,
          "train_metric": 0.0025868245866149664,
          "validation_metric": 0.0022688584867864847,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.13932805900003586
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04096715254127048,
          "train_loss": 0.035376690328121185,
          "validation_loss": 0.030411111190915108,
          "train_metric": 0.0035630110651254654,
          "validation_metric": 0.0030628954991698265,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.15596483199988143
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.032272897311486304,
          "train_loss": 0.02791258879005909,
          "validation_loss": 0.0249360129237175,
          "train_metric": 0.00281125376932323,
          "validation_metric": 0.0025114640593528748,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.17299400000047171
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.030163907096721232,
          "train_loss": 0.023815952241420746,
          "validation_loss": 0.01914181560277939,
          "train_metric": 0.0023986557498574257,
          "validation_metric": 0.0019278933759778738,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.18967086500015284
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.02367148592020385,
          "train_loss": 0.01576712727546692,
          "validation_loss": 0.015368327498435974,
          "train_metric": 0.0015880076680332422,
          "validation_metric": 0.0015478417044505477,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.20665262200054713
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.02088934474159032,
          "train_loss": 0.017886176705360413,
          "validation_loss": 0.013514958322048187,
          "train_metric": 0.001801430364139378,
          "validation_metric": 0.0013611770700663328,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2243048580003233
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.030704395612701774,
          "train_loss": 0.042036984115839005,
          "validation_loss": 0.027378752827644348,
          "train_metric": 0.004233811050653458,
          "validation_metric": 0.0027574878185987473,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.24113349599974754
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.030304638377856463,
          "train_loss": 0.02140866219997406,
          "validation_loss": 0.018161026760935783,
          "train_metric": 0.0021562022157013416,
          "validation_metric": 0.0018291119486093521,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2578625269998156
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.02524929732317105,
          "train_loss": 0.02675558812916279,
          "validation_loss": 0.01959868334233761,
          "train_metric": 0.002694724826142192,
          "validation_metric": 0.001973907696083188,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.27492879800047376
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.016398051695432514,
          "train_loss": 0.0116663109511137,
          "validation_loss": 0.011401837691664696,
          "train_metric": 0.0011749881086871028,
          "validation_metric": 0.0011483512353152037,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2921354270001757
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.018761075945803896,
          "train_loss": 0.019352829083800316,
          "validation_loss": 0.012684072367846966,
          "train_metric": 0.0019491462735459208,
          "validation_metric": 0.0012774931965395808,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3086689640003897
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.021219174785073847,
          "train_loss": 0.01698196306824684,
          "validation_loss": 0.014150894246995449,
          "train_metric": 0.0017103611025959253,
          "validation_metric": 0.0014252262189984322,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3255445539998618
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.014972734381444752,
          "train_loss": 0.017037149518728256,
          "validation_loss": 0.01662309281527996,
          "train_metric": 0.0017159194685518742,
          "validation_metric": 0.001674216939136386,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3444018549998873
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.01308911998057738,
          "train_loss": 0.008796585723757744,
          "validation_loss": 0.00900196935981512,
          "train_metric": 0.0008859598310664296,
          "validation_metric": 0.0009066453785635531,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3621898519995739
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.019950860441895202,
          "train_loss": 0.018170654773712158,
          "validation_loss": 0.012309450656175613,
          "train_metric": 0.0018300821539014578,
          "validation_metric": 0.0012397626414895058,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.38026904899925285
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.016994978883303702,
          "train_loss": 0.009374777786433697,
          "validation_loss": 0.007896075025200844,
          "train_metric": 0.0009441934525966644,
          "validation_metric": 0.000795263797044754,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3974949889998243
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.029353133737458847,
          "train_loss": 0.01924525573849678,
          "validation_loss": 0.01207420788705349,
          "train_metric": 0.0019383116159588099,
          "validation_metric": 0.0012160700280219316,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4144528140004695
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.022796692384872586,
          "train_loss": 0.016337521374225616,
          "validation_loss": 0.015751691535115242,
          "train_metric": 0.0016454552533105016,
          "validation_metric": 0.0015864527085795999,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4310612779991061
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.011801033018855378,
          "train_loss": 0.00814996100962162,
          "validation_loss": 0.0069993347860872746,
          "train_metric": 0.0008208341896533966,
          "validation_metric": 0.0007049473351798952,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4484291549997579
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.010899979388341308,
          "train_loss": 0.011979383416473866,
          "validation_loss": 0.009241105988621712,
          "train_metric": 0.0012065195478498936,
          "validation_metric": 0.0009307301370427012,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.46474517400019977
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.009877517077256925,
          "train_loss": 0.011402061209082603,
          "validation_loss": 0.010338934138417244,
          "train_metric": 0.001148373819887638,
          "validation_metric": 0.0010412995470687747,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.48134979200040107
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.008409884903812781,
          "train_loss": 0.0072839464992284775,
          "validation_loss": 0.0066001443192362785,
          "train_metric": 0.0007336124544963241,
          "validation_metric": 0.0006647423142567277,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.49882327299928875
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.006705837746267207,
          "train_loss": 0.009417664259672165,
          "validation_loss": 0.0068937260657548904,
          "train_metric": 0.0009485126938670874,
          "validation_metric": 0.0006943108746781945,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.516233423999438
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.006956709919904824,
          "train_loss": 0.008516316302120686,
          "validation_loss": 0.007544985506683588,
          "train_metric": 0.0008577320841141045,
          "validation_metric": 0.0007599032833240926,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5329368069997145
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.008933209901442751,
          "train_loss": 0.005669438745826483,
          "validation_loss": 0.004488607868552208,
          "train_metric": 0.0005710050463676453,
          "validation_metric": 0.0004520762013271451,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5502541330006352
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.007937370304716751,
          "train_loss": 0.007125966250896454,
          "validation_loss": 0.006026624236255884,
          "train_metric": 0.0007177012739703059,
          "validation_metric": 0.0006069794180803001,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5673002809999161
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.014948041760362685,
          "train_loss": 0.017453955486416817,
          "validation_loss": 0.012580227106809616,
          "train_metric": 0.0017578984843567014,
          "validation_metric": 0.001267034444026649,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5836747249995824
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.016325433185556903,
          "train_loss": 0.008631350472569466,
          "validation_loss": 0.007195524871349335,
          "train_metric": 0.0008693180861882865,
          "validation_metric": 0.000724707031622529,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6000436989997979
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.009262285922886804,
          "train_loss": 0.009794888086616993,
          "validation_loss": 0.007928835228085518,
          "train_metric": 0.0009865052998065948,
          "validation_metric": 0.0007985633565112948,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6167075769990333
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.015945852239383385,
          "train_loss": 0.009740401990711689,
          "validation_loss": 0.0077167656272649765,
          "train_metric": 0.000981017597950995,
          "validation_metric": 0.0007772044627927244,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6331181469995499
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.008914906269637868,
          "train_loss": 0.009694737382233143,
          "validation_loss": 0.008307373151183128,
          "train_metric": 0.0009764184360392392,
          "validation_metric": 0.0008366882684640586,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6497377619998588
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.014407594164367765,
          "train_loss": 0.03153139352798462,
          "validation_loss": 0.027017883956432343,
          "train_metric": 0.00317572639323771,
          "validation_metric": 0.0027211422566324472,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6660089910001261
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.014854804729111493,
          "train_loss": 0.009229973889887333,
          "validation_loss": 0.006682184524834156,
          "train_metric": 0.0009296091739088297,
          "validation_metric": 0.000673005182761699,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6832991079991189
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.007945938472403213,
          "train_loss": 0.0096619026735425,
          "validation_loss": 0.00861677061766386,
          "train_metric": 0.0009731114259921014,
          "validation_metric": 0.0008678495069034398,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7003330259994982
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010936407052213326,
          "train_loss": 0.012785759754478931,
          "validation_loss": 0.010973222553730011,
          "train_metric": 0.0012877348344773054,
          "validation_metric": 0.0011051826877519488,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8251324400007434
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.010864683878025971,
          "train_loss": 0.01050988957285881,
          "validation_loss": 0.006704314611852169,
          "train_metric": 0.0010585174895823002,
          "validation_metric": 0.0006752340123057365,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8420550779992482
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.010503752462682314,
          "train_loss": 0.0059819952584803104,
          "validation_loss": 0.005525302141904831,
          "train_metric": 0.0006024846225045621,
          "validation_metric": 0.0005564881721511483,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8585722440002428
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.01245467786793597,
          "train_loss": 0.011372080072760582,
          "validation_loss": 0.009315514005720615,
          "train_metric": 0.001145354239270091,
          "validation_metric": 0.0009382244315929711,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8756345710007736
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.02359014138346538,
          "train_loss": 0.014555968344211578,
          "validation_loss": 0.013182478956878185,
          "train_metric": 0.0014660239685326815,
          "validation_metric": 0.0013276911340653896,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8982389859993418
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.009362404744024388,
          "train_loss": 0.010870683938264847,
          "validation_loss": 0.006305714603513479,
          "train_metric": 0.0010948554845526814,
          "validation_metric": 0.0006350885378196836,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9201006879993656
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.013401081014308147,
          "train_loss": 0.010144098661839962,
          "validation_loss": 0.008127384819090366,
          "train_metric": 0.0010216764640063047,
          "validation_metric": 0.0008185603655874729,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9404331259993342
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.019464578275801614,
          "train_loss": 0.019846150651574135,
          "validation_loss": 0.018685292452573776,
          "train_metric": 0.001998831517994404,
          "validation_metric": 0.0018819140968844295,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9630367190002289
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.013847390509909019,
          "train_loss": 0.008957591839134693,
          "validation_loss": 0.005711332894861698,
          "train_metric": 0.0009021758451126516,
          "validation_metric": 0.0005752245197072625,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9858366329990531
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007533411248004995,
          "train_loss": 0.0058450703509151936,
          "validation_loss": 0.004743141122162342,
          "train_metric": 0.0005886941216886044,
          "validation_metric": 0.0004777118156198412,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.023325408999881
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.011110487044788897,
          "train_loss": 0.013963946141302586,
          "validation_loss": 0.012482800520956516,
          "train_metric": 0.0014063975540921092,
          "validation_metric": 0.0012572220293805003,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.0538445040001534
        }
      ],
      "best": {
        "epoch": 32,
        "updates": 512
      },
      "best_validation_metric": 0.0004520762013271451,
      "config": {
        "label": "LR 0.03",
        "batch": 64,
        "lr": 0.03,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.02126208365441038
    },
    "1150d4df4a82_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.725183792412281,
          "train_loss": 0.271520733833313,
          "validation_loss": 0.22441114485263824,
          "train_metric": 0.027346575632691383,
          "validation_metric": 0.022601868957281113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06516694499987352
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.19251352921128273,
          "train_loss": 0.13974319398403168,
          "validation_loss": 0.1179056391119957,
          "train_metric": 0.014074422419071198,
          "validation_metric": 0.011875025928020477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11276881800040428
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12273477180860937,
          "train_loss": 0.10032490640878677,
          "validation_loss": 0.08398673683404922,
          "train_metric": 0.010104356333613396,
          "validation_metric": 0.008458836935460567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15935270299996773
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09152746060863137,
          "train_loss": 0.08048801124095917,
          "validation_loss": 0.06612011045217514,
          "train_metric": 0.008106458932161331,
          "validation_metric": 0.006659375037997961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2016805650000606
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07133611291646957,
          "train_loss": 0.06516025960445404,
          "validation_loss": 0.05265787988901138,
          "train_metric": 0.0065627028234303,
          "validation_metric": 0.005303508602082729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22233523999966565
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06069809605833143,
          "train_loss": 0.05711313337087631,
          "validation_loss": 0.047087281942367554,
          "train_metric": 0.005752225872129202,
          "validation_metric": 0.004742457997053862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24287898799957475
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05503121088258922,
          "train_loss": 0.05024784803390503,
          "validation_loss": 0.041228801012039185,
          "train_metric": 0.005060779396444559,
          "validation_metric": 0.004152413923293352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2843936980007129
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04905405070167035,
          "train_loss": 0.04574151337146759,
          "validation_loss": 0.03793175891041756,
          "train_metric": 0.004606917500495911,
          "validation_metric": 0.0038203480653464794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32602703200063843
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04519221524242312,
          "train_loss": 0.05024033412337303,
          "validation_loss": 0.04192633181810379,
          "train_metric": 0.005060022231191397,
          "validation_metric": 0.004222666844725609,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4794907270006661
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.049359321128576994,
          "train_loss": 0.046037573367357254,
          "validation_loss": 0.039142146706581116,
          "train_metric": 0.004636736121028662,
          "validation_metric": 0.003942254465073347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49600818299950333
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04093247506534681,
          "train_loss": 0.03829994797706604,
          "validation_loss": 0.03261260688304901,
          "train_metric": 0.0038574300706386566,
          "validation_metric": 0.003284622449427843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5129863740003202
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.036824317183345556,
          "train_loss": 0.032820843160152435,
          "validation_loss": 0.028341077268123627,
          "train_metric": 0.003305594902485609,
          "validation_metric": 0.002854409161955118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5297494269998424
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03441062942147255,
          "train_loss": 0.03700372204184532,
          "validation_loss": 0.02941253036260605,
          "train_metric": 0.0037268793676048517,
          "validation_metric": 0.0029623224399983883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5460999219994846
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.032163326046429574,
          "train_loss": 0.02908608876168728,
          "validation_loss": 0.024541635066270828,
          "train_metric": 0.0029294448904693127,
          "validation_metric": 0.0024717431515455246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5629309969990572
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03053868602728471,
          "train_loss": 0.02759496122598648,
          "validation_loss": 0.023318752646446228,
          "train_metric": 0.0027792633045464754,
          "validation_metric": 0.0023485796991735697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5796644010006275
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.026482005312573165,
          "train_loss": 0.027983058243989944,
          "validation_loss": 0.023301327601075172,
          "train_metric": 0.002818351611495018,
          "validation_metric": 0.002346824621781707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5966069360001711
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.02441761939553544,
          "train_loss": 0.023983746767044067,
          "validation_loss": 0.02047683112323284,
          "train_metric": 0.002415555063635111,
          "validation_metric": 0.0020623516757041216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6138663750007254
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.023378114216029644,
          "train_loss": 0.02086450532078743,
          "validation_loss": 0.017464669421315193,
          "train_metric": 0.002101396443322301,
          "validation_metric": 0.0017589774215593934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6307511039995006
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.02140462800161913,
          "train_loss": 0.02289021760225296,
          "validation_loss": 0.018739325925707817,
          "train_metric": 0.002305419184267521,
          "validation_metric": 0.001887356280349195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6470868090000295
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02563406975241378,
          "train_loss": 0.02189042419195175,
          "validation_loss": 0.017806872725486755,
          "train_metric": 0.0022047236561775208,
          "validation_metric": 0.001793443108908832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6630843519997143
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023129807494115084,
          "train_loss": 0.020913323387503624,
          "validation_loss": 0.017478393390774727,
          "train_metric": 0.0021063131280243397,
          "validation_metric": 0.00176035996992141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6794988000001467
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02696609584381804,
          "train_loss": 0.026205450296401978,
          "validation_loss": 0.021345825865864754,
          "train_metric": 0.0026393167208880186,
          "validation_metric": 0.002149873413145542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6955282420003641
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.019721148069947958,
          "train_loss": 0.01606700010597706,
          "validation_loss": 0.01406087540090084,
          "train_metric": 0.0016182095278054476,
          "validation_metric": 0.0014161599101498723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.712171192999449
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.017273633362492546,
          "train_loss": 0.014825965277850628,
          "validation_loss": 0.012481079436838627,
          "train_metric": 0.0014932170743122697,
          "validation_metric": 0.0012570485705509782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7292616979993909
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019757675006985664,
          "train_loss": 0.01432263944298029,
          "validation_loss": 0.012306571938097477,
          "train_metric": 0.001442523906007409,
          "validation_metric": 0.0012394727673381567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.746465485999579
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.017510230012703687,
          "train_loss": 0.014158128760755062,
          "validation_loss": 0.01238947082310915,
          "train_metric": 0.0014259549789130688,
          "validation_metric": 0.0012478221906349063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7629288709995308
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015726784709841013,
          "train_loss": 0.013463691808283329,
          "validation_loss": 0.01134172547608614,
          "train_metric": 0.0013560139341279864,
          "validation_metric": 0.001142297056503594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7801421410003968
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.013707112200791016,
          "train_loss": 0.012812243774533272,
          "validation_loss": 0.01123434491455555,
          "train_metric": 0.001290402258746326,
          "validation_metric": 0.0011314821895211935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7974399089998769
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016051386279286817,
          "train_loss": 0.02502411976456642,
          "validation_loss": 0.020168717950582504,
          "train_metric": 0.0025203379336744547,
          "validation_metric": 0.002031319309026003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8133964699991338
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01542237150715664,
          "train_loss": 0.014315691776573658,
          "validation_loss": 0.011095796711742878,
          "train_metric": 0.001441823784261942,
          "validation_metric": 0.001117528066970408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8302823079993686
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.015414124500239268,
          "train_loss": 0.021503446623682976,
          "validation_loss": 0.018202824518084526,
          "train_metric": 0.0021657487377524376,
          "validation_metric": 0.0018333217594772577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8461716660003731
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.020806275395443663,
          "train_loss": 0.017160432413220406,
          "validation_loss": 0.012870152480900288,
          "train_metric": 0.0017283360939472914,
          "validation_metric": 0.0012962345499545336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8619706120007322
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01630209616268985,
          "train_loss": 0.011724606156349182,
          "validation_loss": 0.009741163812577724,
          "train_metric": 0.0011808593990281224,
          "validation_metric": 0.0009810944320634007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8786086529999011
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013805997223244049,
          "train_loss": 0.023595713078975677,
          "validation_loss": 0.019702864810824394,
          "train_metric": 0.0023764735087752342,
          "validation_metric": 0.0019844004418700933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.89485151900044
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.015536964187049307,
          "train_loss": 0.011063741520047188,
          "validation_loss": 0.00977888610213995,
          "train_metric": 0.0011142995208501816,
          "validation_metric": 0.000984893529675901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9109312769996905
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.012316050342633389,
          "train_loss": 0.011551490053534508,
          "validation_loss": 0.0093732550740242,
          "train_metric": 0.0011634236434474587,
          "validation_metric": 0.0009440399007871747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9275748260006367
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.01322234922554344,
          "train_loss": 0.01286236196756363,
          "validation_loss": 0.010114856995642185,
          "train_metric": 0.0012954497942700982,
          "validation_metric": 0.001018731389194727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9434119859997736
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011155385203892365,
          "train_loss": 0.011489743366837502,
          "validation_loss": 0.010611145757138729,
          "train_metric": 0.0011572049697861075,
          "validation_metric": 0.0010687157046049833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9590387169992027
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.011601212609093636,
          "train_loss": 0.014726401306688786,
          "validation_loss": 0.010290934704244137,
          "train_metric": 0.0014831890584900975,
          "validation_metric": 0.0010364654008299112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9751093619997846
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012412100812071003,
          "train_loss": 0.012658714316785336,
          "validation_loss": 0.01127684861421585,
          "train_metric": 0.0012749392772093415,
          "validation_metric": 0.00113576278090477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9912169849994825
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011675719724735245,
          "train_loss": 0.009901653975248337,
          "validation_loss": 0.007692788261920214,
          "train_metric": 0.0009972583502531052,
          "validation_metric": 0.0007747894851490855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0079418149998673
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010013210485340096,
          "train_loss": 0.008797609247267246,
          "validation_loss": 0.007866949774324894,
          "train_metric": 0.0008860630914568901,
          "validation_metric": 0.0007923304219730198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0242123280004307
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009386818463099189,
          "train_loss": 0.013002975843846798,
          "validation_loss": 0.009492841549217701,
          "train_metric": 0.001309612300246954,
          "validation_metric": 0.0009560841717757285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.040184498000599
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01224605915194843,
          "train_loss": 0.009541220031678677,
          "validation_loss": 0.0073880041018128395,
          "train_metric": 0.0009609567932784557,
          "validation_metric": 0.0007440927438437939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0571654599989415
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.01069485695916228,
          "train_loss": 0.01067455392330885,
          "validation_loss": 0.009969848208129406,
          "train_metric": 0.0010751018999144435,
          "validation_metric": 0.0010041267378255725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.073070854999969
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010661405161954463,
          "train_loss": 0.009650606662034988,
          "validation_loss": 0.008344125933945179,
          "train_metric": 0.0009719737572595477,
          "validation_metric": 0.0008403898100368679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0890405190002639
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011100387433543801,
          "train_loss": 0.015353727154433727,
          "validation_loss": 0.010953432880342007,
          "train_metric": 0.0015463712625205517,
          "validation_metric": 0.0011031896574422717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1051081320001686
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010565859702182934,
          "train_loss": 0.008216824382543564,
          "validation_loss": 0.007050653453916311,
          "train_metric": 0.0008275684667751193,
          "validation_metric": 0.0007101160590536892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1220314660004078
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00831423222552985,
          "train_loss": 0.007751990109682083,
          "validation_loss": 0.007047897204756737,
          "train_metric": 0.0007807521033100784,
          "validation_metric": 0.000709838408511132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1451338639999449
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.008484421006869525,
          "train_loss": 0.008210123516619205,
          "validation_loss": 0.006662710104137659,
          "train_metric": 0.0008268934907391667,
          "validation_metric": 0.0006710437010042369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1696940700003324
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009121570168645121,
          "train_loss": 0.008138127624988556,
          "validation_loss": 0.006662833970040083,
          "train_metric": 0.0008196425042115152,
          "validation_metric": 0.0006710562738589942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.18651318800039
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.012653169003897347,
          "train_loss": 0.008039749227464199,
          "validation_loss": 0.006858522072434425,
          "train_metric": 0.0008097341051325202,
          "validation_metric": 0.0006907652714289725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.203364042999965
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008644956105854362,
          "train_loss": 0.008325216360390186,
          "validation_loss": 0.00712231732904911,
          "train_metric": 0.0008384853135794401,
          "validation_metric": 0.0007173336343839765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2196674839997286
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.00779974230681546,
          "train_loss": 0.008252358995378017,
          "validation_loss": 0.007306789513677359,
          "train_metric": 0.0008311473648063838,
          "validation_metric": 0.0007359131122939289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.235965476000274
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.00925846655445639,
          "train_loss": 0.00652270158752799,
          "validation_loss": 0.0057301875203847885,
          "train_metric": 0.0006569424876943231,
          "validation_metric": 0.0005771234282292426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2527552159990591
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.012231946529936977,
          "train_loss": 0.00858000572770834,
          "validation_loss": 0.00622341874986887,
          "train_metric": 0.0008641468011774123,
          "validation_metric": 0.0006267998833209276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2690471049991174
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008446793217444792,
          "train_loss": 0.00689212791621685,
          "validation_loss": 0.006349670700728893,
          "train_metric": 0.0006941498722881079,
          "validation_metric": 0.0006395156960934401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2848496149999846
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.008316159102832898,
          "train_loss": 0.007118558511137962,
          "validation_loss": 0.006148797459900379,
          "train_metric": 0.0007169551681727171,
          "validation_metric": 0.0006192844011820853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3009323799997219
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007200873791589402,
          "train_loss": 0.006566303316503763,
          "validation_loss": 0.005335439462214708,
          "train_metric": 0.0006613340810872614,
          "validation_metric": 0.0005373659078031778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.317564343000413
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006584036775166169,
          "train_loss": 0.006239462643861771,
          "validation_loss": 0.005712360609322786,
          "train_metric": 0.0006284158444032073,
          "validation_metric": 0.0005753280129283667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.33331590600028
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.006660603481577709,
          "train_loss": 0.006314736790955067,
          "validation_loss": 0.005067717283964157,
          "train_metric": 0.0006359972176142037,
          "validation_metric": 0.0005104019655846059,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3499376380004833
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.006349309740471654,
          "train_loss": 0.005507147405296564,
          "validation_loss": 0.004698514938354492,
          "train_metric": 0.0005546596948988736,
          "validation_metric": 0.00047321722377091646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.367234223000196
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.008569663259549998,
          "train_loss": 0.015207894146442413,
          "validation_loss": 0.0099509721621871,
          "train_metric": 0.001531683374196291,
          "validation_metric": 0.0010022253263741732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3836377949992311
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.010702520652557723,
          "train_loss": 0.0073378500528633595,
          "validation_loss": 0.007015361916273832,
          "train_metric": 0.0007390414248220623,
          "validation_metric": 0.0007065616082400084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4058693690003565
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.007163544389186427,
          "train_loss": 0.0053948210552334785,
          "validation_loss": 0.0044744606129825115,
          "train_metric": 0.0005433466285467148,
          "validation_metric": 0.00045065133599564433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4226259869992646
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005669868871336803,
          "train_loss": 0.005300572142004967,
          "validation_loss": 0.0045133717358112335,
          "train_metric": 0.0005338541814126074,
          "validation_metric": 0.00045457022497430444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.438729126999533
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0069226891282596625,
          "train_loss": 0.007420714013278484,
          "validation_loss": 0.006919790990650654,
          "train_metric": 0.0007473871810361743,
          "validation_metric": 0.000696935981977731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.454768616000365
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00731872973847203,
          "train_loss": 0.006452545523643494,
          "validation_loss": 0.0051038870587944984,
          "train_metric": 0.00064987683435902,
          "validation_metric": 0.0005140448338352144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4740237639998668
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.009972294181352481,
          "train_loss": 0.0056060985662043095,
          "validation_loss": 0.004278527572751045,
          "train_metric": 0.0005646257195621729,
          "validation_metric": 0.00043091760016977787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4906507140003669
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.007715228144661523,
          "train_loss": 0.00778129231184721,
          "validation_loss": 0.007210019044578075,
          "train_metric": 0.0007837032317183912,
          "validation_metric": 0.0007261668215505779,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.506562106000274
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006288037599006202,
          "train_loss": 0.005706281866878271,
          "validation_loss": 0.00420799758285284,
          "train_metric": 0.0005747157265432179,
          "validation_metric": 0.00042381405364722013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5233965530005662
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00576221395022003,
          "train_loss": 0.005137464962899685,
          "validation_loss": 0.004070205148309469,
          "train_metric": 0.000517426582518965,
          "validation_metric": 0.0004099361540284008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.540558847000284
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.005247840948868543,
          "train_loss": 0.006295354105532169,
          "validation_loss": 0.005427327938377857,
          "train_metric": 0.0006340449908748269,
          "validation_metric": 0.0005466206348501146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5569560099993396
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005847955937497318,
          "train_loss": 0.0056092459708452225,
          "validation_loss": 0.004303757566958666,
          "train_metric": 0.0005649426602758467,
          "validation_metric": 0.0004334586556069553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5729046750002453
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00761535503261257,
          "train_loss": 0.0046490756794810295,
          "validation_loss": 0.004038749262690544,
          "train_metric": 0.00046823782031424344,
          "validation_metric": 0.00040676796925254166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5894805719999567
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0055943846455193125,
          "train_loss": 0.005045777186751366,
          "validation_loss": 0.003924061544239521,
          "train_metric": 0.000508192228153348,
          "validation_metric": 0.0003952171537093818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6064609149998432
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.00631263930699788,
          "train_loss": 0.005989368073642254,
          "validation_loss": 0.005381503142416477,
          "train_metric": 0.0006032272358424962,
          "validation_metric": 0.0005420053494162858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6224395640001603
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.005129424927872606,
          "train_loss": 0.004593792837113142,
          "validation_loss": 0.0038168870378285646,
          "train_metric": 0.00046266999561339617,
          "validation_metric": 0.0003844228631351143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6397523029991135
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00618596728600096,
          "train_loss": 0.004681388847529888,
          "validation_loss": 0.004123073071241379,
          "train_metric": 0.0004714923270512372,
          "validation_metric": 0.00041526081622578204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6562634729998535
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004919681974570267,
          "train_loss": 0.004786643199622631,
          "validation_loss": 0.0038449480198323727,
          "train_metric": 0.00048209322267211974,
          "validation_metric": 0.00038724910700693727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6730892599989602
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006906665556016378,
          "train_loss": 0.012297742068767548,
          "validation_loss": 0.009444266557693481,
          "train_metric": 0.0012385835871100426,
          "validation_metric": 0.0009511919342912734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6890348630004155
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.010686725392588414,
          "train_loss": 0.0064812321215868,
          "validation_loss": 0.0057849641889333725,
          "train_metric": 0.00065276597160846,
          "validation_metric": 0.0005826404667459428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7054488369994942
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.007700789232330862,
          "train_loss": 0.004105992149561644,
          "validation_loss": 0.003409368684515357,
          "train_metric": 0.0004135404305998236,
          "validation_metric": 0.0003433791280258447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.722061496000606
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.004603917019267101,
          "train_loss": 0.004845749586820602,
          "validation_loss": 0.0035455594770610332,
          "train_metric": 0.0004880461492575705,
          "validation_metric": 0.00035709579242393374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7385240189996694
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.007054458052152768,
          "train_loss": 0.015696968883275986,
          "validation_loss": 0.012972714379429817,
          "train_metric": 0.0015809413744136691,
          "validation_metric": 0.0013065641978755593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7548241879994748
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.011451014113845304,
          "train_loss": 0.020557669922709465,
          "validation_loss": 0.013468781486153603,
          "train_metric": 0.0020704930648207664,
          "validation_metric": 0.0013565262779593468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7706254120002995
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.015485930867725983,
          "train_loss": 0.00581009965389967,
          "validation_loss": 0.004421717952936888,
          "train_metric": 0.0005851719761267304,
          "validation_metric": 0.0004453392466530204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.786171302999719
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.006737215720931999,
          "train_loss": 0.0061674462631344795,
          "validation_loss": 0.0055349646136164665,
          "train_metric": 0.0006211625295691192,
          "validation_metric": 0.0005574613460339606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.802232379999623
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005034035624703392,
          "train_loss": 0.0039005510043352842,
          "validation_loss": 0.0032008232083171606,
          "train_metric": 0.00039284920785576105,
          "validation_metric": 0.000322375213727355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8189720310001576
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004518714900768828,
          "train_loss": 0.0038000443018972874,
          "validation_loss": 0.0031561877112835646,
          "train_metric": 0.0003827265463769436,
          "validation_metric": 0.0003178796614520252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8364899699990929
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.006248384044738486,
          "train_loss": 0.005066498648375273,
          "validation_loss": 0.004222129005938768,
          "train_metric": 0.0005102792056277394,
          "validation_metric": 0.00042523740557953715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8527870420002728
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.005300535114656668,
          "train_loss": 0.004341588355600834,
          "validation_loss": 0.0038378548342734575,
          "train_metric": 0.0004372688417788595,
          "validation_metric": 0.00038653466617688537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8687205679998442
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004943848747643642,
          "train_loss": 0.0077809556387364864,
          "validation_loss": 0.005293975584208965,
          "train_metric": 0.0007836693548597395,
          "validation_metric": 0.0005331897991709411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8848028539996449
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004287212868803181,
          "train_loss": 0.008774102665483952,
          "validation_loss": 0.007212408818304539,
          "train_metric": 0.0008836955530568957,
          "validation_metric": 0.0007264073938131332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.90120210600071
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0035014182722079568,
          "train_loss": 0.004472360014915466,
          "validation_loss": 0.004246356897056103,
          "train_metric": 0.00045043969294056296,
          "validation_metric": 0.0004276775289326906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9179141839995282
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004747023289382923,
          "train_loss": 0.00342956674285233,
          "validation_loss": 0.0028597929049283266,
          "train_metric": 0.000345413398463279,
          "validation_metric": 0.00028802783344872296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9349904430000606
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005751226031861734,
          "train_loss": 0.006062327418476343,
          "validation_loss": 0.004158920142799616,
          "train_metric": 0.0006105754291638732,
          "validation_metric": 0.0004188711754977703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9518379080000159
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.006527146659209393,
          "train_loss": 0.004064810462296009,
          "validation_loss": 0.003415434155613184,
          "train_metric": 0.000409392814617604,
          "validation_metric": 0.00034399001742713153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9684718089993112
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.005003856465918943,
          "train_loss": 0.0043891118839383125,
          "validation_loss": 0.0036584429908543825,
          "train_metric": 0.00044205531594343483,
          "validation_metric": 0.000368465029168874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9851106089990935
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.005998564207402524,
          "train_loss": 0.004563015885651112,
          "validation_loss": 0.004187696613371372,
          "train_metric": 0.00045957029215060174,
          "validation_metric": 0.0004217694513499737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.001916937000715
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005680662710801698,
          "train_loss": 0.004552732687443495,
          "validation_loss": 0.0037168257404118776,
          "train_metric": 0.0004585345450323075,
          "validation_metric": 0.0003743450506590307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0185530340004334
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.006389895494066877,
          "train_loss": 0.004864814691245556,
          "validation_loss": 0.004019264131784439,
          "train_metric": 0.000489966303575784,
          "validation_metric": 0.00040480552706867456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0358792709994304
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.007791414565872401,
          "train_loss": 0.005660008639097214,
          "validation_loss": 0.004888178315013647,
          "train_metric": 0.000570055388379842,
          "validation_metric": 0.0004923194646835327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0527536810004676
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0048040308320196345,
          "train_loss": 0.0035663405433297157,
          "validation_loss": 0.0027851785998791456,
          "train_metric": 0.0003591887070797384,
          "validation_metric": 0.00028051293338648975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.070050996999271
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004540373971394729,
          "train_loss": 0.004792868625372648,
          "validation_loss": 0.003553792368620634,
          "train_metric": 0.0004827201773878187,
          "validation_metric": 0.0003579249605536461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0865510299991
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004516240820521489,
          "train_loss": 0.0034971204586327076,
          "validation_loss": 0.00261951913125813,
          "train_metric": 0.00035221720463596284,
          "validation_metric": 0.0002638284058775753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1041244080006436
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0035208531917305663,
          "train_loss": 0.00888960249722004,
          "validation_loss": 0.0073181455954909325,
          "train_metric": 0.0008953282376751304,
          "validation_metric": 0.0007370567764155567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1203946309997264
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.004127585325477412,
          "train_loss": 0.005249641835689545,
          "validation_loss": 0.0033487235195934772,
          "train_metric": 0.0005287246312946081,
          "validation_metric": 0.00033727110712789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1367001790003997
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.004842965594434645,
          "train_loss": 0.002905305940657854,
          "validation_loss": 0.0024432791396975517,
          "train_metric": 0.00029261180316098034,
          "validation_metric": 0.0002460781252011657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.153982030999032
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.005225695929766516,
          "train_loss": 0.0038044615648686886,
          "validation_loss": 0.0030250484123826027,
          "train_metric": 0.0003831714275293052,
          "validation_metric": 0.0003046718193218112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1710557410006004
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004379622194392141,
          "train_loss": 0.004060350824147463,
          "validation_loss": 0.0038772965781390667,
          "train_metric": 0.00040894359699450433,
          "validation_metric": 0.00039050710620358586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1873139349991106
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004134063194214832,
          "train_loss": 0.004081107676029205,
          "validation_loss": 0.0038114036433398724,
          "train_metric": 0.00041103424155153334,
          "validation_metric": 0.0003838706179521978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.20354847300041
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0034449776394467335,
          "train_loss": 0.002791373757645488,
          "validation_loss": 0.00257500889711082,
          "train_metric": 0.0002811369486153126,
          "validation_metric": 0.00025934548466466367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.220341140999153
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.002757111069513485,
          "train_loss": 0.004488422069698572,
          "validation_loss": 0.0030822553671896458,
          "train_metric": 0.0004520574293565005,
          "validation_metric": 0.0003104334173258394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2367998930003523
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004882394205196761,
          "train_loss": 0.005478053819388151,
          "validation_loss": 0.004331367090344429,
          "train_metric": 0.0005517295212484896,
          "validation_metric": 0.0004362394392956048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2539337950001936
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0031214265327434987,
          "train_loss": 0.0033513247035443783,
          "validation_loss": 0.002528879791498184,
          "train_metric": 0.0003375331871211529,
          "validation_metric": 0.00025469952379353344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.271008924999478
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003535403222485911,
          "train_loss": 0.003856767201796174,
          "validation_loss": 0.0025257794186472893,
          "train_metric": 0.0003884395118802786,
          "validation_metric": 0.0002543872396927327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.288462585000161
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003580433942261152,
          "train_loss": 0.0028860922902822495,
          "validation_loss": 0.0023965039290487766,
          "train_metric": 0.0002906766312662512,
          "validation_metric": 0.00024136710271704942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.308082290999664
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0032227757619693875,
          "train_loss": 0.0024330851156264544,
          "validation_loss": 0.001969155389815569,
          "train_metric": 0.000245051400270313,
          "validation_metric": 0.00019832611724268645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3263789949996863
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.00312216387828812,
          "train_loss": 0.002400434808805585,
          "validation_loss": 0.001938469591550529,
          "train_metric": 0.00024176300212275237,
          "validation_metric": 0.00019523555238265544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.344250281999848
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.005049306331784464,
          "train_loss": 0.00326095730997622,
          "validation_loss": 0.0026451763696968555,
          "train_metric": 0.0003284316626377404,
          "validation_metric": 0.00026641247677616775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.360301819999222
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0034608831701916642,
          "train_loss": 0.0028723659925162792,
          "validation_loss": 0.0023677286226302385,
          "train_metric": 0.00028929419931955636,
          "validation_metric": 0.0002384689578320831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3764631359990744
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0028234652345418,
          "train_loss": 0.0027469599153846502,
          "validation_loss": 0.0020657279528677464,
          "train_metric": 0.0002766637480817735,
          "validation_metric": 0.00020805251551792026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3928054989992233
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0026031358720501885,
          "train_loss": 0.002991296350955963,
          "validation_loss": 0.0021256254985928535,
          "train_metric": 0.00030127246282063425,
          "validation_metric": 0.00021408521570265293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4094479830000637
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0029182495054556057,
          "train_loss": 0.002263699658215046,
          "validation_loss": 0.0020225406624376774,
          "train_metric": 0.00022799152065999806,
          "validation_metric": 0.00020370287529658526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.426535746999434
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.00309807939993334,
          "train_loss": 0.002908478956669569,
          "validation_loss": 0.0021945068147033453,
          "train_metric": 0.0002929313632193953,
          "validation_metric": 0.00022102269576862454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.444012722999105
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.00266599341557594,
          "train_loss": 0.0024597709998488426,
          "validation_loss": 0.0018571208929643035,
          "train_metric": 0.0002477391099091619,
          "validation_metric": 0.00018704244575928897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4617029830005777
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.004796148758032359,
          "train_loss": 0.004223702475428581,
          "validation_loss": 0.0028237407095730305,
          "train_metric": 0.0004253957886248827,
          "validation_metric": 0.00028439678135327995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.478684878000422
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.009019669574627187,
          "train_loss": 0.006652167998254299,
          "validation_loss": 0.005630786530673504,
          "train_metric": 0.0006699820514768362,
          "validation_metric": 0.0005671122344210744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4953997989996424
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.010253318818286061,
          "train_loss": 0.0058790575712919235,
          "validation_loss": 0.0038958757650107145,
          "train_metric": 0.0005921171396039426,
          "validation_metric": 0.00039237833698280156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.512399262999679
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0057141728029819205,
          "train_loss": 0.007645321078598499,
          "validation_loss": 0.00648478651419282,
          "train_metric": 0.0007700088317506015,
          "validation_metric": 0.0006531239487230778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5295370029998594
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.008052829813095741,
          "train_loss": 0.016303688287734985,
          "validation_loss": 0.011124902404844761,
          "train_metric": 0.001642047893255949,
          "validation_metric": 0.0011204594047740102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5467680570000084
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.008326188559294678,
          "train_loss": 0.004383569583296776,
          "validation_loss": 0.004234254360198975,
          "train_metric": 0.00044149704626761377,
          "validation_metric": 0.00042645863140933216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.563857117000225
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0033992926000792067,
          "train_loss": 0.0024024329613894224,
          "validation_loss": 0.0019044072832912207,
          "train_metric": 0.0002419642696622759,
          "validation_metric": 0.0001918049092637375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5807014820002223
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.002910113355028443,
          "train_loss": 0.005612793378531933,
          "validation_loss": 0.0055975476279854774,
          "train_metric": 0.0005652999971061945,
          "validation_metric": 0.0005637644208036363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.597438921999128
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.00336274285291438,
          "train_loss": 0.0024300217628479004,
          "validation_loss": 0.0018143948400393128,
          "train_metric": 0.0002447428705636412,
          "validation_metric": 0.00018273918249178678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.622388623000006
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003300657535874052,
          "train_loss": 0.00337297678925097,
          "validation_loss": 0.002545601222664118,
          "train_metric": 0.00033971384982578456,
          "validation_metric": 0.00025638361694291234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6402852009996423
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0036947657208656892,
          "train_loss": 0.002643207088112831,
          "validation_loss": 0.0021333610638976097,
          "train_metric": 0.00026621416327543557,
          "validation_metric": 0.00021486429614014924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.656942010999046
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0027905169190489687,
          "train_loss": 0.0048367613926529884,
          "validation_loss": 0.003739910200238228,
          "train_metric": 0.0004871408746112138,
          "validation_metric": 0.0003766701265703887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6734822990001703
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.005829701800394105,
          "train_loss": 0.010636413469910622,
          "validation_loss": 0.009382311254739761,
          "train_metric": 0.0010712605435401201,
          "validation_metric": 0.0009449520148336887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.689768178999657
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.010631502969772555,
          "train_loss": 0.008176790550351143,
          "validation_loss": 0.0050178635865449905,
          "train_metric": 0.0008235364221036434,
          "validation_metric": 0.0005053807981312275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.706594977998975
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.013285969442222267,
          "train_loss": 0.005460844840854406,
          "validation_loss": 0.004616655874997377,
          "train_metric": 0.0005499962717294693,
          "validation_metric": 0.0004649726615753025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7239815359989734
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.004386250668176217,
          "train_loss": 0.003142002271488309,
          "validation_loss": 0.002617886755615473,
          "train_metric": 0.0003164509544149041,
          "validation_metric": 0.00026366396923549473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.741631131999384
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0034448182450432796,
          "train_loss": 0.0038700609002262354,
          "validation_loss": 0.003437807783484459,
          "train_metric": 0.0003897783753927797,
          "validation_metric": 0.0003462434688117355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.758759092999753
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.004726875435153488,
          "train_loss": 0.0074235061183571815,
          "validation_loss": 0.0051461271941661835,
          "train_metric": 0.0007476683822460473,
          "validation_metric": 0.000518299057148397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.776344464000431
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0032514285885554273,
          "train_loss": 0.0020597921684384346,
          "validation_loss": 0.0017143039731308818,
          "train_metric": 0.00020745473739225417,
          "validation_metric": 0.00017265844508074224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7951777700000093
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0028847064895671792,
          "train_loss": 0.00222614873200655,
          "validation_loss": 0.0017028801376000047,
          "train_metric": 0.00022420956520363688,
          "validation_metric": 0.00017150785424746573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8143280379990756
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0026161493515246548,
          "train_loss": 0.002388717606663704,
          "validation_loss": 0.0022249927278608084,
          "train_metric": 0.0002405828854534775,
          "validation_metric": 0.0002240930771222338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.831329676999303
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0024451421122648753,
          "train_loss": 0.002188307698816061,
          "validation_loss": 0.0019238610984757543,
          "train_metric": 0.00022039833129383624,
          "validation_metric": 0.00019376425188966095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8478350330005924
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.002076177905109944,
          "train_loss": 0.002661527367308736,
          "validation_loss": 0.0017820227658376098,
          "train_metric": 0.000268059317022562,
          "validation_metric": 0.00017947881133295596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.864508766000654
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0022713253238180187,
          "train_loss": 0.0021308159921318293,
          "validation_loss": 0.0016065847594290972,
          "train_metric": 0.000214607993257232,
          "validation_metric": 0.00016180932288989425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8823666669995873
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.002485668381268624,
          "train_loss": 0.0022483482025563717,
          "validation_loss": 0.0015288699651136994,
          "train_metric": 0.00022644540877081454,
          "validation_metric": 0.00015398218238260597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.899968658999569
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0044211139174876735,
          "train_loss": 0.0026093204505741596,
          "validation_loss": 0.0019262847490608692,
          "train_metric": 0.00026280118618160486,
          "validation_metric": 0.00019400837481953204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.916795789000389
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0023705231069470756,
          "train_loss": 0.0022000048775225878,
          "validation_loss": 0.001563854981213808,
          "train_metric": 0.00022157642524689436,
          "validation_metric": 0.00015750573948025703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9332689299990307
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0022128868331492413,
          "train_loss": 0.0028951175045222044,
          "validation_loss": 0.002665025880560279,
          "train_metric": 0.00029158563120290637,
          "validation_metric": 0.00026841164799407125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.95015958499971
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0028586407934199087,
          "train_loss": 0.002053749980404973,
          "validation_loss": 0.0015130533138290048,
          "train_metric": 0.0002068461908493191,
          "validation_metric": 0.0001523892133263871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.967633341999317
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0024523594947822858,
          "train_loss": 0.0020305104553699493,
          "validation_loss": 0.0018025942845270038,
          "train_metric": 0.00020450558804441243,
          "validation_metric": 0.00018155068391934037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9851529659990774
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.002135887203621678,
          "train_loss": 0.001838091528043151,
          "validation_loss": 0.001428671763278544,
          "train_metric": 0.000185125827556476,
          "validation_metric": 0.0001438906037947163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.002869489000659
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0023210191720863804,
          "train_loss": 0.0025264176074415445,
          "validation_loss": 0.0016575446352362633,
          "train_metric": 0.0002544515300542116,
          "validation_metric": 0.00016694182704668492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.019762833999266
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0023631388357898686,
          "train_loss": 0.0028521050699055195,
          "validation_loss": 0.002574087819084525,
          "train_metric": 0.000287253555143252,
          "validation_metric": 0.0002592527016531676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0371513179998146
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.002485522913048044,
          "train_loss": 0.002309029456228018,
          "validation_loss": 0.001501815509982407,
          "train_metric": 0.00023255700943991542,
          "validation_metric": 0.0001512573508080095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0548855470005947
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.002716261169553036,
          "train_loss": 0.002138255164027214,
          "validation_loss": 0.0016314181266352534,
          "train_metric": 0.00021535724226851016,
          "validation_metric": 0.00016431046242360026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.071858966999571
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.00289134705599281,
          "train_loss": 0.0017966300947591662,
          "validation_loss": 0.0015389397740364075,
          "train_metric": 0.0001809499808587134,
          "validation_metric": 0.000154996378114447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0885621060006088
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0024329685511474963,
          "train_loss": 0.0017711227992549539,
          "validation_loss": 0.0013358104042708874,
          "train_metric": 0.00017838101484812796,
          "validation_metric": 0.00013453791325446218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1056945839991386
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0024775342462817207,
          "train_loss": 0.007562030106782913,
          "validation_loss": 0.004347797017544508,
          "train_metric": 0.0007616200600750744,
          "validation_metric": 0.0004378942248877138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1226377919992956
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.003393010872969171,
          "train_loss": 0.004969136323779821,
          "validation_loss": 0.003458495484665036,
          "train_metric": 0.0005004731356166303,
          "validation_metric": 0.0003483269829303026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.139881200999298
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0035988839917990845,
          "train_loss": 0.0024098376743495464,
          "validation_loss": 0.002166891237720847,
          "train_metric": 0.00024271004076581448,
          "validation_metric": 0.00021824135910719633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.15608856200015
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0026283382467227057,
          "train_loss": 0.0037163584493100643,
          "validation_loss": 0.002459761220961809,
          "train_metric": 0.00037429804797284305,
          "validation_metric": 0.0002477381203789264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1728321440004947
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.003228494020731887,
          "train_loss": 0.002183516975492239,
          "validation_loss": 0.002094058785587549,
          "train_metric": 0.0002199158479925245,
          "validation_metric": 0.00021090592781547457,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1894602480006142
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0027687562760547735,
          "train_loss": 0.00206978851929307,
          "validation_loss": 0.002118093427270651,
          "train_metric": 0.00020846155530307442,
          "validation_metric": 0.00021332658070605248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2056628619993717
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.00244362272133003,
          "train_loss": 0.0024799888487905264,
          "validation_loss": 0.0014960987027734518,
          "train_metric": 0.00024977544671855867,
          "validation_metric": 0.00015068158973008394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2220124790001137
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0024938411006587557,
          "train_loss": 0.0029078428633511066,
          "validation_loss": 0.0021193339489400387,
          "train_metric": 0.00029286730568856,
          "validation_metric": 0.00021345156710594893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.238734418999229
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.002222596980573144,
          "train_loss": 0.0017096850788220763,
          "validation_loss": 0.0011506980517879128,
          "train_metric": 0.0001721931912470609,
          "validation_metric": 0.00011589410132728517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.255577854999501
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0020034787467011483,
          "train_loss": 0.003791782772168517,
          "validation_loss": 0.0024487022310495377,
          "train_metric": 0.00038189446786418557,
          "validation_metric": 0.0002466243167873472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2716502629991737
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.002604292611067649,
          "train_loss": 0.001992400037124753,
          "validation_loss": 0.0013801869936287403,
          "train_metric": 0.00020066724391654134,
          "validation_metric": 0.00013900737394578755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.287637174000338
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0015919984834908973,
          "train_loss": 0.002042272360995412,
          "validation_loss": 0.0015085138147696853,
          "train_metric": 0.00020569018670357764,
          "validation_metric": 0.0001519319775979966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.303751256000396
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.002792971406961442,
          "train_loss": 0.0023457002826035023,
          "validation_loss": 0.0019018009770661592,
          "train_metric": 0.00023625030007679015,
          "validation_metric": 0.00019154239271301776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3207745010004146
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0028879059718747158,
          "train_loss": 0.002785990247502923,
          "validation_loss": 0.0017061191610991955,
          "train_metric": 0.00028059474425390363,
          "validation_metric": 0.00017183407908305526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3372318639994774
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0039992633974179626,
          "train_loss": 0.0028155019972473383,
          "validation_loss": 0.001958691980689764,
          "train_metric": 0.00028356711845844984,
          "validation_metric": 0.00019727229664567858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3541305779999675
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0038601312116952613,
          "train_loss": 0.0020218396093696356,
          "validation_loss": 0.0013067623367533088,
          "train_metric": 0.0002036323130596429,
          "validation_metric": 0.00013161230890545994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3713823260004574
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002202932177169714,
          "train_loss": 0.0015829962212592363,
          "validation_loss": 0.001153468736447394,
          "train_metric": 0.00015943360631354153,
          "validation_metric": 0.00011617315612966195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3882870159995946
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.002359665710173431,
          "train_loss": 0.0038902978412806988,
          "validation_loss": 0.0025316239334642887,
          "train_metric": 0.00039181651663966477,
          "validation_metric": 0.0002549758937675506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4051039879996097
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0027126112690893933,
          "train_loss": 0.0029516371432691813,
          "validation_loss": 0.0022234974894672632,
          "train_metric": 0.0002972780494019389,
          "validation_metric": 0.00022394253755919635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4217802080002
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.002648346364367171,
          "train_loss": 0.0014680414460599422,
          "validation_loss": 0.0011522223940119147,
          "train_metric": 0.00014785575331188738,
          "validation_metric": 0.00011604760948102921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.438608952999857
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0019272344979981426,
          "train_loss": 0.0014646945055574179,
          "validation_loss": 0.0010348103241994977,
          "train_metric": 0.00014751864364370704,
          "validation_metric": 0.00010422229388495907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4562167879994377
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.001740282565151574,
          "train_loss": 0.003867876483127475,
          "validation_loss": 0.0030748567078262568,
          "train_metric": 0.0003895583504345268,
          "validation_metric": 0.0003096883010584861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4725913759994
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0018642846553120762,
          "train_loss": 0.002669702982529998,
          "validation_loss": 0.0014482561964541674,
          "train_metric": 0.0002688826934900135,
          "validation_metric": 0.00014586304314434528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4893280630003574
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0022091412465670146,
          "train_loss": 0.00132090097758919,
          "validation_loss": 0.0010026126401498914,
          "train_metric": 0.000133036301122047,
          "validation_metric": 0.00010097946505993605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.507033190000584
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.001993655550904805,
          "train_loss": 0.004154097754508257,
          "validation_loss": 0.002466202713549137,
          "train_metric": 0.0004183854616712779,
          "validation_metric": 0.0002483869029674679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.523516668999946
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.005789592658402398,
          "train_loss": 0.005244763568043709,
          "validation_loss": 0.004753930028527975,
          "train_metric": 0.0005282334168441594,
          "validation_metric": 0.0004787984653376043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.539756961999956
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.004537009765044786,
          "train_loss": 0.0023086173459887505,
          "validation_loss": 0.0021977645810693502,
          "train_metric": 0.0002325154928257689,
          "validation_metric": 0.00022135081235319376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5557255139992776
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002118617449013982,
          "train_loss": 0.0018439913401380181,
          "validation_loss": 0.0013823781628161669,
          "train_metric": 0.00018572005501482636,
          "validation_metric": 0.00013922806829214096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.571762129999115
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0063240105446311645,
          "train_loss": 0.004834144376218319,
          "validation_loss": 0.004053998738527298,
          "train_metric": 0.0004868773394264281,
          "validation_metric": 0.0004083039239048958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5882767389994115
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.008810377592453733,
          "train_loss": 0.006248880177736282,
          "validation_loss": 0.003924248274415731,
          "train_metric": 0.0006293643964454532,
          "validation_metric": 0.0003952359256800264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6048156149990973
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.007666255594813265,
          "train_loss": 0.008354502730071545,
          "validation_loss": 0.00667897704988718,
          "train_metric": 0.0008414349285885692,
          "validation_metric": 0.0006726821884512901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6208195429990155
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.005779173345217714,
          "train_loss": 0.003191235475242138,
          "validation_loss": 0.0027579041197896004,
          "train_metric": 0.0003214095486328006,
          "validation_metric": 0.0002777659974526614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6370441330000176
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0025131213333224878,
          "train_loss": 0.0019074729643762112,
          "validation_loss": 0.001513271825388074,
          "train_metric": 0.00019211368635296822,
          "validation_metric": 0.00015241120127029717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.653670706999037
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.003088339220994385,
          "train_loss": 0.0034172069281339645,
          "validation_loss": 0.003257881850004196,
          "train_metric": 0.0003441685694269836,
          "validation_metric": 0.00032812191057018936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.670392333999189
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.005713075057428796,
          "train_loss": 0.005796504206955433,
          "validation_loss": 0.0038430329877883196,
          "train_metric": 0.0005838025826960802,
          "validation_metric": 0.0003870561777148396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.687379491999309
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0032024843167164363,
          "train_loss": 0.0018125363858416677,
          "validation_loss": 0.001240603276528418,
          "train_metric": 0.00018255203031003475,
          "validation_metric": 0.00012494901602622122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.703846008000255
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.003016645601746859,
          "train_loss": 0.00453146593645215,
          "validation_loss": 0.004193053115159273,
          "train_metric": 0.00045639261952601373,
          "validation_metric": 0.0004223089781589806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.720300828999825
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0029971036783535965,
          "train_loss": 0.004020341672003269,
          "validation_loss": 0.0024494784884154797,
          "train_metric": 0.0004049141425639391,
          "validation_metric": 0.000246702489675954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7368905549992633
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.00235589488511323,
          "train_loss": 0.0017048449954017997,
          "validation_loss": 0.0012626835377886891,
          "train_metric": 0.00017170571663882583,
          "validation_metric": 0.00012717285426333547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.753219825999622
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0028452727401599986,
          "train_loss": 0.0015206175157800317,
          "validation_loss": 0.0011538210092112422,
          "train_metric": 0.00015315102064050734,
          "validation_metric": 0.00011620861914707348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7695468139991135
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.001778132893377915,
          "train_loss": 0.0013422046322375536,
          "validation_loss": 0.0010648226598277688,
          "train_metric": 0.00013518192281480879,
          "validation_metric": 0.00010724503226811066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.786012999999002
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.001539646216770052,
          "train_loss": 0.002451565582305193,
          "validation_loss": 0.0016310777282342315,
          "train_metric": 0.00024691279395483434,
          "validation_metric": 0.00016427617811132222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8026274889998604
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.002069160109385848,
          "train_loss": 0.0017999086994677782,
          "validation_loss": 0.0014677378349006176,
          "train_metric": 0.00018128022202290595,
          "validation_metric": 0.00014782517973799258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9204356230002304
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.001500716622103937,
          "train_loss": 0.001165090361610055,
          "validation_loss": 0.0009577301098033786,
          "train_metric": 0.00011734363215509802,
          "validation_metric": 9.64590726653114e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.938917453999238
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0012462312806746922,
          "train_loss": 0.0012078055879101157,
          "validation_loss": 0.0009420915739610791,
          "train_metric": 0.00012164574582129717,
          "validation_metric": 9.488400974078104e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9568362589998287
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0015528826133959228,
          "train_loss": 0.0012302887625992298,
          "validation_loss": 0.0008866148418746889,
          "train_metric": 0.0001239101547980681,
          "validation_metric": 8.929659088607877e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.974714670999674
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0018140974425477907,
          "train_loss": 0.0030379188247025013,
          "validation_loss": 0.001903293072246015,
          "train_metric": 0.0003059680457226932,
          "validation_metric": 0.00019169269944541156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9914287209994654
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0018797156881191768,
          "train_loss": 0.0026393351145088673,
          "validation_loss": 0.0020763089414685965,
          "train_metric": 0.00026582417194731534,
          "validation_metric": 0.00020911823958158493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.007602228999531
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.001619533770281123,
          "train_loss": 0.0015312281902879477,
          "validation_loss": 0.0012264344841241837,
          "train_metric": 0.00015421969874296337,
          "validation_metric": 0.00012352198245935142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.023953926000104
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0013887197437725263,
          "train_loss": 0.0015659501077607274,
          "validation_loss": 0.0013197255320847034,
          "train_metric": 0.00015771675680298358,
          "validation_metric": 0.00013291792129166424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.040231815999505
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0018667944241315126,
          "train_loss": 0.0011906923027709126,
          "validation_loss": 0.0008893643389455974,
          "train_metric": 0.00011992214422207326,
          "validation_metric": 8.957352110883221e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.05664772199998
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0016443443892057985,
          "train_loss": 0.0032144198194146156,
          "validation_loss": 0.0027647782117128372,
          "train_metric": 0.0003237446362618357,
          "validation_metric": 0.00027845834847539663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0910105149996525
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.002690692439500708,
          "train_loss": 0.0013074286980554461,
          "validation_loss": 0.0009409919730387628,
          "train_metric": 0.0001316794368904084,
          "validation_metric": 9.477326238993555e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.132534012000178
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0022044496472517494,
          "train_loss": 0.0014647165080532432,
          "validation_loss": 0.0011294521391391754,
          "train_metric": 0.00014752087008673698,
          "validation_metric": 0.00011375428584869951,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.175277467998967
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.006401148326403927,
          "train_loss": 0.007007676642388105,
          "validation_loss": 0.004653202835470438,
          "train_metric": 0.0007057875045575202,
          "validation_metric": 0.00046865351032465696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.220475836998958
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.012935809136251919,
          "train_loss": 0.008811923675239086,
          "validation_loss": 0.005613419692963362,
          "train_metric": 0.000887504662387073,
          "validation_metric": 0.0005653630360029638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.262223783000081
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.004336539092037128,
          "train_loss": 0.001828602864407003,
          "validation_loss": 0.0016719965497031808,
          "train_metric": 0.0001841701741795987,
          "validation_metric": 0.0001683973678154871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.286411060998944
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.002135910042852629,
          "train_loss": 0.0012996513396501541,
          "validation_loss": 0.0010780578013509512,
          "train_metric": 0.00013089612184558064,
          "validation_metric": 0.00010857803135877475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.303077508999195
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.0013592706836789148,
          "train_loss": 0.0026916873175650835,
          "validation_loss": 0.0022469633258879185,
          "train_metric": 0.0002710969129111618,
          "validation_metric": 0.0002263059141114354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.328074204999211
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0033625226005824516,
          "train_loss": 0.004325009882450104,
          "validation_loss": 0.0026807733811438084,
          "train_metric": 0.00043559918412938714,
          "validation_metric": 0.0002699976903386414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.363873325999521
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0019646740111056715,
          "train_loss": 0.0016606331337243319,
          "validation_loss": 0.0011869031004607677,
          "train_metric": 0.00016725290333852172,
          "validation_metric": 0.00011954052752116695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.380514712000149
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0026108824167749844,
          "train_loss": 0.0033218779135495424,
          "validation_loss": 0.0028448477387428284,
          "train_metric": 0.0003345673903822899,
          "validation_metric": 0.0002865226415451616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.398353196000244
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.003730219839781057,
          "train_loss": 0.00299733760766685,
          "validation_loss": 0.0021620038896799088,
          "train_metric": 0.0003018809075001627,
          "validation_metric": 0.0002177490823669359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.415112114000294
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.004173698813247029,
          "train_loss": 0.0038332168478518724,
          "validation_loss": 0.002613670425489545,
          "train_metric": 0.00038606757880188525,
          "validation_metric": 0.000263239344349131,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.431836047999241
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.0021333415097615216,
          "train_loss": 0.001521300058811903,
          "validation_loss": 0.0013603053521364927,
          "train_metric": 0.0001532197347842157,
          "validation_metric": 0.00013700498675461859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.44848453799932
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0016451239243906457,
          "train_loss": 0.006430890876799822,
          "validation_loss": 0.004499776288866997,
          "train_metric": 0.0006476957350969315,
          "validation_metric": 0.00045320094795897603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.46480133499972
        }
      ],
      "best": {
        "epoch": 210,
        "updates": 3360
      },
      "best_validation_metric": 8.929659088607877e-05,
      "config": {
        "label": "Batch 64",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009449687343297595
    },
    "be8cbd994d03_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.28720599552616477,
          "train_loss": 0.10027504712343216,
          "validation_loss": 0.06795430928468704,
          "train_metric": 0.010099336504936218,
          "validation_metric": 0.006844108924269676,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.013117929000145523
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07358372444286942,
          "train_loss": 0.05031083524227142,
          "validation_loss": 0.04620775207877159,
          "train_metric": 0.005067123100161552,
          "validation_metric": 0.004653874784708023,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.025745856000867207
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.04152324388269335,
          "train_loss": 0.02450956217944622,
          "validation_loss": 0.024288460612297058,
          "train_metric": 0.002468513324856758,
          "validation_metric": 0.0024462444707751274,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.037927183000647346
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.030721052084118128,
          "train_loss": 0.019903583452105522,
          "validation_loss": 0.01614859327673912,
          "train_metric": 0.0020046161953359842,
          "validation_metric": 0.0016264274017885327,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.050333037001109915
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.03007853135932237,
          "train_loss": 0.01922285556793213,
          "validation_loss": 0.01699378341436386,
          "train_metric": 0.0019360554870218039,
          "validation_metric": 0.0017115514492616057,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.06216217700057314
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.019236974709201604,
          "train_loss": 0.014049062505364418,
          "validation_loss": 0.01285537052899599,
          "train_metric": 0.001414970145560801,
          "validation_metric": 0.0012947458308190107,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.0745495490009489
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.02988182828994468,
          "train_loss": 0.01814492978155613,
          "validation_loss": 0.014583639800548553,
          "train_metric": 0.0018274907488375902,
          "validation_metric": 0.0014688110677525401,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.08634218699990015
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.02921431016875431,
          "train_loss": 0.02494482696056366,
          "validation_loss": 0.017591319978237152,
          "train_metric": 0.0025123516097664833,
          "validation_metric": 0.0017717332812026143,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.09817119699982868
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.01624043204355985,
          "train_loss": 0.01268675271421671,
          "validation_loss": 0.012614678591489792,
          "train_metric": 0.0012777632800862193,
          "validation_metric": 0.001270504086278379,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.11073772700001427
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.011456204825663008,
          "train_loss": 0.008430266752839088,
          "validation_loss": 0.006608125753700733,
          "train_metric": 0.0008490656036883593,
          "validation_metric": 0.0006655462202616036,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.12273648700102058
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.017939994955668226,
          "train_loss": 0.021665014326572418,
          "validation_loss": 0.01686611771583557,
          "train_metric": 0.002182021038606763,
          "validation_metric": 0.0016986934933811426,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.13472347700007958
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.02899911953136325,
          "train_loss": 0.050915367901325226,
          "validation_loss": 0.03300299122929573,
          "train_metric": 0.005128009710460901,
          "validation_metric": 0.0033239400945603848,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.14675585300028615
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.020491767616476864,
          "train_loss": 0.010313348844647408,
          "validation_loss": 0.0076930890791118145,
          "train_metric": 0.0010387225775048137,
          "validation_metric": 0.0007748197531327605,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.1583975570010807
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.010805208366946317,
          "train_loss": 0.007607579231262207,
          "validation_loss": 0.0067960722371935844,
          "train_metric": 0.0007662074640393257,
          "validation_metric": 0.0006844754097983241,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.16995195799972862
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.011422676965594292,
          "train_loss": 0.012221421115100384,
          "validation_loss": 0.008611799217760563,
          "train_metric": 0.00123089668340981,
          "validation_metric": 0.0008673489210195839,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.1825046720005048
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.014641293848399073,
          "train_loss": 0.015165723860263824,
          "validation_loss": 0.009078958071768284,
          "train_metric": 0.0015274363104254007,
          "validation_metric": 0.0009143993374891579,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.1944049850008014
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.011987970123300329,
          "train_loss": 0.007402829825878143,
          "validation_loss": 0.005906306207180023,
          "train_metric": 0.0007455858867615461,
          "validation_metric": 0.0005948615144006908,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.20632181700057117
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.007795702826115303,
          "train_loss": 0.01208858098834753,
          "validation_loss": 0.008384054526686668,
          "train_metric": 0.0012175176525488496,
          "validation_metric": 0.0008444110862910748,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.21814386600090074
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.02450304222293198,
          "train_loss": 0.03168167173862457,
          "validation_loss": 0.024178452789783478,
          "train_metric": 0.003190862014889717,
          "validation_metric": 0.0024351654574275017,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.22913333399992553
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.03533009276725352,
          "train_loss": 0.021493714302778244,
          "validation_loss": 0.016783149912953377,
          "train_metric": 0.002164768520742655,
          "validation_metric": 0.0016903374344110489,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.24062304899962328
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.02198279701406136,
          "train_loss": 0.022090841084718704,
          "validation_loss": 0.01836252771317959,
          "train_metric": 0.0022249084431678057,
          "validation_metric": 0.0018494065152481198,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2521160710002732
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.01384031368070282,
          "train_loss": 0.019328860566020012,
          "validation_loss": 0.016052378341555595,
          "train_metric": 0.0019467318197712302,
          "validation_metric": 0.0016167368739843369,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.26723380499970517
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.034380319935735315,
          "train_loss": 0.0270090252161026,
          "validation_loss": 0.017199797555804253,
          "train_metric": 0.0027202502824366093,
          "validation_metric": 0.0017323006177321076,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2790830060002918
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.02473000477766618,
          "train_loss": 0.014617206528782845,
          "validation_loss": 0.01442934200167656,
          "train_metric": 0.0014721915358677506,
          "validation_metric": 0.001453270553611219,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.290749923000476
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.012536097783595324,
          "train_loss": 0.008472515270113945,
          "validation_loss": 0.006177864037454128,
          "train_metric": 0.0008533207583241165,
          "validation_metric": 0.0006222118972800672,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3036648060005973
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.014900184498401359,
          "train_loss": 0.010290983133018017,
          "validation_loss": 0.008825329132378101,
          "train_metric": 0.0010364701738581061,
          "validation_metric": 0.0008888549637049437,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3154464960007317
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.006562665716046467,
          "train_loss": 0.005829630419611931,
          "validation_loss": 0.004888406954705715,
          "train_metric": 0.0005871389876119792,
          "validation_metric": 0.0004923424567095935,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.33112818700101343
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.03579463259666227,
          "train_loss": 0.054020702838897705,
          "validation_loss": 0.03743675351142883,
          "train_metric": 0.005440767854452133,
          "validation_metric": 0.0037704934366047382,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.34628553699985787
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.048576382687315345,
          "train_loss": 0.045170124620199203,
          "validation_loss": 0.03742726147174835,
          "train_metric": 0.004549369215965271,
          "validation_metric": 0.003769536968320608,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3579398180008866
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.023950692149810493,
          "train_loss": 0.013094546273350716,
          "validation_loss": 0.012978743761777878,
          "train_metric": 0.0013188348384574056,
          "validation_metric": 0.0013071715366095304,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3697909870006697
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008997186436317861,
          "train_loss": 0.007829777896404266,
          "validation_loss": 0.007799853570759296,
          "train_metric": 0.0007885865634307265,
          "validation_metric": 0.0007855726289562881,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.38302900199960277
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.0099111139861634,
          "train_loss": 0.008233767934143543,
          "validation_loss": 0.007921918295323849,
          "train_metric": 0.0008292749989777803,
          "validation_metric": 0.0007978666690178216,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.39476541099975293
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.009982078612665646,
          "train_loss": 0.03548198565840721,
          "validation_loss": 0.021188847720623016,
          "train_metric": 0.003573615802451968,
          "validation_metric": 0.002134063048288226,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4064888400007476
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.032195260166190565,
          "train_loss": 0.03886880725622177,
          "validation_loss": 0.024918075650930405,
          "train_metric": 0.003914723638445139,
          "validation_metric": 0.0025096575263887644,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4180189150010847
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.02292870939709246,
          "train_loss": 0.011712197214365005,
          "validation_loss": 0.010247309692203999,
          "train_metric": 0.0011796095641329885,
          "validation_metric": 0.0010320714209228754,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4296816570004012
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.00946548931824509,
          "train_loss": 0.006585792638361454,
          "validation_loss": 0.00526101840659976,
          "train_metric": 0.0006632969016209245,
          "validation_metric": 0.0005298705073073506,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4419972300001973
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.00687539171485696,
          "train_loss": 0.004304701928049326,
          "validation_loss": 0.0035073189064860344,
          "train_metric": 0.0004335537669248879,
          "validation_metric": 0.000353244278812781,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.45465298899944173
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.004790732622495852,
          "train_loss": 0.003941986244171858,
          "validation_loss": 0.0032377822790294886,
          "train_metric": 0.0003970224061049521,
          "validation_metric": 0.0003260975645389408,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4796032960002776
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01523710141191259,
          "train_loss": 0.04167903959751129,
          "validation_loss": 0.02493266947567463,
          "train_metric": 0.0041977595537900925,
          "validation_metric": 0.0025111271534115076,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.49112676799995825
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.027572898456128314,
          "train_loss": 0.025668809190392494,
          "validation_loss": 0.023735934868454933,
          "train_metric": 0.0025852685794234276,
          "validation_metric": 0.0023905965499579906,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5026523140004429
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.01213258811912965,
          "train_loss": 0.006195357069373131,
          "validation_loss": 0.005203015636652708,
          "train_metric": 0.0006239736103452742,
          "validation_metric": 0.0005240287282504141,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5151278049997927
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.007967685291077942,
          "train_loss": 0.005827602464705706,
          "validation_loss": 0.00463262852281332,
          "train_metric": 0.0005869347369298339,
          "validation_metric": 0.0004665814049076289,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5269688190001034
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.012229986146849114,
          "train_loss": 0.019719503819942474,
          "validation_loss": 0.015394461341202259,
          "train_metric": 0.0019860758911818266,
          "validation_metric": 0.001550473622046411,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5384063700003026
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014187382854288444,
          "train_loss": 0.015036116354167461,
          "validation_loss": 0.01136680319905281,
          "train_metric": 0.0015143825439736247,
          "validation_metric": 0.001144822919741273,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5618846050001594
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.009103457115998026,
          "train_loss": 0.010499464347958565,
          "validation_loss": 0.009565547108650208,
          "train_metric": 0.0010574676562100649,
          "validation_metric": 0.0009634069283492863,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5871007910009212
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.006742212557583116,
          "train_loss": 0.0035566245205700397,
          "validation_loss": 0.003232615767046809,
          "train_metric": 0.00035821020719595253,
          "validation_metric": 0.0003255772462580353,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.619072571000288
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.006809061858803034,
          "train_loss": 0.007267671637237072,
          "validation_loss": 0.005217796191573143,
          "train_metric": 0.0007319732685573399,
          "validation_metric": 0.0005255173309706151,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6392801120000513
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.005300939876178745,
          "train_loss": 0.004363919608294964,
          "validation_loss": 0.0028659459203481674,
          "train_metric": 0.00043951792758889496,
          "validation_metric": 0.0002886475413106382,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6519175510002242
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.0063775235175853595,
          "train_loss": 0.006212638691067696,
          "validation_loss": 0.0046859001740813255,
          "train_metric": 0.0006257142522372305,
          "validation_metric": 0.00047194669605232775,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6635100250005053
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009290229179896414,
          "train_loss": 0.005882439203560352,
          "validation_loss": 0.0035485061816871166,
          "train_metric": 0.0005924576544202864,
          "validation_metric": 0.0003573925350792706,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6755409729994426
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007678797483094968,
          "train_loss": 0.00825569685548544,
          "validation_loss": 0.00746193528175354,
          "train_metric": 0.0008314835140481591,
          "validation_metric": 0.0007515387260355055,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6871433159994922
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.006898298132000491,
          "train_loss": 0.008538627065718174,
          "validation_loss": 0.007948487997055054,
          "train_metric": 0.0008599793072789907,
          "validation_metric": 0.0008005426498129964,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6988074430009874
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.012449757312424481,
          "train_loss": 0.012699867598712444,
          "validation_loss": 0.00751863420009613,
          "train_metric": 0.0012790842447429895,
          "validation_metric": 0.0007572493050247431,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7107023770004162
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.00757809889910277,
          "train_loss": 0.008869035169482231,
          "validation_loss": 0.006452339235693216,
          "train_metric": 0.0008932567434385419,
          "validation_metric": 0.000649855996016413,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7220526260007318
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.005471556789416354,
          "train_loss": 0.00942834373563528,
          "validation_loss": 0.006418560165911913,
          "train_metric": 0.0009495881968177855,
          "validation_metric": 0.0006464537582360208,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7341031010000734
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.007935200934298337,
          "train_loss": 0.004900040104985237,
          "validation_loss": 0.004253825172781944,
          "train_metric": 0.0004935141187161207,
          "validation_metric": 0.000428429659223184,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7487130669996986
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.014929735407349654,
          "train_loss": 0.012183059006929398,
          "validation_loss": 0.009452519938349724,
          "train_metric": 0.0012270330917090178,
          "validation_metric": 0.0009520230814814568,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7615598169995792
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.021169812825974077,
          "train_loss": 0.030843041837215424,
          "validation_loss": 0.020674344152212143,
          "train_metric": 0.0031063982751220465,
          "validation_metric": 0.002082244260236621,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7737730509998073
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.025349690346047282,
          "train_loss": 0.014885598793625832,
          "validation_loss": 0.014750920236110687,
          "train_metric": 0.0014992229407653213,
          "validation_metric": 0.0014856585767120123,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7853346800002328
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.011478115542558953,
          "train_loss": 0.005493846721947193,
          "validation_loss": 0.004880827385932207,
          "train_metric": 0.000553320103790611,
          "validation_metric": 0.0004915790050290525,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7966473640008189
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004808930782019161,
          "train_loss": 0.004848894663155079,
          "validation_loss": 0.003477165475487709,
          "train_metric": 0.0004883629153482616,
          "validation_metric": 0.0003502073814161122,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8085342060003313
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0051082603094982915,
          "train_loss": 0.008438479155302048,
          "validation_loss": 0.006683470215648413,
          "train_metric": 0.0008498926181346178,
          "validation_metric": 0.0006731346365995705,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8199786080003832
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.029912532554590143,
          "train_loss": 0.03629003092646599,
          "validation_loss": 0.029090482741594315,
          "train_metric": 0.0036549989599734545,
          "validation_metric": 0.0029298868030309677,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8318443749994913
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.029588691832032055,
          "train_loss": 0.012756697833538055,
          "validation_loss": 0.012664585374295712,
          "train_metric": 0.0012848079204559326,
          "validation_metric": 0.0012755306670442224,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8433738170006109
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.012370453827315941,
          "train_loss": 0.016269046813249588,
          "validation_loss": 0.014624612405896187,
          "train_metric": 0.001638558809645474,
          "validation_metric": 0.0014729374088346958,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8546411999996053
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.010272265542880632,
          "train_loss": 0.0073212371207773685,
          "validation_loss": 0.007599780336022377,
          "train_metric": 0.0007373682456091046,
          "validation_metric": 0.0007654220680706203,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8665501909999875
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007476677725207992,
          "train_loss": 0.01119996514171362,
          "validation_loss": 0.006029813084751368,
          "train_metric": 0.0011280194157734513,
          "validation_metric": 0.0006073006661608815,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8783582280011615
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0065902114729397,
          "train_loss": 0.005133438389748335,
          "validation_loss": 0.004623817279934883,
          "train_metric": 0.0005170211079530418,
          "validation_metric": 0.00046569397090934217,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8901004020008259
        }
      ],
      "best": {
        "epoch": 48,
        "updates": 768
      },
      "best_validation_metric": 0.0002886475413106382,
      "config": {
        "label": "LR 0.03",
        "batch": 64,
        "lr": 0.03,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.016989630405357208
    },
    "8ac87378bb5b_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 64,
          "samples_seen": 1024,
          "online_loss": 0.30405234018689953,
          "train_loss": 0.09432892501354218,
          "validation_loss": 0.07780647277832031,
          "train_metric": 0.009500463493168354,
          "validation_metric": 0.00783638097345829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03287953899962304
        },
        {
          "epoch": 2,
          "updates": 128,
          "samples_seen": 2048,
          "online_loss": 0.07553436110902112,
          "train_loss": 0.06811955571174622,
          "validation_loss": 0.056377023458480835,
          "train_metric": 0.006860751658678055,
          "validation_metric": 0.005678087472915649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06358485300006578
        },
        {
          "epoch": 3,
          "updates": 192,
          "samples_seen": 3072,
          "online_loss": 0.058350919804070145,
          "train_loss": 0.05071042850613594,
          "validation_loss": 0.04290137439966202,
          "train_metric": 0.005107368342578411,
          "validation_metric": 0.004320869222283363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09862146799969196
        },
        {
          "epoch": 4,
          "updates": 256,
          "samples_seen": 4096,
          "online_loss": 0.05183539605786791,
          "train_loss": 0.04143005609512329,
          "validation_loss": 0.03292068466544151,
          "train_metric": 0.004172683227807283,
          "validation_metric": 0.003315650625154376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12827697700049612
        },
        {
          "epoch": 5,
          "updates": 320,
          "samples_seen": 5120,
          "online_loss": 0.041577326410333626,
          "train_loss": 0.0330607108771801,
          "validation_loss": 0.025740105658769608,
          "train_metric": 0.0033297536429017782,
          "validation_metric": 0.0025924488436430693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15730650300065463
        },
        {
          "epoch": 6,
          "updates": 384,
          "samples_seen": 6144,
          "online_loss": 0.03692157486148062,
          "train_loss": 0.028630802407860756,
          "validation_loss": 0.02386183850467205,
          "train_metric": 0.0028835893608629704,
          "validation_metric": 0.0024032772053033113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18620443700092437
        },
        {
          "epoch": 7,
          "updates": 448,
          "samples_seen": 7168,
          "online_loss": 0.029337134736124426,
          "train_loss": 0.021653776988387108,
          "validation_loss": 0.018267223611474037,
          "train_metric": 0.0021808892488479614,
          "validation_metric": 0.0018398077227175236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2149342409993551
        },
        {
          "epoch": 8,
          "updates": 512,
          "samples_seen": 8192,
          "online_loss": 0.028584459039848298,
          "train_loss": 0.0523504801094532,
          "validation_loss": 0.036134831607341766,
          "train_metric": 0.005272548645734787,
          "validation_metric": 0.0036393676418811083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24402199200085306
        },
        {
          "epoch": 9,
          "updates": 576,
          "samples_seen": 9216,
          "online_loss": 0.034820369830413256,
          "train_loss": 0.027965635061264038,
          "validation_loss": 0.02425423450767994,
          "train_metric": 0.0028165963012725115,
          "validation_metric": 0.002442797878757119,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37747159699938493
        },
        {
          "epoch": 10,
          "updates": 640,
          "samples_seen": 10240,
          "online_loss": 0.022823747025540797,
          "train_loss": 0.02376246452331543,
          "validation_loss": 0.021696094423532486,
          "train_metric": 0.002393268281593919,
          "validation_metric": 0.0021851512137800455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42475303299943334
        },
        {
          "epoch": 11,
          "updates": 704,
          "samples_seen": 11264,
          "online_loss": 0.021480128412804333,
          "train_loss": 0.01560952141880989,
          "validation_loss": 0.012483786791563034,
          "train_metric": 0.0015721336239948869,
          "validation_metric": 0.0012573213316500187,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47032447200035676
        },
        {
          "epoch": 12,
          "updates": 768,
          "samples_seen": 12288,
          "online_loss": 0.016016627560020424,
          "train_loss": 0.015190854668617249,
          "validation_loss": 0.013768654316663742,
          "train_metric": 0.0015299671795219183,
          "validation_metric": 0.0013867283705621958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5136074729998654
        },
        {
          "epoch": 13,
          "updates": 832,
          "samples_seen": 13312,
          "online_loss": 0.013792937425932905,
          "train_loss": 0.013874605298042297,
          "validation_loss": 0.01084647886455059,
          "train_metric": 0.00139739946462214,
          "validation_metric": 0.0010924176312983036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5591684760001954
        },
        {
          "epoch": 14,
          "updates": 896,
          "samples_seen": 14336,
          "online_loss": 0.015341836156949284,
          "train_loss": 0.019977113232016563,
          "validation_loss": 0.018726734444499016,
          "train_metric": 0.002012021606788039,
          "validation_metric": 0.0018860880518332124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6068408240007557
        },
        {
          "epoch": 15,
          "updates": 960,
          "samples_seen": 15360,
          "online_loss": 0.01804348247424059,
          "train_loss": 0.026645472273230553,
          "validation_loss": 0.021020296961069107,
          "train_metric": 0.002683634404093027,
          "validation_metric": 0.0021170873660594225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6532289719998516
        },
        {
          "epoch": 16,
          "updates": 1024,
          "samples_seen": 16384,
          "online_loss": 0.015056374375490122,
          "train_loss": 0.011738265864551067,
          "validation_loss": 0.010271968320012093,
          "train_metric": 0.0011822350788861513,
          "validation_metric": 0.0010345550253987312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.699105714000325
        },
        {
          "epoch": 17,
          "updates": 1088,
          "samples_seen": 17408,
          "online_loss": 0.013324697471034597,
          "train_loss": 0.01271234080195427,
          "validation_loss": 0.011665533296763897,
          "train_metric": 0.0012803403660655022,
          "validation_metric": 0.0011749097611755133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7422313150000264
        },
        {
          "epoch": 18,
          "updates": 1152,
          "samples_seen": 18432,
          "online_loss": 0.01210375746632053,
          "train_loss": 0.014222957193851471,
          "validation_loss": 0.00854982528835535,
          "train_metric": 0.0014324841322377324,
          "validation_metric": 0.0008611070807091892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7851540560004651
        },
        {
          "epoch": 19,
          "updates": 1216,
          "samples_seen": 19456,
          "online_loss": 0.016799078177427873,
          "train_loss": 0.012104406021535397,
          "validation_loss": 0.01112705934792757,
          "train_metric": 0.0012191114947199821,
          "validation_metric": 0.0011206766357645392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8286657450007624
        },
        {
          "epoch": 20,
          "updates": 1280,
          "samples_seen": 20480,
          "online_loss": 0.012688635461017839,
          "train_loss": 0.010795699432492256,
          "validation_loss": 0.00687992200255394,
          "train_metric": 0.0010873033897951245,
          "validation_metric": 0.0006929205846972764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8723887390005984
        },
        {
          "epoch": 21,
          "updates": 1344,
          "samples_seen": 21504,
          "online_loss": 0.010700059143346152,
          "train_loss": 0.00927622802555561,
          "validation_loss": 0.006946666166186333,
          "train_metric": 0.0009342675912193954,
          "validation_metric": 0.000699642754625529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9159535529997811
        },
        {
          "epoch": 22,
          "updates": 1408,
          "samples_seen": 22528,
          "online_loss": 0.010594090073936968,
          "train_loss": 0.007905232720077038,
          "validation_loss": 0.005931931547820568,
          "train_metric": 0.0007961862138472497,
          "validation_metric": 0.0005974425002932549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.981398777999857
        },
        {
          "epoch": 23,
          "updates": 1472,
          "samples_seen": 23552,
          "online_loss": 0.01006094761987697,
          "train_loss": 0.00798938050866127,
          "validation_loss": 0.005514916963875294,
          "train_metric": 0.0008046611328609288,
          "validation_metric": 0.0005554422968998551,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.029762469999696
        },
        {
          "epoch": 24,
          "updates": 1536,
          "samples_seen": 24576,
          "online_loss": 0.011372572630534705,
          "train_loss": 0.007471478544175625,
          "validation_loss": 0.006100369617342949,
          "train_metric": 0.0007525000255554914,
          "validation_metric": 0.0006144069484435022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0594164180001826
        },
        {
          "epoch": 25,
          "updates": 1600,
          "samples_seen": 25600,
          "online_loss": 0.014801576882746303,
          "train_loss": 0.00864070188254118,
          "validation_loss": 0.007496920879930258,
          "train_metric": 0.0008702597697265446,
          "validation_metric": 0.0007550624432042241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0894920709997677
        },
        {
          "epoch": 26,
          "updates": 1664,
          "samples_seen": 26624,
          "online_loss": 0.009414805506821722,
          "train_loss": 0.007016154937446117,
          "validation_loss": 0.005437467247247696,
          "train_metric": 0.0007066414109431207,
          "validation_metric": 0.0005476417718455195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1201851170008013
        },
        {
          "epoch": 27,
          "updates": 1728,
          "samples_seen": 27648,
          "online_loss": 0.00947493024796131,
          "train_loss": 0.006848525255918503,
          "validation_loss": 0.005577326286584139,
          "train_metric": 0.0006897583953104913,
          "validation_metric": 0.0005617278511635959,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1499166670000704
        },
        {
          "epoch": 28,
          "updates": 1792,
          "samples_seen": 28672,
          "online_loss": 0.007366378178630839,
          "train_loss": 0.014364744536578655,
          "validation_loss": 0.01322558056563139,
          "train_metric": 0.001446764450520277,
          "validation_metric": 0.0013320320285856724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1797830640007305
        },
        {
          "epoch": 29,
          "updates": 1856,
          "samples_seen": 29696,
          "online_loss": 0.009038718269948731,
          "train_loss": 0.005967139732092619,
          "validation_loss": 0.005006393417716026,
          "train_metric": 0.0006009884527884424,
          "validation_metric": 0.0005042256671003997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2101842329993815
        },
        {
          "epoch": 30,
          "updates": 1920,
          "samples_seen": 30720,
          "online_loss": 0.007269305172940221,
          "train_loss": 0.006070842035114765,
          "validation_loss": 0.00526451226323843,
          "train_metric": 0.0006114329444244504,
          "validation_metric": 0.0005302223726175725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2401836680000997
        },
        {
          "epoch": 31,
          "updates": 1984,
          "samples_seen": 31744,
          "online_loss": 0.01183827543809457,
          "train_loss": 0.01838619075715542,
          "validation_loss": 0.013810832984745502,
          "train_metric": 0.0018517898861318827,
          "validation_metric": 0.0013909767149016261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2700377340006526
        },
        {
          "epoch": 32,
          "updates": 2048,
          "samples_seen": 32768,
          "online_loss": 0.009677641161033534,
          "train_loss": 0.00983097217977047,
          "validation_loss": 0.007972660474479198,
          "train_metric": 0.0009901394369080663,
          "validation_metric": 0.0008029771852307022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2998010600003909
        },
        {
          "epoch": 33,
          "updates": 2112,
          "samples_seen": 33792,
          "online_loss": 0.008866046568073216,
          "train_loss": 0.010993272997438908,
          "validation_loss": 0.009732788428664207,
          "train_metric": 0.0011072022607550025,
          "validation_metric": 0.000980250770226121,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3292513199994573
        },
        {
          "epoch": 34,
          "updates": 2176,
          "samples_seen": 34816,
          "online_loss": 0.007996867822839704,
          "train_loss": 0.005679445806890726,
          "validation_loss": 0.005176261067390442,
          "train_metric": 0.0005720129702240229,
          "validation_metric": 0.0005213340627960861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3584517990002496
        },
        {
          "epoch": 35,
          "updates": 2240,
          "samples_seen": 35840,
          "online_loss": 0.007470270575140603,
          "train_loss": 0.009562221355736256,
          "validation_loss": 0.007828319445252419,
          "train_metric": 0.0009630719432607293,
          "validation_metric": 0.0007884397055022418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.387737062999804
        },
        {
          "epoch": 36,
          "updates": 2304,
          "samples_seen": 36864,
          "online_loss": 0.011500852127028338,
          "train_loss": 0.013290808536112309,
          "validation_loss": 0.008989092893898487,
          "train_metric": 0.0013386014616116881,
          "validation_metric": 0.000905348570086062,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.417224367000017
        },
        {
          "epoch": 37,
          "updates": 2368,
          "samples_seen": 37888,
          "online_loss": 0.006825256305091898,
          "train_loss": 0.007449793629348278,
          "validation_loss": 0.005964143201708794,
          "train_metric": 0.0007503159577026963,
          "validation_metric": 0.000600686646066606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4570613859996229
        },
        {
          "epoch": 38,
          "updates": 2432,
          "samples_seen": 38912,
          "online_loss": 0.009277378966544347,
          "train_loss": 0.012609496712684631,
          "validation_loss": 0.009771067649126053,
          "train_metric": 0.0012699823128059506,
          "validation_metric": 0.0009841059800237417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4966022139997222
        },
        {
          "epoch": 39,
          "updates": 2496,
          "samples_seen": 39936,
          "online_loss": 0.014622356507970835,
          "train_loss": 0.005917272996157408,
          "validation_loss": 0.004529896192252636,
          "train_metric": 0.0005959660629741848,
          "validation_metric": 0.0004562345566228032,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5263820769996528
        },
        {
          "epoch": 40,
          "updates": 2560,
          "samples_seen": 40960,
          "online_loss": 0.007362722289144585,
          "train_loss": 0.021916955709457397,
          "validation_loss": 0.020337406545877457,
          "train_metric": 0.002207395387813449,
          "validation_metric": 0.0020483091939240694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5562306600004376
        },
        {
          "epoch": 41,
          "updates": 2624,
          "samples_seen": 41984,
          "online_loss": 0.008980282688753505,
          "train_loss": 0.013478385284543037,
          "validation_loss": 0.012544209137558937,
          "train_metric": 0.0013574936892837286,
          "validation_metric": 0.0012634068261831999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5860983290003787
        },
        {
          "epoch": 42,
          "updates": 2688,
          "samples_seen": 43008,
          "online_loss": 0.015158131408497866,
          "train_loss": 0.00685978913679719,
          "validation_loss": 0.00523852976039052,
          "train_metric": 0.0006908929208293557,
          "validation_metric": 0.0005276055308058858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.616077568000037
        },
        {
          "epoch": 43,
          "updates": 2752,
          "samples_seen": 44032,
          "online_loss": 0.007505802137075079,
          "train_loss": 0.004884110763669014,
          "validation_loss": 0.0033445097506046295,
          "train_metric": 0.0004919096827507019,
          "validation_metric": 0.00033684674417600036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6471752840006957
        },
        {
          "epoch": 44,
          "updates": 2816,
          "samples_seen": 45056,
          "online_loss": 0.005758723398685106,
          "train_loss": 0.005221392959356308,
          "validation_loss": 0.0042118774726986885,
          "train_metric": 0.0005258794990368187,
          "validation_metric": 0.0004242048889864236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6982048930003657
        },
        {
          "epoch": 45,
          "updates": 2880,
          "samples_seen": 46080,
          "online_loss": 0.006443665019105538,
          "train_loss": 0.005131413694471121,
          "validation_loss": 0.0036247121170163155,
          "train_metric": 0.0005168172065168619,
          "validation_metric": 0.00036506770993582904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.731387136000194
        },
        {
          "epoch": 46,
          "updates": 2944,
          "samples_seen": 47104,
          "online_loss": 0.008460028840090672,
          "train_loss": 0.004031615797430277,
          "validation_loss": 0.003203451167792082,
          "train_metric": 0.00040604957030154765,
          "validation_metric": 0.00032263988396152854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7623255310008972
        },
        {
          "epoch": 47,
          "updates": 3008,
          "samples_seen": 48128,
          "online_loss": 0.004597357190959883,
          "train_loss": 0.004257503431290388,
          "validation_loss": 0.002866714494302869,
          "train_metric": 0.0004288001509848982,
          "validation_metric": 0.0002887249574996531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.792745520000608
        },
        {
          "epoch": 48,
          "updates": 3072,
          "samples_seen": 49152,
          "online_loss": 0.005335754691259353,
          "train_loss": 0.004363509826362133,
          "validation_loss": 0.003632086329162121,
          "train_metric": 0.0004394768038764596,
          "validation_metric": 0.0003658104396890849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8226801159999013
        },
        {
          "epoch": 49,
          "updates": 3136,
          "samples_seen": 50176,
          "online_loss": 0.00720228835598391,
          "train_loss": 0.017325375229120255,
          "validation_loss": 0.012586082331836224,
          "train_metric": 0.0017449482111260295,
          "validation_metric": 0.0012676242040470243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8523054670004058
        },
        {
          "epoch": 50,
          "updates": 3200,
          "samples_seen": 51200,
          "online_loss": 0.009371280392770132,
          "train_loss": 0.006701665930449963,
          "validation_loss": 0.004037297796458006,
          "train_metric": 0.000674967363011092,
          "validation_metric": 0.0004066218389198184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8823505410000507
        },
        {
          "epoch": 51,
          "updates": 3264,
          "samples_seen": 52224,
          "online_loss": 0.006193068369611865,
          "train_loss": 0.0040632253512740135,
          "validation_loss": 0.0030864723958075047,
          "train_metric": 0.00040923318010754883,
          "validation_metric": 0.00031085815862752497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9123537970008329
        },
        {
          "epoch": 52,
          "updates": 3328,
          "samples_seen": 53248,
          "online_loss": 0.0054987036969578185,
          "train_loss": 0.004838605411350727,
          "validation_loss": 0.003937999717891216,
          "train_metric": 0.0004873266152571887,
          "validation_metric": 0.000396620889659971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9425773139992089
        },
        {
          "epoch": 53,
          "updates": 3392,
          "samples_seen": 54272,
          "online_loss": 0.005621636893010873,
          "train_loss": 0.005704227834939957,
          "validation_loss": 0.005611371714621782,
          "train_metric": 0.0005745089147239923,
          "validation_metric": 0.0005651568062603474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9724944299996423
        },
        {
          "epoch": 54,
          "updates": 3456,
          "samples_seen": 55296,
          "online_loss": 0.005382112888582924,
          "train_loss": 0.0059754676185548306,
          "validation_loss": 0.004852229729294777,
          "train_metric": 0.0006018272833898664,
          "validation_metric": 0.0004886988317593932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0019149649997416
        },
        {
          "epoch": 55,
          "updates": 3520,
          "samples_seen": 56320,
          "online_loss": 0.00605211429137853,
          "train_loss": 0.005337027832865715,
          "validation_loss": 0.003999953158199787,
          "train_metric": 0.000537525920663029,
          "validation_metric": 0.0004028606344945729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.032097564000651
        },
        {
          "epoch": 56,
          "updates": 3584,
          "samples_seen": 57344,
          "online_loss": 0.005341795178537723,
          "train_loss": 0.004583438392728567,
          "validation_loss": 0.0030073358211666346,
          "train_metric": 0.00046162717626430094,
          "validation_metric": 0.00030288787093013525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0622510079992935
        },
        {
          "epoch": 57,
          "updates": 3648,
          "samples_seen": 58368,
          "online_loss": 0.0067485419640433975,
          "train_loss": 0.00502941245213151,
          "validation_loss": 0.004121781326830387,
          "train_metric": 0.0005065440200269222,
          "validation_metric": 0.00041513072210364044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.092349385000489
        },
        {
          "epoch": 58,
          "updates": 3712,
          "samples_seen": 59392,
          "online_loss": 0.008932914560773497,
          "train_loss": 0.004488174803555012,
          "validation_loss": 0.004202807322144508,
          "train_metric": 0.0004520324873737991,
          "validation_metric": 0.00042329137795604765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.122161748000508
        },
        {
          "epoch": 59,
          "updates": 3776,
          "samples_seen": 60416,
          "online_loss": 0.00446444940598667,
          "train_loss": 0.0035428660921752453,
          "validation_loss": 0.0027048583142459393,
          "train_metric": 0.00035682451562024653,
          "validation_metric": 0.0002724234655033797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1530412350002734
        },
        {
          "epoch": 60,
          "updates": 3840,
          "samples_seen": 61440,
          "online_loss": 0.005137224774443894,
          "train_loss": 0.006164331454783678,
          "validation_loss": 0.00480642868205905,
          "train_metric": 0.0006208489066921175,
          "validation_metric": 0.0004840858746320009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.183503653999651
        },
        {
          "epoch": 61,
          "updates": 3904,
          "samples_seen": 62464,
          "online_loss": 0.004551064762381429,
          "train_loss": 0.003381602931767702,
          "validation_loss": 0.003083042800426483,
          "train_metric": 0.00034058268647640944,
          "validation_metric": 0.0003105127834714949,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.214177641999413
        },
        {
          "epoch": 62,
          "updates": 3968,
          "samples_seen": 63488,
          "online_loss": 0.008478669217311108,
          "train_loss": 0.008639921434223652,
          "validation_loss": 0.008510851301252842,
          "train_metric": 0.0008701811893843114,
          "validation_metric": 0.0008571818470954895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2442351869995036
        },
        {
          "epoch": 63,
          "updates": 4032,
          "samples_seen": 64512,
          "online_loss": 0.006250769816233515,
          "train_loss": 0.004123435355722904,
          "validation_loss": 0.0031041509937494993,
          "train_metric": 0.0004152973706368357,
          "validation_metric": 0.00031263873097486794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2747030240007007
        },
        {
          "epoch": 64,
          "updates": 4096,
          "samples_seen": 65536,
          "online_loss": 0.006697426538039508,
          "train_loss": 0.0043337345123291016,
          "validation_loss": 0.0036273489240556955,
          "train_metric": 0.00043647788697853684,
          "validation_metric": 0.0003653333114925772,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.304349887999706
        },
        {
          "epoch": 65,
          "updates": 4160,
          "samples_seen": 66560,
          "online_loss": 0.0053494626167776005,
          "train_loss": 0.006131673697382212,
          "validation_loss": 0.004584598354995251,
          "train_metric": 0.0006175597663968801,
          "validation_metric": 0.0004617439699359238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3348272149996774
        },
        {
          "epoch": 66,
          "updates": 4224,
          "samples_seen": 67584,
          "online_loss": 0.006699013464185555,
          "train_loss": 0.00401175394654274,
          "validation_loss": 0.0033636940643191338,
          "train_metric": 0.00040404911851510406,
          "validation_metric": 0.00033877897658385336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.376576960999955
        },
        {
          "epoch": 67,
          "updates": 4288,
          "samples_seen": 68608,
          "online_loss": 0.004993303856281273,
          "train_loss": 0.003618378657847643,
          "validation_loss": 0.0027476095128804445,
          "train_metric": 0.00036442981217987835,
          "validation_metric": 0.0002767291443888098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.409548605000964
        },
        {
          "epoch": 68,
          "updates": 4352,
          "samples_seen": 69632,
          "online_loss": 0.005318216229170503,
          "train_loss": 0.0036440554540604353,
          "validation_loss": 0.0031076068989932537,
          "train_metric": 0.0003670159203466028,
          "validation_metric": 0.0003129867836833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4413416250008595
        },
        {
          "epoch": 69,
          "updates": 4416,
          "samples_seen": 70656,
          "online_loss": 0.004738453658319486,
          "train_loss": 0.0029254285618662834,
          "validation_loss": 0.0022887960076332092,
          "train_metric": 0.00029463841929100454,
          "validation_metric": 0.00023051915923133492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4743745849991683
        },
        {
          "epoch": 70,
          "updates": 4480,
          "samples_seen": 71680,
          "online_loss": 0.005582029704783054,
          "train_loss": 0.003437668550759554,
          "validation_loss": 0.0027475026436150074,
          "train_metric": 0.00034622938255779445,
          "validation_metric": 0.00027671840507537127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.508026900000914
        },
        {
          "epoch": 71,
          "updates": 4544,
          "samples_seen": 72704,
          "online_loss": 0.004770427421590284,
          "train_loss": 0.003331385087221861,
          "validation_loss": 0.0022828378714621067,
          "train_metric": 0.0003355249355081469,
          "validation_metric": 0.00022991906735114753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5415378300003795
        },
        {
          "epoch": 72,
          "updates": 4608,
          "samples_seen": 73728,
          "online_loss": 0.004814205785351078,
          "train_loss": 0.003334868000820279,
          "validation_loss": 0.0028704041615128517,
          "train_metric": 0.00033587569487281144,
          "validation_metric": 0.00028909658431075513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.586473534000106
        },
        {
          "epoch": 73,
          "updates": 4672,
          "samples_seen": 74752,
          "online_loss": 0.004629666373148211,
          "train_loss": 0.008785701356828213,
          "validation_loss": 0.008273785933852196,
          "train_metric": 0.0008848637226037681,
          "validation_metric": 0.0008333054138347507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.655793600000834
        },
        {
          "epoch": 74,
          "updates": 4736,
          "samples_seen": 75776,
          "online_loss": 0.007456771199031209,
          "train_loss": 0.0055289119482040405,
          "validation_loss": 0.00377639546059072,
          "train_metric": 0.0005568517372012138,
          "validation_metric": 0.00038034471799619496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.688204403000782
        },
        {
          "epoch": 75,
          "updates": 4800,
          "samples_seen": 76800,
          "online_loss": 0.005208477526139177,
          "train_loss": 0.00735034653916955,
          "validation_loss": 0.003949092235416174,
          "train_metric": 0.0007402999326586723,
          "validation_metric": 0.0003977381275035441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7198360479997064
        },
        {
          "epoch": 76,
          "updates": 4864,
          "samples_seen": 77824,
          "online_loss": 0.006433867940813798,
          "train_loss": 0.004554219543933868,
          "validation_loss": 0.004850180819630623,
          "train_metric": 0.00045868437155149877,
          "validation_metric": 0.0004884924273937941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7522697929998685
        },
        {
          "epoch": 77,
          "updates": 4928,
          "samples_seen": 78848,
          "online_loss": 0.008853975837155303,
          "train_loss": 0.004178787115961313,
          "validation_loss": 0.00434187101200223,
          "train_metric": 0.0004208721511531621,
          "validation_metric": 0.0004372973635327071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.788058775000536
        },
        {
          "epoch": 78,
          "updates": 4992,
          "samples_seen": 79872,
          "online_loss": 0.006105519566517614,
          "train_loss": 0.013112669810652733,
          "validation_loss": 0.010665046982467175,
          "train_metric": 0.00132066011428833,
          "validation_metric": 0.0010741446167230606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8290999620003277
        },
        {
          "epoch": 79,
          "updates": 5056,
          "samples_seen": 80896,
          "online_loss": 0.009076230382561334,
          "train_loss": 0.0043055531568825245,
          "validation_loss": 0.003981491085141897,
          "train_metric": 0.00043363950680941343,
          "validation_metric": 0.0004010012198705226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8642294040000706
        },
        {
          "epoch": 80,
          "updates": 5120,
          "samples_seen": 81920,
          "online_loss": 0.00494230776303084,
          "train_loss": 0.010593020357191563,
          "validation_loss": 0.009459535591304302,
          "train_metric": 0.0010668901959434152,
          "validation_metric": 0.000952729897107929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.911109654000029
        },
        {
          "epoch": 81,
          "updates": 5184,
          "samples_seen": 82944,
          "online_loss": 0.004856095925561021,
          "train_loss": 0.003249396802857518,
          "validation_loss": 0.0029310123063623905,
          "train_metric": 0.00032726730569265783,
          "validation_metric": 0.0002952008508145809,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9641375449991756
        },
        {
          "epoch": 82,
          "updates": 5248,
          "samples_seen": 83968,
          "online_loss": 0.007762419725168002,
          "train_loss": 0.010052097961306572,
          "validation_loss": 0.007531917653977871,
          "train_metric": 0.0010124106192961335,
          "validation_metric": 0.0007585871499031782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0180731140007993
        },
        {
          "epoch": 83,
          "updates": 5312,
          "samples_seen": 84992,
          "online_loss": 0.004788763601027313,
          "train_loss": 0.011373627930879593,
          "validation_loss": 0.009264699183404446,
          "train_metric": 0.0011455100029706955,
          "validation_metric": 0.0009331065230071545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0679314370008797
        },
        {
          "epoch": 84,
          "updates": 5376,
          "samples_seen": 86016,
          "online_loss": 0.006873411657579709,
          "train_loss": 0.004976554308086634,
          "validation_loss": 0.003740814048796892,
          "train_metric": 0.0005012203473597765,
          "validation_metric": 0.0003767611051443964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1343090160007705
        },
        {
          "epoch": 85,
          "updates": 5440,
          "samples_seen": 87040,
          "online_loss": 0.006468521253282233,
          "train_loss": 0.008581401780247688,
          "validation_loss": 0.006798807997256517,
          "train_metric": 0.0008642873726785183,
          "validation_metric": 0.0006847510812804103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1780756959997234
        },
        {
          "epoch": 86,
          "updates": 5504,
          "samples_seen": 88064,
          "online_loss": 0.004428734755038022,
          "train_loss": 0.0067888484336435795,
          "validation_loss": 0.006650186609476805,
          "train_metric": 0.0006837479304522276,
          "validation_metric": 0.0006697825156152248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.293941571000687
        },
        {
          "epoch": 87,
          "updates": 5568,
          "samples_seen": 89088,
          "online_loss": 0.004600659552124853,
          "train_loss": 0.0037590707652270794,
          "validation_loss": 0.004042862914502621,
          "train_metric": 0.0003785997978411615,
          "validation_metric": 0.00040718226227909327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.360570316999656
        },
        {
          "epoch": 88,
          "updates": 5632,
          "samples_seen": 90112,
          "online_loss": 0.005690874304718818,
          "train_loss": 0.006972610019147396,
          "validation_loss": 0.00513526750728488,
          "train_metric": 0.0007022557547315955,
          "validation_metric": 0.0005172053934074938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.405818695999187
        },
        {
          "epoch": 89,
          "updates": 5696,
          "samples_seen": 91136,
          "online_loss": 0.004117303341899969,
          "train_loss": 0.0030482769943773746,
          "validation_loss": 0.0021788303274661303,
          "train_metric": 0.0003070112725254148,
          "validation_metric": 0.00021944381296634674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.452249206000488
        },
        {
          "epoch": 90,
          "updates": 5760,
          "samples_seen": 92160,
          "online_loss": 0.007077329594267212,
          "train_loss": 0.008106725290417671,
          "validation_loss": 0.005950107239186764,
          "train_metric": 0.0008164796163327992,
          "validation_metric": 0.0005992730148136616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.497163885000191
        },
        {
          "epoch": 91,
          "updates": 5824,
          "samples_seen": 93184,
          "online_loss": 0.004493783553243702,
          "train_loss": 0.0041456460021436214,
          "validation_loss": 0.0029534432105720043,
          "train_metric": 0.0004175342619419098,
          "validation_metric": 0.00029745997744612396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.545674405999307
        },
        {
          "epoch": 92,
          "updates": 5888,
          "samples_seen": 94208,
          "online_loss": 0.005081416660459581,
          "train_loss": 0.008835327811539173,
          "validation_loss": 0.006014087703078985,
          "train_metric": 0.000889861723408103,
          "validation_metric": 0.0006057168357074261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6328451720000885
        },
        {
          "epoch": 93,
          "updates": 5952,
          "samples_seen": 95232,
          "online_loss": 0.010254840841753321,
          "train_loss": 0.017273858189582825,
          "validation_loss": 0.01278870552778244,
          "train_metric": 0.0017397596966475248,
          "validation_metric": 0.001288031693547964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.71691225900031
        },
        {
          "epoch": 94,
          "updates": 6016,
          "samples_seen": 96256,
          "online_loss": 0.0053323273386922665,
          "train_loss": 0.006619919557124376,
          "validation_loss": 0.004498146940022707,
          "train_metric": 0.0006667341222055256,
          "validation_metric": 0.0004530368896666914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7691804419991968
        },
        {
          "epoch": 95,
          "updates": 6080,
          "samples_seen": 97280,
          "online_loss": 0.0029970165874146915,
          "train_loss": 0.003940886352211237,
          "validation_loss": 0.0024327002465724945,
          "train_metric": 0.0003969116078224033,
          "validation_metric": 0.00024501263396814466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8156009150006867
        },
        {
          "epoch": 96,
          "updates": 6144,
          "samples_seen": 98304,
          "online_loss": 0.005573884599471057,
          "train_loss": 0.002654040465131402,
          "validation_loss": 0.0020992420613765717,
          "train_metric": 0.0002673052658792585,
          "validation_metric": 0.00021142794867046177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.863383968000562
        },
        {
          "epoch": 97,
          "updates": 6208,
          "samples_seen": 99328,
          "online_loss": 0.0032278571275128343,
          "train_loss": 0.002344634383916855,
          "validation_loss": 0.001625742414034903,
          "train_metric": 0.00023614296515006572,
          "validation_metric": 0.00016373881953768432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.911645041000156
        },
        {
          "epoch": 98,
          "updates": 6272,
          "samples_seen": 100352,
          "online_loss": 0.0032956773536625406,
          "train_loss": 0.0027507536578923464,
          "validation_loss": 0.001542727928608656,
          "train_metric": 0.0002770458231680095,
          "validation_metric": 0.00015537792933173478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.959944075999374
        },
        {
          "epoch": 99,
          "updates": 6336,
          "samples_seen": 101376,
          "online_loss": 0.003888637761519931,
          "train_loss": 0.005359954666346312,
          "validation_loss": 0.005575936753302813,
          "train_metric": 0.0005398350185714662,
          "validation_metric": 0.0005615879199467599,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.005970273999992
        },
        {
          "epoch": 100,
          "updates": 6400,
          "samples_seen": 102400,
          "online_loss": 0.0071557096280230326,
          "train_loss": 0.008553983643651009,
          "validation_loss": 0.005157074425369501,
          "train_metric": 0.0008615258848294616,
          "validation_metric": 0.0005194016848690808,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.051137820999429
        },
        {
          "epoch": 101,
          "updates": 6464,
          "samples_seen": 103424,
          "online_loss": 0.005227677026596211,
          "train_loss": 0.0037773295771330595,
          "validation_loss": 0.003218442667275667,
          "train_metric": 0.0003804388106800616,
          "validation_metric": 0.0003241497906856239,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.096296255000198
        },
        {
          "epoch": 102,
          "updates": 6528,
          "samples_seen": 104448,
          "online_loss": 0.010747707759946934,
          "train_loss": 0.0025641112588346004,
          "validation_loss": 0.0018489259527996182,
          "train_metric": 0.0002582478919066489,
          "validation_metric": 0.000186217061127536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.141606008999588
        },
        {
          "epoch": 103,
          "updates": 6592,
          "samples_seen": 105472,
          "online_loss": 0.003153159838575448,
          "train_loss": 0.0024327083956450224,
          "validation_loss": 0.0020581642165780067,
          "train_metric": 0.00024501350708305836,
          "validation_metric": 0.00020729076641146094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.188371170999744
        },
        {
          "epoch": 104,
          "updates": 6656,
          "samples_seen": 106496,
          "online_loss": 0.0047518743335785985,
          "train_loss": 0.0023689470253884792,
          "validation_loss": 0.0018883863231167197,
          "train_metric": 0.0002385916595812887,
          "validation_metric": 0.00019019133469555527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.235082819999661
        },
        {
          "epoch": 105,
          "updates": 6720,
          "samples_seen": 107520,
          "online_loss": 0.002864625022994005,
          "train_loss": 0.002149714156985283,
          "validation_loss": 0.0017709290841594338,
          "train_metric": 0.00021651132556144148,
          "validation_metric": 0.00017836150072980672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2820347159995436
        },
        {
          "epoch": 106,
          "updates": 6784,
          "samples_seen": 108544,
          "online_loss": 0.003044066114398447,
          "train_loss": 0.002749798819422722,
          "validation_loss": 0.0025093320291489363,
          "train_metric": 0.00027694969321601093,
          "validation_metric": 0.0002527307369746268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.327685006999673
        },
        {
          "epoch": 107,
          "updates": 6848,
          "samples_seen": 109568,
          "online_loss": 0.010788415421302489,
          "train_loss": 0.004282890819013119,
          "validation_loss": 0.0033033054787665606,
          "train_metric": 0.0004313570389058441,
          "validation_metric": 0.0003326968289911747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.3725771570007055
        },
        {
          "epoch": 108,
          "updates": 6912,
          "samples_seen": 110592,
          "online_loss": 0.006096399119087437,
          "train_loss": 0.0036718002520501614,
          "validation_loss": 0.002482930663973093,
          "train_metric": 0.00036981026642024517,
          "validation_metric": 0.00025007169460877776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.418461589999424
        },
        {
          "epoch": 109,
          "updates": 6976,
          "samples_seen": 111616,
          "online_loss": 0.005053864512774453,
          "train_loss": 0.003872285597026348,
          "validation_loss": 0.0018221137579530478,
          "train_metric": 0.0003900024457834661,
          "validation_metric": 0.00018351660401094705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.465802713000812
        },
        {
          "epoch": 110,
          "updates": 7040,
          "samples_seen": 112640,
          "online_loss": 0.0031148710509114608,
          "train_loss": 0.002265114802867174,
          "validation_loss": 0.0018608060199767351,
          "train_metric": 0.00022813407122157514,
          "validation_metric": 0.00018741357780527323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.511604831999648
        },
        {
          "epoch": 111,
          "updates": 7104,
          "samples_seen": 113664,
          "online_loss": 0.002669593961513783,
          "train_loss": 0.0022198681253939867,
          "validation_loss": 0.0019114331807941198,
          "train_metric": 0.00022357699344865978,
          "validation_metric": 0.00019251253979746252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.557833254000798
        },
        {
          "epoch": 112,
          "updates": 7168,
          "samples_seen": 114688,
          "online_loss": 0.003870617797701925,
          "train_loss": 0.00514136953279376,
          "validation_loss": 0.0034159133210778236,
          "train_metric": 0.0005178199498914182,
          "validation_metric": 0.0003440382715780288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.604496680000011
        },
        {
          "epoch": 113,
          "updates": 7232,
          "samples_seen": 115712,
          "online_loss": 0.00461149069815292,
          "train_loss": 0.0026698634028434753,
          "validation_loss": 0.0018044740427285433,
          "train_metric": 0.00026889884611591697,
          "validation_metric": 0.00018174001888837665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.650824193000517
        },
        {
          "epoch": 114,
          "updates": 7296,
          "samples_seen": 116736,
          "online_loss": 0.003885151026111089,
          "train_loss": 0.0018876653630286455,
          "validation_loss": 0.001373938866890967,
          "train_metric": 0.00019011874974239618,
          "validation_metric": 0.00013837809092365205,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.696511960000862
        },
        {
          "epoch": 115,
          "updates": 7360,
          "samples_seen": 117760,
          "online_loss": 0.0038666413464625293,
          "train_loss": 0.004962870851159096,
          "validation_loss": 0.0035862033255398273,
          "train_metric": 0.0004998421645723283,
          "validation_metric": 0.00036118930438533425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.74397278700053
        },
        {
          "epoch": 116,
          "updates": 7424,
          "samples_seen": 118784,
          "online_loss": 0.002704373834603757,
          "train_loss": 0.0037212141323834658,
          "validation_loss": 0.002075754338875413,
          "train_metric": 0.00037478707963600755,
          "validation_metric": 0.00020906234567519277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.790368592000959
        },
        {
          "epoch": 117,
          "updates": 7488,
          "samples_seen": 119808,
          "online_loss": 0.002402125404387334,
          "train_loss": 0.0024109091609716415,
          "validation_loss": 0.0021638888865709305,
          "train_metric": 0.0002428179286653176,
          "validation_metric": 0.00021793895575683564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.837922767999771
        },
        {
          "epoch": 118,
          "updates": 7552,
          "samples_seen": 120832,
          "online_loss": 0.003067298592895895,
          "train_loss": 0.002319727325811982,
          "validation_loss": 0.0015160581097006798,
          "train_metric": 0.00023363441869150847,
          "validation_metric": 0.00015269183495547622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.883703555000466
        },
        {
          "epoch": 119,
          "updates": 7616,
          "samples_seen": 121856,
          "online_loss": 0.004464163337615901,
          "train_loss": 0.002965671941637993,
          "validation_loss": 0.002165667014196515,
          "train_metric": 0.00029869165155105293,
          "validation_metric": 0.00021811800252180547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.931317357999433
        },
        {
          "epoch": 120,
          "updates": 7680,
          "samples_seen": 122880,
          "online_loss": 0.003751158600607596,
          "train_loss": 0.005794346798211336,
          "validation_loss": 0.005087094381451607,
          "train_metric": 0.000583585468120873,
          "validation_metric": 0.0005123535520397127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.978004713999326
        },
        {
          "epoch": 121,
          "updates": 7744,
          "samples_seen": 123904,
          "online_loss": 0.003790390614540229,
          "train_loss": 0.0034614610485732555,
          "validation_loss": 0.003749551484361291,
          "train_metric": 0.0003486256464384496,
          "validation_metric": 0.0003776410885620862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.024185538000893
        },
        {
          "epoch": 122,
          "updates": 7808,
          "samples_seen": 124928,
          "online_loss": 0.0035518588713330246,
          "train_loss": 0.0027786085847765207,
          "validation_loss": 0.002021545311436057,
          "train_metric": 0.0002798513160087168,
          "validation_metric": 0.0002036026562564075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.07059862099959
        },
        {
          "epoch": 123,
          "updates": 7872,
          "samples_seen": 125952,
          "online_loss": 0.00433643936912631,
          "train_loss": 0.005810137838125229,
          "validation_loss": 0.004873223602771759,
          "train_metric": 0.0005851758760400116,
          "validation_metric": 0.0004908132832497358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.115647963999436
        },
        {
          "epoch": 124,
          "updates": 7936,
          "samples_seen": 126976,
          "online_loss": 0.00952824796513596,
          "train_loss": 0.0051563093438744545,
          "validation_loss": 0.003785326611250639,
          "train_metric": 0.0005193246179260314,
          "validation_metric": 0.0003812442591879517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.162615716999426
        },
        {
          "epoch": 125,
          "updates": 8000,
          "samples_seen": 128000,
          "online_loss": 0.007006684886164294,
          "train_loss": 0.007816600613296032,
          "validation_loss": 0.00660160044208169,
          "train_metric": 0.0007872594287618995,
          "validation_metric": 0.0006648890557698905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.209227848999944
        },
        {
          "epoch": 126,
          "updates": 8064,
          "samples_seen": 129024,
          "online_loss": 0.007803040703038278,
          "train_loss": 0.004308783449232578,
          "validation_loss": 0.002919509308412671,
          "train_metric": 0.0004339648876339197,
          "validation_metric": 0.00029404228553175926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.256145359999209
        },
        {
          "epoch": 127,
          "updates": 8128,
          "samples_seen": 130048,
          "online_loss": 0.0034580576611915603,
          "train_loss": 0.007300990633666515,
          "validation_loss": 0.004674061667174101,
          "train_metric": 0.0007353290566243231,
          "validation_metric": 0.00047075439943000674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.300765441999829
        },
        {
          "epoch": 128,
          "updates": 8192,
          "samples_seen": 131072,
          "online_loss": 0.00612249744995097,
          "train_loss": 0.011094090528786182,
          "validation_loss": 0.008920840919017792,
          "train_metric": 0.0011173561215400696,
          "validation_metric": 0.0008984744781628251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.3464464110002154
        },
        {
          "epoch": 129,
          "updates": 8256,
          "samples_seen": 132096,
          "online_loss": 0.002979872282253382,
          "train_loss": 0.0033686941023916006,
          "validation_loss": 0.003133278340101242,
          "train_metric": 0.00033928256016224623,
          "validation_metric": 0.0003155723097734153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.392830791000961
        },
        {
          "epoch": 130,
          "updates": 8320,
          "samples_seen": 133120,
          "online_loss": 0.003830464023167224,
          "train_loss": 0.004075348377227783,
          "validation_loss": 0.002914399141445756,
          "train_metric": 0.0004104541440028697,
          "validation_metric": 0.0002935276133939624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.439658896000765
        },
        {
          "epoch": 131,
          "updates": 8384,
          "samples_seen": 134144,
          "online_loss": 0.0061447716802831565,
          "train_loss": 0.0027744146063923836,
          "validation_loss": 0.001773764961399138,
          "train_metric": 0.0002794288448058069,
          "validation_metric": 0.00017864711116999388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.48660057699999
        },
        {
          "epoch": 132,
          "updates": 8448,
          "samples_seen": 135168,
          "online_loss": 0.0035228589065354754,
          "train_loss": 0.0025315466336905956,
          "validation_loss": 0.0022327147889882326,
          "train_metric": 0.00025496812304481864,
          "validation_metric": 0.00022487084788735956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.532748794999861
        },
        {
          "epoch": 133,
          "updates": 8512,
          "samples_seen": 136192,
          "online_loss": 0.0045763755833831965,
          "train_loss": 0.0027543872129172087,
          "validation_loss": 0.0023243133910000324,
          "train_metric": 0.0002774117747321725,
          "validation_metric": 0.0002340963255846873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.579422189999605
        },
        {
          "epoch": 134,
          "updates": 8576,
          "samples_seen": 137216,
          "online_loss": 0.0038265666821644118,
          "train_loss": 0.0034312757197767496,
          "validation_loss": 0.001498499303124845,
          "train_metric": 0.0003455854894127697,
          "validation_metric": 0.00015092332614585757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.625406197999837
        }
      ],
      "best": {
        "epoch": 114,
        "updates": 7296
      },
      "best_validation_metric": 0.00013837809092365205,
      "config": {
        "label": "Batch 16",
        "batch": 16,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.011763421735347758
    },
    "8ac87378bb5b_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 64,
          "samples_seen": 1024,
          "online_loss": 0.30496128514641896,
          "train_loss": 0.09889104962348938,
          "validation_loss": 0.08465031534433365,
          "train_metric": 0.00995994359254837,
          "validation_metric": 0.008525669574737549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.052612999999837484
        },
        {
          "epoch": 2,
          "updates": 128,
          "samples_seen": 2048,
          "online_loss": 0.08240926322469022,
          "train_loss": 0.060156285762786865,
          "validation_loss": 0.04975259304046631,
          "train_metric": 0.006058720406144857,
          "validation_metric": 0.005010898690670729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0979012360003253
        },
        {
          "epoch": 3,
          "updates": 192,
          "samples_seen": 3072,
          "online_loss": 0.055069187670596875,
          "train_loss": 0.05514586344361305,
          "validation_loss": 0.04595014825463295,
          "train_metric": 0.005554089788347483,
          "validation_metric": 0.00462793093174696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14323819300079776
        },
        {
          "epoch": 4,
          "updates": 256,
          "samples_seen": 4096,
          "online_loss": 0.042041635198984295,
          "train_loss": 0.03377453610301018,
          "validation_loss": 0.02893630415201187,
          "train_metric": 0.0034016475547105074,
          "validation_metric": 0.002914358163252473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1869516320002731
        },
        {
          "epoch": 5,
          "updates": 320,
          "samples_seen": 5120,
          "online_loss": 0.042440216333488934,
          "train_loss": 0.034161485731601715,
          "validation_loss": 0.0324280671775341,
          "train_metric": 0.003440620144829154,
          "validation_metric": 0.0032660362776368856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23789498600126535
        },
        {
          "epoch": 6,
          "updates": 384,
          "samples_seen": 6144,
          "online_loss": 0.03160349497193238,
          "train_loss": 0.024577395990490913,
          "validation_loss": 0.020045440644025803,
          "train_metric": 0.00247534504160285,
          "validation_metric": 0.0020189033821225166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31932215000051656
        },
        {
          "epoch": 7,
          "updates": 448,
          "samples_seen": 7168,
          "online_loss": 0.024795499339234084,
          "train_loss": 0.0337798036634922,
          "validation_loss": 0.025481587275862694,
          "train_metric": 0.003402178408578038,
          "validation_metric": 0.002566412091255188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4218930000006367
        },
        {
          "epoch": 8,
          "updates": 512,
          "samples_seen": 8192,
          "online_loss": 0.02799120203962957,
          "train_loss": 0.021791907027363777,
          "validation_loss": 0.015822535380721092,
          "train_metric": 0.002194801578298211,
          "validation_metric": 0.0015935878036543727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47978080500070064
        },
        {
          "epoch": 9,
          "updates": 576,
          "samples_seen": 9216,
          "online_loss": 0.02110055333650962,
          "train_loss": 0.016142243519425392,
          "validation_loss": 0.012869763188064098,
          "train_metric": 0.00162578746676445,
          "validation_metric": 0.0012961954344063997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6231778370001848
        },
        {
          "epoch": 10,
          "updates": 640,
          "samples_seen": 10240,
          "online_loss": 0.020033576111018192,
          "train_loss": 0.015082152560353279,
          "validation_loss": 0.012922706082463264,
          "train_metric": 0.0015190191334113479,
          "validation_metric": 0.0013015277218073606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6848866710006405
        },
        {
          "epoch": 11,
          "updates": 704,
          "samples_seen": 11264,
          "online_loss": 0.018872276048568892,
          "train_loss": 0.01877380535006523,
          "validation_loss": 0.016785724088549614,
          "train_metric": 0.0018908288329839706,
          "validation_metric": 0.0016905968077480793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.746307515000808
        },
        {
          "epoch": 12,
          "updates": 768,
          "samples_seen": 12288,
          "online_loss": 0.015564517714665271,
          "train_loss": 0.019443897530436516,
          "validation_loss": 0.016875337809324265,
          "train_metric": 0.001958318054676056,
          "validation_metric": 0.001699622254818678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.789054851000401
        },
        {
          "epoch": 13,
          "updates": 832,
          "samples_seen": 13312,
          "online_loss": 0.02062957177622593,
          "train_loss": 0.013845331966876984,
          "validation_loss": 0.011532033793628216,
          "train_metric": 0.001394451130181551,
          "validation_metric": 0.0011614641407504678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.834125508001307
        },
        {
          "epoch": 14,
          "updates": 896,
          "samples_seen": 14336,
          "online_loss": 0.014163261696012341,
          "train_loss": 0.013045270927250385,
          "validation_loss": 0.011753086000680923,
          "train_metric": 0.0013138719368726015,
          "validation_metric": 0.001183727872557938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8783180120008183
        },
        {
          "epoch": 15,
          "updates": 960,
          "samples_seen": 15360,
          "online_loss": 0.018057347393551026,
          "train_loss": 0.010781795717775822,
          "validation_loss": 0.00936091784387827,
          "train_metric": 0.0010859029134735465,
          "validation_metric": 0.0009427974000573158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.922916626001097
        },
        {
          "epoch": 16,
          "updates": 1024,
          "samples_seen": 16384,
          "online_loss": 0.01201313555611705,
          "train_loss": 0.010476985014975071,
          "validation_loss": 0.009610496461391449,
          "train_metric": 0.0010552034946158528,
          "validation_metric": 0.0009679339709691703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9688877370008413
        },
        {
          "epoch": 17,
          "updates": 1088,
          "samples_seen": 17408,
          "online_loss": 0.014973890263718204,
          "train_loss": 0.017886243760585785,
          "validation_loss": 0.013730121776461601,
          "train_metric": 0.0018014369998127222,
          "validation_metric": 0.001382847665809095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0128747569997358
        },
        {
          "epoch": 18,
          "updates": 1152,
          "samples_seen": 18432,
          "online_loss": 0.013085429758575629,
          "train_loss": 0.011683696880936623,
          "validation_loss": 0.009333208203315735,
          "train_metric": 0.0011767392279580235,
          "validation_metric": 0.0009400066337548196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0579825719996734
        },
        {
          "epoch": 19,
          "updates": 1216,
          "samples_seen": 19456,
          "online_loss": 0.017297398122536833,
          "train_loss": 0.02513301372528076,
          "validation_loss": 0.020134035497903824,
          "train_metric": 0.0025313054211437702,
          "validation_metric": 0.0020278263837099075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.102040871999634
        },
        {
          "epoch": 20,
          "updates": 1280,
          "samples_seen": 20480,
          "online_loss": 0.011894399105585762,
          "train_loss": 0.008214673027396202,
          "validation_loss": 0.007204909808933735,
          "train_metric": 0.0008273517014458776,
          "validation_metric": 0.000725652149412781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1465496410000924
        },
        {
          "epoch": 21,
          "updates": 1344,
          "samples_seen": 21504,
          "online_loss": 0.010655492063051497,
          "train_loss": 0.0107549037784338,
          "validation_loss": 0.010437926277518272,
          "train_metric": 0.0010831945110112429,
          "validation_metric": 0.001051269588060677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1890568279995932
        },
        {
          "epoch": 22,
          "updates": 1408,
          "samples_seen": 22528,
          "online_loss": 0.012774481894666678,
          "train_loss": 0.008350655436515808,
          "validation_loss": 0.006709737237542868,
          "train_metric": 0.0008410474401898682,
          "validation_metric": 0.0006757801165804267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2449800690010306
        },
        {
          "epoch": 23,
          "updates": 1472,
          "samples_seen": 23552,
          "online_loss": 0.010650439867276873,
          "train_loss": 0.007794754579663277,
          "validation_loss": 0.007431033067405224,
          "train_metric": 0.0007850591791793704,
          "validation_metric": 0.0007484264788217843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2768182979998528
        },
        {
          "epoch": 24,
          "updates": 1536,
          "samples_seen": 24576,
          "online_loss": 0.009276645722820831,
          "train_loss": 0.007059791125357151,
          "validation_loss": 0.006250465288758278,
          "train_metric": 0.0007110362639650702,
          "validation_metric": 0.000629523943644017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.307732761000807
        },
        {
          "epoch": 25,
          "updates": 1600,
          "samples_seen": 25600,
          "online_loss": 0.008429100917965116,
          "train_loss": 0.006332362070679665,
          "validation_loss": 0.005964370444417,
          "train_metric": 0.000637772202026099,
          "validation_metric": 0.000600709521677345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.339141655000276
        },
        {
          "epoch": 26,
          "updates": 1664,
          "samples_seen": 26624,
          "online_loss": 0.008373810198463616,
          "train_loss": 0.007454246748238802,
          "validation_loss": 0.0069554271176457405,
          "train_metric": 0.0007507645059376955,
          "validation_metric": 0.0007005251245573163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3700396029998956
        },
        {
          "epoch": 27,
          "updates": 1728,
          "samples_seen": 27648,
          "online_loss": 0.009199855850056338,
          "train_loss": 0.00636113528162241,
          "validation_loss": 0.004939631558954716,
          "train_metric": 0.0006406703032553196,
          "validation_metric": 0.0004975015763193369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4012001110004348
        },
        {
          "epoch": 28,
          "updates": 1792,
          "samples_seen": 28672,
          "online_loss": 0.00930646253618761,
          "train_loss": 0.007356925401836634,
          "validation_loss": 0.006398562341928482,
          "train_metric": 0.0007409625686705112,
          "validation_metric": 0.0006444397149607539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4326455210011773
        },
        {
          "epoch": 29,
          "updates": 1856,
          "samples_seen": 29696,
          "online_loss": 0.01466073271149071,
          "train_loss": 0.023007825016975403,
          "validation_loss": 0.01775229535996914,
          "train_metric": 0.002317263977602124,
          "validation_metric": 0.001787946093827486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4640299279999454
        },
        {
          "epoch": 30,
          "updates": 1920,
          "samples_seen": 30720,
          "online_loss": 0.0096389428990733,
          "train_loss": 0.007106661330908537,
          "validation_loss": 0.0064910296350717545,
          "train_metric": 0.0007157569634728134,
          "validation_metric": 0.0006537525914609432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4952359520011669
        },
        {
          "epoch": 31,
          "updates": 1984,
          "samples_seen": 31744,
          "online_loss": 0.008078788097918732,
          "train_loss": 0.008794700726866722,
          "validation_loss": 0.007007361855357885,
          "train_metric": 0.0008857700740918517,
          "validation_metric": 0.0007057557813823223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5258068290004303
        },
        {
          "epoch": 32,
          "updates": 2048,
          "samples_seen": 32768,
          "online_loss": 0.012073777893419901,
          "train_loss": 0.011158484034240246,
          "validation_loss": 0.00796123594045639,
          "train_metric": 0.00112384173553437,
          "validation_metric": 0.0008018265361897647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5568662929999846
        },
        {
          "epoch": 33,
          "updates": 2112,
          "samples_seen": 33792,
          "online_loss": 0.007957637692925346,
          "train_loss": 0.006528130732476711,
          "validation_loss": 0.004314207937568426,
          "train_metric": 0.0006574893486686051,
          "validation_metric": 0.00043451119563542306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5877918879996287
        },
        {
          "epoch": 34,
          "updates": 2176,
          "samples_seen": 34816,
          "online_loss": 0.006950728659830929,
          "train_loss": 0.00783917773514986,
          "validation_loss": 0.007652789819985628,
          "train_metric": 0.0007895331946201622,
          "validation_metric": 0.0007707609329372644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.618816460000744
        },
        {
          "epoch": 35,
          "updates": 2240,
          "samples_seen": 35840,
          "online_loss": 0.011103222006568103,
          "train_loss": 0.01054596621543169,
          "validation_loss": 0.007749403361231089,
          "train_metric": 0.0010621510446071625,
          "validation_metric": 0.0007804914494045079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6503224030002457
        },
        {
          "epoch": 36,
          "updates": 2304,
          "samples_seen": 36864,
          "online_loss": 0.013304896634508623,
          "train_loss": 0.029836036264896393,
          "validation_loss": 0.021317539736628532,
          "train_metric": 0.0030049760825932026,
          "validation_metric": 0.0021470242645591497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6809207950009295
        },
        {
          "epoch": 37,
          "updates": 2368,
          "samples_seen": 37888,
          "online_loss": 0.016832370956763043,
          "train_loss": 0.006985950283706188,
          "validation_loss": 0.006190818268805742,
          "train_metric": 0.0007035993621684611,
          "validation_metric": 0.0006235165637917817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7120163250001497
        },
        {
          "epoch": 38,
          "updates": 2432,
          "samples_seen": 38912,
          "online_loss": 0.008982905978882627,
          "train_loss": 0.005726327188313007,
          "validation_loss": 0.004953449591994286,
          "train_metric": 0.0005767346592620015,
          "validation_metric": 0.000498893263284117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7424043950013584
        },
        {
          "epoch": 39,
          "updates": 2496,
          "samples_seen": 39936,
          "online_loss": 0.010092989678923914,
          "train_loss": 0.005423896946012974,
          "validation_loss": 0.004181575030088425,
          "train_metric": 0.0005462749977596104,
          "validation_metric": 0.0004211529449094087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.772858582000481
        },
        {
          "epoch": 40,
          "updates": 2560,
          "samples_seen": 40960,
          "online_loss": 0.0060147262302052695,
          "train_loss": 0.004635312594473362,
          "validation_loss": 0.003806495340541005,
          "train_metric": 0.0004668517503887415,
          "validation_metric": 0.00038337628939189017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.80481364499974
        },
        {
          "epoch": 41,
          "updates": 2624,
          "samples_seen": 41984,
          "online_loss": 0.008449524229490635,
          "train_loss": 0.009961613453924656,
          "validation_loss": 0.007583096623420715,
          "train_metric": 0.0010032971622422338,
          "validation_metric": 0.0007637416711077094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.835598781999579
        },
        {
          "epoch": 42,
          "updates": 2688,
          "samples_seen": 43008,
          "online_loss": 0.010964764693198958,
          "train_loss": 0.006015809252858162,
          "validation_loss": 0.004535278771072626,
          "train_metric": 0.0006058902363292873,
          "validation_metric": 0.00045677664456889033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.86635274100081
        },
        {
          "epoch": 43,
          "updates": 2752,
          "samples_seen": 44032,
          "online_loss": 0.007416923617711291,
          "train_loss": 0.02131902612745762,
          "validation_loss": 0.015902819111943245,
          "train_metric": 0.00214717467315495,
          "validation_metric": 0.0016016735462471843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8965189900009136
        },
        {
          "epoch": 44,
          "updates": 2816,
          "samples_seen": 45056,
          "online_loss": 0.009653466448071413,
          "train_loss": 0.006203942932188511,
          "validation_loss": 0.006065302528440952,
          "train_metric": 0.0006248384597711265,
          "validation_metric": 0.0006108751404099166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.989852394999616
        },
        {
          "epoch": 45,
          "updates": 2880,
          "samples_seen": 46080,
          "online_loss": 0.0053833390011277515,
          "train_loss": 0.0045046391896903515,
          "validation_loss": 0.003960290923714638,
          "train_metric": 0.0004536907363217324,
          "validation_metric": 0.00039886601734906435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0766153479999048
        },
        {
          "epoch": 46,
          "updates": 2944,
          "samples_seen": 47104,
          "online_loss": 0.006484148193521833,
          "train_loss": 0.003978617489337921,
          "validation_loss": 0.002959044650197029,
          "train_metric": 0.00040071175317279994,
          "validation_metric": 0.0002980241260956973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1226251019998017
        },
        {
          "epoch": 47,
          "updates": 3008,
          "samples_seen": 48128,
          "online_loss": 0.016418997463006235,
          "train_loss": 0.010933425277471542,
          "validation_loss": 0.008622393012046814,
          "train_metric": 0.0011011746246367693,
          "validation_metric": 0.0008684158674441278,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1689507860010053
        },
        {
          "epoch": 48,
          "updates": 3072,
          "samples_seen": 49152,
          "online_loss": 0.007820113906291226,
          "train_loss": 0.006071016192436218,
          "validation_loss": 0.004279268905520439,
          "train_metric": 0.0006114504649303854,
          "validation_metric": 0.00043099233880639076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2156898160010314
        },
        {
          "epoch": 49,
          "updates": 3136,
          "samples_seen": 50176,
          "online_loss": 0.006070688801628421,
          "train_loss": 0.0166839100420475,
          "validation_loss": 0.009765399619936943,
          "train_metric": 0.001680342247709632,
          "validation_metric": 0.0009835351957008243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2623374000013428
        },
        {
          "epoch": 50,
          "updates": 3200,
          "samples_seen": 51200,
          "online_loss": 0.015345618134233519,
          "train_loss": 0.008599977940320969,
          "validation_loss": 0.0065785921178758144,
          "train_metric": 0.0008661583997309208,
          "validation_metric": 0.0006625717505812645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3127661470007297
        },
        {
          "epoch": 51,
          "updates": 3264,
          "samples_seen": 52224,
          "online_loss": 0.009465241448197048,
          "train_loss": 0.006852529477328062,
          "validation_loss": 0.006664183922111988,
          "train_metric": 0.0006901616579852998,
          "validation_metric": 0.0006711921887472272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.359110277000582
        },
        {
          "epoch": 52,
          "updates": 3328,
          "samples_seen": 53248,
          "online_loss": 0.007058567637159285,
          "train_loss": 0.005111368838697672,
          "validation_loss": 0.004578143358230591,
          "train_metric": 0.0005147983320057392,
          "validation_metric": 0.0004610938485711813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4059704080009396
        },
        {
          "epoch": 53,
          "updates": 3392,
          "samples_seen": 54272,
          "online_loss": 0.006453309148128028,
          "train_loss": 0.008909109979867935,
          "validation_loss": 0.009453926235437393,
          "train_metric": 0.0008972930372692645,
          "validation_metric": 0.000952164817135781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.454169119000653
        },
        {
          "epoch": 54,
          "updates": 3456,
          "samples_seen": 55296,
          "online_loss": 0.007128656175154902,
          "train_loss": 0.011307896114885807,
          "validation_loss": 0.007331364788115025,
          "train_metric": 0.001138889929279685,
          "validation_metric": 0.0007383882184512913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5302964900001825
        },
        {
          "epoch": 55,
          "updates": 3520,
          "samples_seen": 56320,
          "online_loss": 0.005784142787888413,
          "train_loss": 0.008087782189249992,
          "validation_loss": 0.00510256364941597,
          "train_metric": 0.0008145718602463603,
          "validation_metric": 0.0005139115964993834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6236765140001808
        },
        {
          "epoch": 56,
          "updates": 3584,
          "samples_seen": 57344,
          "online_loss": 0.008943350635945535,
          "train_loss": 0.006036847829818726,
          "validation_loss": 0.0046520228497684,
          "train_metric": 0.0006080091698095202,
          "validation_metric": 0.00046853473759256303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6958539890001703
        },
        {
          "epoch": 57,
          "updates": 3648,
          "samples_seen": 58368,
          "online_loss": 0.008918125940908794,
          "train_loss": 0.005299156531691551,
          "validation_loss": 0.003823417704552412,
          "train_metric": 0.0005337115726433694,
          "validation_metric": 0.00038508063880726695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7523524070002168
        },
        {
          "epoch": 58,
          "updates": 3712,
          "samples_seen": 59392,
          "online_loss": 0.00691817045208154,
          "train_loss": 0.006041401065886021,
          "validation_loss": 0.00387282925657928,
          "train_metric": 0.0006084677879698575,
          "validation_metric": 0.0003900571900885552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7982210280006257
        },
        {
          "epoch": 59,
          "updates": 3776,
          "samples_seen": 60416,
          "online_loss": 0.008093491730960523,
          "train_loss": 0.003821650054305792,
          "validation_loss": 0.003288802457973361,
          "train_metric": 0.000384902668884024,
          "validation_metric": 0.0003312361368443817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8455640690008295
        },
        {
          "epoch": 60,
          "updates": 3840,
          "samples_seen": 61440,
          "online_loss": 0.004463181084929602,
          "train_loss": 0.003290013875812292,
          "validation_loss": 0.003012512344866991,
          "train_metric": 0.0003313581692054868,
          "validation_metric": 0.00030340917874127626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8925769950001268
        },
        {
          "epoch": 61,
          "updates": 3904,
          "samples_seen": 62464,
          "online_loss": 0.004352869989816099,
          "train_loss": 0.0034724974539130926,
          "validation_loss": 0.002703029429540038,
          "train_metric": 0.00034973720903508365,
          "validation_metric": 0.0002722392091527581,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9399428169999737
        },
        {
          "epoch": 62,
          "updates": 3968,
          "samples_seen": 63488,
          "online_loss": 0.004165498351085262,
          "train_loss": 0.006252744235098362,
          "validation_loss": 0.006505170837044716,
          "train_metric": 0.0006297535146586597,
          "validation_metric": 0.0006551771075464785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.020723765999719
        },
        {
          "epoch": 63,
          "updates": 4032,
          "samples_seen": 64512,
          "online_loss": 0.007432392657392484,
          "train_loss": 0.012663591653108597,
          "validation_loss": 0.009089037775993347,
          "train_metric": 0.0012754305498674512,
          "validation_metric": 0.0009154145373031497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.074764603999938
        },
        {
          "epoch": 64,
          "updates": 4096,
          "samples_seen": 65536,
          "online_loss": 0.012173054402410344,
          "train_loss": 0.004426517523825169,
          "validation_loss": 0.003216005163267255,
          "train_metric": 0.00044582263217307627,
          "validation_metric": 0.00032390429987572134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1248252960012906
        },
        {
          "epoch": 65,
          "updates": 4160,
          "samples_seen": 66560,
          "online_loss": 0.004716051106697705,
          "train_loss": 0.0128056975081563,
          "validation_loss": 0.010926077142357826,
          "train_metric": 0.0012897428823634982,
          "validation_metric": 0.0011004344560205936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.170885896000982
        },
        {
          "epoch": 66,
          "updates": 4224,
          "samples_seen": 67584,
          "online_loss": 0.006257149811517593,
          "train_loss": 0.008384747430682182,
          "validation_loss": 0.0063104466535151005,
          "train_metric": 0.0008444810518994927,
          "validation_metric": 0.0006355652003549039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2161901960007526
        },
        {
          "epoch": 67,
          "updates": 4288,
          "samples_seen": 68608,
          "online_loss": 0.006724030254190438,
          "train_loss": 0.00698777986690402,
          "validation_loss": 0.004529524594545364,
          "train_metric": 0.0007037835312075913,
          "validation_metric": 0.0004561970999930054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.260663569000826
        },
        {
          "epoch": 68,
          "updates": 4352,
          "samples_seen": 69632,
          "online_loss": 0.005272652818803181,
          "train_loss": 0.004513835534453392,
          "validation_loss": 0.0032315505668520927,
          "train_metric": 0.00045461702393367887,
          "validation_metric": 0.00032546991133131087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.306085136000547
        },
        {
          "epoch": 69,
          "updates": 4416,
          "samples_seen": 70656,
          "online_loss": 0.006018569374191429,
          "train_loss": 0.008701360784471035,
          "validation_loss": 0.009028051048517227,
          "train_metric": 0.0008763692458160222,
          "validation_metric": 0.000909272232092917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3522403330007364
        },
        {
          "epoch": 70,
          "updates": 4480,
          "samples_seen": 71680,
          "online_loss": 0.005956256103900159,
          "train_loss": 0.010017328895628452,
          "validation_loss": 0.009249208495020866,
          "train_metric": 0.0010089086135849357,
          "validation_metric": 0.0009315463248640299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.482099107999602
        },
        {
          "epoch": 71,
          "updates": 4544,
          "samples_seen": 72704,
          "online_loss": 0.007939313002680137,
          "train_loss": 0.004941737279295921,
          "validation_loss": 0.002979236887767911,
          "train_metric": 0.0004977137432433665,
          "validation_metric": 0.000300057785352692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5690747540011216
        },
        {
          "epoch": 72,
          "updates": 4608,
          "samples_seen": 73728,
          "online_loss": 0.005594404125531582,
          "train_loss": 0.0040687499567866325,
          "validation_loss": 0.003510915907099843,
          "train_metric": 0.00040978958713822067,
          "validation_metric": 0.00035360653419047594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.616039993999948
        },
        {
          "epoch": 73,
          "updates": 4672,
          "samples_seen": 74752,
          "online_loss": 0.004871920091318316,
          "train_loss": 0.0078071728348731995,
          "validation_loss": 0.006149535998702049,
          "train_metric": 0.0007863097125664353,
          "validation_metric": 0.0006193587323650718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6631765230013116
        },
        {
          "epoch": 74,
          "updates": 4736,
          "samples_seen": 75776,
          "online_loss": 0.004173669505689759,
          "train_loss": 0.002356225624680519,
          "validation_loss": 0.001823133323341608,
          "train_metric": 0.00023731040710117668,
          "validation_metric": 0.00018361931142862886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.710893372999635
        },
        {
          "epoch": 75,
          "updates": 4800,
          "samples_seen": 76800,
          "online_loss": 0.0037187473305948515,
          "train_loss": 0.004010722041130066,
          "validation_loss": 0.0020855965558439493,
          "train_metric": 0.00040394527604803443,
          "validation_metric": 0.00021005365124437958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.756966070000999
        },
        {
          "epoch": 76,
          "updates": 4864,
          "samples_seen": 77824,
          "online_loss": 0.0043592026927399274,
          "train_loss": 0.003421632805839181,
          "validation_loss": 0.0020599341951310635,
          "train_metric": 0.0003446143527980894,
          "validation_metric": 0.0002074690128210932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8027231420001044
        },
        {
          "epoch": 77,
          "updates": 4928,
          "samples_seen": 78848,
          "online_loss": 0.004864849133127791,
          "train_loss": 0.0025047315284609795,
          "validation_loss": 0.0023085603024810553,
          "train_metric": 0.0002522673748899251,
          "validation_metric": 0.0002325097593711689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8557454600013443
        },
        {
          "epoch": 78,
          "updates": 4992,
          "samples_seen": 79872,
          "online_loss": 0.003288205200874472,
          "train_loss": 0.002371746115386486,
          "validation_loss": 0.001864442601799965,
          "train_metric": 0.00023887358838692307,
          "validation_metric": 0.000187779834959656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9019557499996154
        },
        {
          "epoch": 79,
          "updates": 5056,
          "samples_seen": 80896,
          "online_loss": 0.004240113017203839,
          "train_loss": 0.010669159702956676,
          "validation_loss": 0.008121848106384277,
          "train_metric": 0.0010745585896074772,
          "validation_metric": 0.0008180027361959219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9488994050007022
        },
        {
          "epoch": 80,
          "updates": 5120,
          "samples_seen": 81920,
          "online_loss": 0.0045353219061325944,
          "train_loss": 0.005180539563298225,
          "validation_loss": 0.003077478613704443,
          "train_metric": 0.0005217649741098285,
          "validation_metric": 0.0003099524474237114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.995771035000871
        },
        {
          "epoch": 81,
          "updates": 5184,
          "samples_seen": 82944,
          "online_loss": 0.003719690798789088,
          "train_loss": 0.002674693241715431,
          "validation_loss": 0.002011821372434497,
          "train_metric": 0.0002693852875381708,
          "validation_metric": 0.00020262326870579273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.042472467999687
        },
        {
          "epoch": 82,
          "updates": 5248,
          "samples_seen": 83968,
          "online_loss": 0.0036845469480795146,
          "train_loss": 0.004945295862853527,
          "validation_loss": 0.00383618101477623,
          "train_metric": 0.0004980720113962889,
          "validation_metric": 0.0003863660676870495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.088154873999883
        },
        {
          "epoch": 83,
          "updates": 5312,
          "samples_seen": 84992,
          "online_loss": 0.004408694331118568,
          "train_loss": 0.00442283321171999,
          "validation_loss": 0.0034648089203983545,
          "train_metric": 0.00044545161654241383,
          "validation_metric": 0.00034896290162578225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.139420755000174
        },
        {
          "epoch": 84,
          "updates": 5376,
          "samples_seen": 86016,
          "online_loss": 0.007575932747386105,
          "train_loss": 0.010662883520126343,
          "validation_loss": 0.00888698361814022,
          "train_metric": 0.001073926454409957,
          "validation_metric": 0.0008950645569711924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.192755150999801
        },
        {
          "epoch": 85,
          "updates": 5440,
          "samples_seen": 87040,
          "online_loss": 0.0038827320995551418,
          "train_loss": 0.0019891983829438686,
          "validation_loss": 0.0016552208689972758,
          "train_metric": 0.00020034477347508073,
          "validation_metric": 0.00016670780314598233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.241699096000957
        },
        {
          "epoch": 86,
          "updates": 5504,
          "samples_seen": 88064,
          "online_loss": 0.007294223024473467,
          "train_loss": 0.007203786168247461,
          "validation_loss": 0.004230677615851164,
          "train_metric": 0.0007255389355123043,
          "validation_metric": 0.00042609835509210825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.287435068999912
        },
        {
          "epoch": 87,
          "updates": 5568,
          "samples_seen": 89088,
          "online_loss": 0.004738493926652154,
          "train_loss": 0.007943075150251389,
          "validation_loss": 0.004688540939241648,
          "train_metric": 0.000799997360445559,
          "validation_metric": 0.0004722127050627023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.333419456001138
        },
        {
          "epoch": 88,
          "updates": 5632,
          "samples_seen": 90112,
          "online_loss": 0.011721419115474419,
          "train_loss": 0.01060440018773079,
          "validation_loss": 0.009423981420695782,
          "train_metric": 0.0010680364212021232,
          "validation_metric": 0.0009491488453932106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.379703234000772
        },
        {
          "epoch": 89,
          "updates": 5696,
          "samples_seen": 91136,
          "online_loss": 0.00591257628184394,
          "train_loss": 0.005927795544266701,
          "validation_loss": 0.006248519755899906,
          "train_metric": 0.0005970258498564363,
          "validation_metric": 0.0006293280748650432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.42555289899974
        },
        {
          "epoch": 90,
          "updates": 5760,
          "samples_seen": 92160,
          "online_loss": 0.0035983590005344013,
          "train_loss": 0.002484923927113414,
          "validation_loss": 0.0022280151024460793,
          "train_metric": 0.0002502724528312683,
          "validation_metric": 0.00022439751774072647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.4708355920010945
        },
        {
          "epoch": 91,
          "updates": 5824,
          "samples_seen": 93184,
          "online_loss": 0.003199404984115972,
          "train_loss": 0.005665736738592386,
          "validation_loss": 0.0027013318613171577,
          "train_metric": 0.0005706321680918336,
          "validation_metric": 0.0002720682823564857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.518886437001129
        },
        {
          "epoch": 92,
          "updates": 5888,
          "samples_seen": 94208,
          "online_loss": 0.00798844245946384,
          "train_loss": 0.004337705671787262,
          "validation_loss": 0.0032742165494710207,
          "train_metric": 0.0004368778318166733,
          "validation_metric": 0.0003297670918982476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.565783146001195
        },
        {
          "epoch": 93,
          "updates": 5952,
          "samples_seen": 95232,
          "online_loss": 0.003971661042669439,
          "train_loss": 0.002238109242171049,
          "validation_loss": 0.001718203886412084,
          "train_metric": 0.00022541415819432586,
          "validation_metric": 0.00017305120127275586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.615604836000784
        },
        {
          "epoch": 94,
          "updates": 6016,
          "samples_seen": 96256,
          "online_loss": 0.0026776953263834002,
          "train_loss": 0.010557277128100395,
          "validation_loss": 0.0067505002953112125,
          "train_metric": 0.0010632902849465609,
          "validation_metric": 0.0006798856193199754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.6614838799996505
        },
        {
          "epoch": 95,
          "updates": 6080,
          "samples_seen": 97280,
          "online_loss": 0.011971426592026546,
          "train_loss": 0.003737105056643486,
          "validation_loss": 0.0029039476066827774,
          "train_metric": 0.00037638749927282333,
          "validation_metric": 0.00029247498605400324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.732259136000721
        },
        {
          "epoch": 96,
          "updates": 6144,
          "samples_seen": 98304,
          "online_loss": 0.004304741366013332,
          "train_loss": 0.0035538412630558014,
          "validation_loss": 0.002336077392101288,
          "train_metric": 0.0003579298499971628,
          "validation_metric": 0.0002352811861783266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.790437699000904
        },
        {
          "epoch": 97,
          "updates": 6208,
          "samples_seen": 99328,
          "online_loss": 0.0037614408904573793,
          "train_loss": 0.0037960237823426723,
          "validation_loss": 0.0028114269953221083,
          "train_metric": 0.0003823215956799686,
          "validation_metric": 0.0002831566089298576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.836690479000026
        },
        {
          "epoch": 98,
          "updates": 6272,
          "samples_seen": 100352,
          "online_loss": 0.006343052502415958,
          "train_loss": 0.0027959514409303665,
          "validation_loss": 0.002410676795989275,
          "train_metric": 0.00028159801149740815,
          "validation_metric": 0.00024279451463371515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.88643648300058
        },
        {
          "epoch": 99,
          "updates": 6336,
          "samples_seen": 101376,
          "online_loss": 0.003013926684729995,
          "train_loss": 0.002501944312825799,
          "validation_loss": 0.0017032683826982975,
          "train_metric": 0.0002519866975490004,
          "validation_metric": 0.00017154694069176912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.951787621999756
        },
        {
          "epoch": 100,
          "updates": 6400,
          "samples_seen": 102400,
          "online_loss": 0.008490440212199246,
          "train_loss": 0.008317200466990471,
          "validation_loss": 0.00631683086976409,
          "train_metric": 0.0008376780315302312,
          "validation_metric": 0.0006362080457620323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.998659298000348
        },
        {
          "epoch": 101,
          "updates": 6464,
          "samples_seen": 103424,
          "online_loss": 0.01879721933073597,
          "train_loss": 0.003605456557124853,
          "validation_loss": 0.002894693985581398,
          "train_metric": 0.0003631283761933446,
          "validation_metric": 0.00029154293588362634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.044854215000669
        },
        {
          "epoch": 102,
          "updates": 6528,
          "samples_seen": 104448,
          "online_loss": 0.004222330051106837,
          "train_loss": 0.002692158566787839,
          "validation_loss": 0.002565619768574834,
          "train_metric": 0.0002711443230509758,
          "validation_metric": 0.0002583998430054635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.090014463999978
        },
        {
          "epoch": 103,
          "updates": 6592,
          "samples_seen": 105472,
          "online_loss": 0.00272432444739934,
          "train_loss": 0.0023383116349577904,
          "validation_loss": 0.0017755122389644384,
          "train_metric": 0.00023550620244350284,
          "validation_metric": 0.00017882310203276575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.141193659999772
        },
        {
          "epoch": 104,
          "updates": 6656,
          "samples_seen": 106496,
          "online_loss": 0.003044774387490179,
          "train_loss": 0.002394182374700904,
          "validation_loss": 0.002257047686725855,
          "train_metric": 0.0002411332679912448,
          "validation_metric": 0.00022732156503479928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.209285631999592
        },
        {
          "epoch": 105,
          "updates": 6720,
          "samples_seen": 107520,
          "online_loss": 0.0034756207386408278,
          "train_loss": 0.001918851980008185,
          "validation_loss": 0.001382662681862712,
          "train_metric": 0.0001932597515406087,
          "validation_metric": 0.00013925672101322562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.2647794090007665
        },
        {
          "epoch": 106,
          "updates": 6784,
          "samples_seen": 108544,
          "online_loss": 0.003361191079193304,
          "train_loss": 0.00826833676546812,
          "validation_loss": 0.004804221447557211,
          "train_metric": 0.0008327566320076585,
          "validation_metric": 0.00048386355047114193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.312651698999616
        },
        {
          "epoch": 107,
          "updates": 6848,
          "samples_seen": 109568,
          "online_loss": 0.0040781988800517865,
          "train_loss": 0.005453233607113361,
          "validation_loss": 0.0035039670765399933,
          "train_metric": 0.0005492297350428998,
          "validation_metric": 0.0003529067325871438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.359563619000255
        },
        {
          "epoch": 108,
          "updates": 6912,
          "samples_seen": 110592,
          "online_loss": 0.00817460371490597,
          "train_loss": 0.0022446054499596357,
          "validation_loss": 0.0015055319527164102,
          "train_metric": 0.00022606845595873892,
          "validation_metric": 0.00015163168427534401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.407217855999988
        },
        {
          "epoch": 109,
          "updates": 6976,
          "samples_seen": 111616,
          "online_loss": 0.005550383959189276,
          "train_loss": 0.0048477742820978165,
          "validation_loss": 0.0044470420107245445,
          "train_metric": 0.00048825007979758084,
          "validation_metric": 0.0004478897899389267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.454903935000402
        },
        {
          "epoch": 110,
          "updates": 7040,
          "samples_seen": 112640,
          "online_loss": 0.004945261117882183,
          "train_loss": 0.00450520683079958,
          "validation_loss": 0.003524243598803878,
          "train_metric": 0.0004537478962447494,
          "validation_metric": 0.0003549489192664623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.501856748000137
        },
        {
          "epoch": 111,
          "updates": 7104,
          "samples_seen": 113664,
          "online_loss": 0.003949655144424469,
          "train_loss": 0.002953133312985301,
          "validation_loss": 0.0022415088023990393,
          "train_metric": 0.00029742883634753525,
          "validation_metric": 0.00022575657931156456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.548228181000013
        },
        {
          "epoch": 112,
          "updates": 7168,
          "samples_seen": 114688,
          "online_loss": 0.003407112381751176,
          "train_loss": 0.003416326129809022,
          "validation_loss": 0.0027199655305594206,
          "train_metric": 0.000344079831847921,
          "validation_metric": 0.0002739449846558273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.594824499999959
        },
        {
          "epoch": 113,
          "updates": 7232,
          "samples_seen": 115712,
          "online_loss": 0.009466864075420744,
          "train_loss": 0.01245089527219534,
          "validation_loss": 0.008253354579210281,
          "train_metric": 0.0012540085008367896,
          "validation_metric": 0.0008312475983984768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.64206127000034
        },
        {
          "epoch": 114,
          "updates": 7296,
          "samples_seen": 116736,
          "online_loss": 0.005046320752171596,
          "train_loss": 0.0029177602846175432,
          "validation_loss": 0.0016532635781913996,
          "train_metric": 0.0002938661491498351,
          "validation_metric": 0.0001665106537984684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.689296333001039
        },
        {
          "epoch": 115,
          "updates": 7360,
          "samples_seen": 117760,
          "online_loss": 0.002212162263049322,
          "train_loss": 0.0016986411064863205,
          "validation_loss": 0.0013938716147094965,
          "train_metric": 0.00017108090105466545,
          "validation_metric": 0.00014038562949281186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.735843921000196
        },
        {
          "epoch": 116,
          "updates": 7424,
          "samples_seen": 118784,
          "online_loss": 0.0026622235166087194,
          "train_loss": 0.0020963100250810385,
          "validation_loss": 0.0015677736373618245,
          "train_metric": 0.00021113270486239344,
          "validation_metric": 0.00015790040197316557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.782005046999984
        },
        {
          "epoch": 117,
          "updates": 7488,
          "samples_seen": 119808,
          "online_loss": 0.011872993472024973,
          "train_loss": 0.020118914544582367,
          "validation_loss": 0.020075948908925056,
          "train_metric": 0.002026303205639124,
          "validation_metric": 0.0020219760481268167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.837210433001019
        },
        {
          "epoch": 118,
          "updates": 7552,
          "samples_seen": 120832,
          "online_loss": 0.010087263644436462,
          "train_loss": 0.002743080258369446,
          "validation_loss": 0.0020887700375169516,
          "train_metric": 0.00027627302915789187,
          "validation_metric": 0.00021037328406237066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.886785572000008
        },
        {
          "epoch": 119,
          "updates": 7616,
          "samples_seen": 121856,
          "online_loss": 0.002670980557695657,
          "train_loss": 0.002120147692039609,
          "validation_loss": 0.0014766647946089506,
          "train_metric": 0.00021353350894059986,
          "validation_metric": 0.00014872428437229246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.919974643000387
        },
        {
          "epoch": 120,
          "updates": 7680,
          "samples_seen": 122880,
          "online_loss": 0.0025485546545951365,
          "train_loss": 0.0019730094354599714,
          "validation_loss": 0.001693853409960866,
          "train_metric": 0.00019871428958140314,
          "validation_metric": 0.00017059869423974305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.952320962000158
        },
        {
          "epoch": 121,
          "updates": 7744,
          "samples_seen": 123904,
          "online_loss": 0.0037176685943904886,
          "train_loss": 0.006629216019064188,
          "validation_loss": 0.0037708920426666737,
          "train_metric": 0.0006676703342236578,
          "validation_metric": 0.0003797904937528074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.985989927001356
        },
        {
          "epoch": 122,
          "updates": 7808,
          "samples_seen": 124928,
          "online_loss": 0.005823216934913944,
          "train_loss": 0.01072394847869873,
          "validation_loss": 0.008465679362416267,
          "train_metric": 0.0010800767922773957,
          "validation_metric": 0.0008526321616955101,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.0193182109997
        },
        {
          "epoch": 123,
          "updates": 7872,
          "samples_seen": 125952,
          "online_loss": 0.007210088224383071,
          "train_loss": 0.008426467888057232,
          "validation_loss": 0.007178532425314188,
          "train_metric": 0.000848683004733175,
          "validation_metric": 0.0007229954353533685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.0524829879996105
        },
        {
          "epoch": 124,
          "updates": 7936,
          "samples_seen": 126976,
          "online_loss": 0.012556607656733831,
          "train_loss": 0.008719703182578087,
          "validation_loss": 0.007070390973240137,
          "train_metric": 0.0008782166405580938,
          "validation_metric": 0.0007121039088815451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.0856215940002585
        },
        {
          "epoch": 125,
          "updates": 8000,
          "samples_seen": 128000,
          "online_loss": 0.004346389055626787,
          "train_loss": 0.011611470952630043,
          "validation_loss": 0.00921515841037035,
          "train_metric": 0.0011694647837430239,
          "validation_metric": 0.000928116962313652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.118897700000161
        }
      ],
      "best": {
        "epoch": 105,
        "updates": 6720
      },
      "best_validation_metric": 0.00013925672101322562,
      "config": {
        "label": "Batch 16",
        "batch": 16,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.011800708496239775
    },
    "be8cbd994d03_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.25128196459263563,
          "train_loss": 0.1026863232254982,
          "validation_loss": 0.07346689701080322,
          "train_metric": 0.010342191904783249,
          "validation_metric": 0.007399316877126694,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.0171097889997327
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07234213163610548,
          "train_loss": 0.07510144263505936,
          "validation_loss": 0.06354568153619766,
          "train_metric": 0.007563942577689886,
          "validation_metric": 0.0064000883139669895,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.034030349999738974
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.05712086718995124,
          "train_loss": 0.04298853874206543,
          "validation_loss": 0.03944144397974014,
          "train_metric": 0.004329648334532976,
          "validation_metric": 0.003972397651523352,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.0542697919991042
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.04124185326509178,
          "train_loss": 0.039410609751939774,
          "validation_loss": 0.030355975031852722,
          "train_metric": 0.003969292622059584,
          "validation_metric": 0.003057342953979969,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.08361588600018877
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.033004145545419306,
          "train_loss": 0.03479968383908272,
          "validation_loss": 0.02701583132147789,
          "train_metric": 0.003504896769300103,
          "validation_metric": 0.0027209355030208826,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.10916537399862136
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.02561670570867136,
          "train_loss": 0.025141824036836624,
          "validation_loss": 0.020702961832284927,
          "train_metric": 0.002532192040234804,
          "validation_metric": 0.002085126703605056,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.128768657999899
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.01748245913768187,
          "train_loss": 0.016836190596222878,
          "validation_loss": 0.013181811198592186,
          "train_metric": 0.0016956792678683996,
          "validation_metric": 0.0013276238460093737,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.1515134119999857
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.014318117406219244,
          "train_loss": 0.046528950333595276,
          "validation_loss": 0.033481065183877945,
          "train_metric": 0.004686225205659866,
          "validation_metric": 0.003372090170159936,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.17413294699872495
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.025819073955062777,
          "train_loss": 0.012293417006731033,
          "validation_loss": 0.009942252188920975,
          "train_metric": 0.001238148077391088,
          "validation_metric": 0.0010013472056016326,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2041552849987056
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.017492826853413135,
          "train_loss": 0.01880115084350109,
          "validation_loss": 0.013885610736906528,
          "train_metric": 0.0018935827538371086,
          "validation_metric": 0.0013985078549012542,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.22059326699854864
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.027304516901494935,
          "train_loss": 0.06755799055099487,
          "validation_loss": 0.06958584487438202,
          "train_metric": 0.006804194301366806,
          "validation_metric": 0.007008431013673544,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.23709135899844114
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.03362681093858555,
          "train_loss": 0.016535228118300438,
          "validation_loss": 0.016072580590844154,
          "train_metric": 0.0016653676284477115,
          "validation_metric": 0.0016187714645639062,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2531466429991269
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.019894773024134338,
          "train_loss": 0.01881503313779831,
          "validation_loss": 0.019192788749933243,
          "train_metric": 0.0018949813675135374,
          "validation_metric": 0.0019330274080857635,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.26972774599926197
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.016378937638364732,
          "train_loss": 0.012517486698925495,
          "validation_loss": 0.011131730861961842,
          "train_metric": 0.0012607155367732048,
          "validation_metric": 0.0011211471864953637,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.2864539619986317
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.011700731120072305,
          "train_loss": 0.015722764655947685,
          "validation_loss": 0.010592243634164333,
          "train_metric": 0.001583539298735559,
          "validation_metric": 0.0010668118484318256,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3031916079999064
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.010635019571054727,
          "train_loss": 0.008123190142214298,
          "validation_loss": 0.006314739119261503,
          "train_metric": 0.0008181380107998848,
          "validation_metric": 0.0006359974504448473,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.32023616900005436
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.012662724140682258,
          "train_loss": 0.010757599025964737,
          "validation_loss": 0.006970019545406103,
          "train_metric": 0.0010834659915417433,
          "validation_metric": 0.0007019948097877204,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3363411479986098
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.02474428154528141,
          "train_loss": 0.01928664743900299,
          "validation_loss": 0.018743446096777916,
          "train_metric": 0.001942480681464076,
          "validation_metric": 0.001887771300971508,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.35282553599972744
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.013873522257199511,
          "train_loss": 0.015882382169365883,
          "validation_loss": 0.012220388278365135,
          "train_metric": 0.001599615439772606,
          "validation_metric": 0.0012307927245274186,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.36924474899933557
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.018884545381297357,
          "train_loss": 0.02976568043231964,
          "validation_loss": 0.030144669115543365,
          "train_metric": 0.0029978903476148844,
          "validation_metric": 0.0030360606033354998,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.3853080989993032
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.026033966714749113,
          "train_loss": 0.018643256276845932,
          "validation_loss": 0.01549321599304676,
          "train_metric": 0.0018776804208755493,
          "validation_metric": 0.0015604199143126607,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.40119257100013783
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.019462034601019695,
          "train_loss": 0.013437370769679546,
          "validation_loss": 0.010714075528085232,
          "train_metric": 0.0013533626915886998,
          "validation_metric": 0.0010790824890136719,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.41781865499979176
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.015429269667947665,
          "train_loss": 0.029222818091511726,
          "validation_loss": 0.020153198391199112,
          "train_metric": 0.002943215426057577,
          "validation_metric": 0.0020297563169151545,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.43437488499876054
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.015224907554511447,
          "train_loss": 0.008092422969639301,
          "validation_loss": 0.007439020089805126,
          "train_metric": 0.0008150392095558345,
          "validation_metric": 0.0007492308504879475,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.45091912599855277
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.011184759699972346,
          "train_loss": 0.009577702730894089,
          "validation_loss": 0.008495678193867207,
          "train_metric": 0.0009646312682889402,
          "validation_metric": 0.0008556536049582064,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4672348059993965
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.009334446061984636,
          "train_loss": 0.013145903125405312,
          "validation_loss": 0.011392178013920784,
          "train_metric": 0.001324007287621498,
          "validation_metric": 0.0011473784688860178,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.4835097759987548
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.008522517862729728,
          "train_loss": 0.014611808583140373,
          "validation_loss": 0.010562802664935589,
          "train_metric": 0.001471647759899497,
          "validation_metric": 0.0010638468666002154,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.49993070299933606
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.008859038454829715,
          "train_loss": 0.00542509276419878,
          "validation_loss": 0.004215569235384464,
          "train_metric": 0.0005463954876177013,
          "validation_metric": 0.00042457671952433884,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5169883250000566
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.014543485260219313,
          "train_loss": 0.036845505237579346,
          "validation_loss": 0.03355100750923157,
          "train_metric": 0.0037109442055225372,
          "validation_metric": 0.003379134926944971,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5337397869989218
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.023210454295622185,
          "train_loss": 0.013243937864899635,
          "validation_loss": 0.014316743239760399,
          "train_metric": 0.0013338810531422496,
          "validation_metric": 0.001441929955035448,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5503107009990345
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.009630044354707934,
          "train_loss": 0.02320416271686554,
          "validation_loss": 0.014523934572935104,
          "train_metric": 0.0023370382841676474,
          "validation_metric": 0.001462797517888248,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5665790139992168
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.024923962133470923,
          "train_loss": 0.01068384014070034,
          "validation_loss": 0.009201528504490852,
          "train_metric": 0.001076037180610001,
          "validation_metric": 0.0009267442510463297,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.5830414679985552
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.015101594268344343,
          "train_loss": 0.01345081813633442,
          "validation_loss": 0.009596533142030239,
          "train_metric": 0.0013547170674428344,
          "validation_metric": 0.000966527673881501,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.59956678799972
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.011989339807769284,
          "train_loss": 0.011377265676856041,
          "validation_loss": 0.008948974311351776,
          "train_metric": 0.0011458764784038067,
          "validation_metric": 0.0009013079106807709,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6161531629986712
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.009089398430660367,
          "train_loss": 0.008982693776488304,
          "validation_loss": 0.0063102287240326405,
          "train_metric": 0.000904704153072089,
          "validation_metric": 0.0006355431396514177,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6325809929985553
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.0085994685505284,
          "train_loss": 0.009032921865582466,
          "validation_loss": 0.008105510845780373,
          "train_metric": 0.0009097626898437738,
          "validation_metric": 0.0008163573802448809,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6501641839986405
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.017655255607678555,
          "train_loss": 0.012858407571911812,
          "validation_loss": 0.010684921406209469,
          "train_metric": 0.0012950518867000937,
          "validation_metric": 0.0010761460289359093,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6665193789995101
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.030525729845976457,
          "train_loss": 0.03246269375085831,
          "validation_loss": 0.0400688610970974,
          "train_metric": 0.0032695233821868896,
          "validation_metric": 0.0040355888195335865,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.6822656309996091
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.018862196244299412,
          "train_loss": 0.012744036503136158,
          "validation_loss": 0.008532511070370674,
          "train_metric": 0.0012835327070206404,
          "validation_metric": 0.0008593632956035435,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.698352598999918
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.015243953181197867,
          "train_loss": 0.009687630459666252,
          "validation_loss": 0.008076634258031845,
          "train_metric": 0.0009757025400176644,
          "validation_metric": 0.0008134489762596786,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7150710349997098
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.00911190127953887,
          "train_loss": 0.005455123260617256,
          "validation_loss": 0.0039056253153830767,
          "train_metric": 0.0005494199576787651,
          "validation_metric": 0.0003933603293262422,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7327404440002283
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.012805219739675522,
          "train_loss": 0.009123409166932106,
          "validation_loss": 0.006591612007468939,
          "train_metric": 0.0009188763797283173,
          "validation_metric": 0.0006638829945586622,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7491121209986886
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009364904239191674,
          "train_loss": 0.006235888693481684,
          "validation_loss": 0.005914786830544472,
          "train_metric": 0.0006280558300204575,
          "validation_metric": 0.000595715653616935,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7655078169991612
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.012263939235708676,
          "train_loss": 0.014504166319966316,
          "validation_loss": 0.011388527229428291,
          "train_metric": 0.0014608065830543637,
          "validation_metric": 0.0011470107128843665,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7822953949998919
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.012308380930335261,
          "train_loss": 0.0076313382014632225,
          "validation_loss": 0.004995668772608042,
          "train_metric": 0.0007686004973948002,
          "validation_metric": 0.0005031453911215067,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.7992409269991185
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.00741905091854278,
          "train_loss": 0.0047460878267884254,
          "validation_loss": 0.004004006274044514,
          "train_metric": 0.00047800858737900853,
          "validation_metric": 0.00040326890302821994,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8159778949993779
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.006276226238696836,
          "train_loss": 0.007447235286235809,
          "validation_loss": 0.005911177024245262,
          "train_metric": 0.0007500583888031542,
          "validation_metric": 0.0005953520303592086,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8328100660000928
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0064359061943832785,
          "train_loss": 0.00587816908955574,
          "validation_loss": 0.004326930269598961,
          "train_metric": 0.0005920277326367795,
          "validation_metric": 0.0004357926081866026,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8491713539988268
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.006368354261212517,
          "train_loss": 0.023492200300097466,
          "validation_loss": 0.017734156921505928,
          "train_metric": 0.002366048516705632,
          "validation_metric": 0.0017861193045973778,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8652760489985667
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.019153180968714878,
          "train_loss": 0.014953430742025375,
          "validation_loss": 0.014811434783041477,
          "train_metric": 0.001506054773926735,
          "validation_metric": 0.0014917533844709396,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.881714662999002
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009679946291726083,
          "train_loss": 0.006901535205543041,
          "validation_loss": 0.00687922490760684,
          "train_metric": 0.0006950973765924573,
          "validation_metric": 0.0006928503280505538,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.8992955289995734
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.007933065528050065,
          "train_loss": 0.003397311782464385,
          "validation_loss": 0.0031686872243881226,
          "train_metric": 0.00034216479980386794,
          "validation_metric": 0.00031913857674226165,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9168470509994222
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.006384828622685745,
          "train_loss": 0.004297993611544371,
          "validation_loss": 0.0034600994549691677,
          "train_metric": 0.0004328781506046653,
          "validation_metric": 0.0003484885673969984,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9337609369995334
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.011624256323557347,
          "train_loss": 0.010057921521365643,
          "validation_loss": 0.00859124306589365,
          "train_metric": 0.0010129970032721758,
          "validation_metric": 0.0008652785909362137,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9501886170000944
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.014000004041008651,
          "train_loss": 0.016081657260656357,
          "validation_loss": 0.016746465116739273,
          "train_metric": 0.0016196856740862131,
          "validation_metric": 0.0016866425285115838,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9670504539990361
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.009316337265772745,
          "train_loss": 0.009057899005711079,
          "validation_loss": 0.006404774263501167,
          "train_metric": 0.0009122784831561148,
          "validation_metric": 0.0006450655055232346,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9836392789984529
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.011993870371952653,
          "train_loss": 0.005980128888040781,
          "validation_loss": 0.004284940659999847,
          "train_metric": 0.0006022967281751335,
          "validation_metric": 0.0004315635596867651,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 0.9998718819988426
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.006850437413959298,
          "train_loss": 0.005441918503493071,
          "validation_loss": 0.003961381036788225,
          "train_metric": 0.0005480900872498751,
          "validation_metric": 0.0003989757678937167,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.0166520759994455
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.015075413655722514,
          "train_loss": 0.028963297605514526,
          "validation_loss": 0.023454945534467697,
          "train_metric": 0.0029170773923397064,
          "validation_metric": 0.00236229645088315,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.0332357839997712
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.02258795697707683,
          "train_loss": 0.02257859706878662,
          "validation_loss": 0.01701636053621769,
          "train_metric": 0.0022740333806723356,
          "validation_metric": 0.0017138253897428513,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.0498805999995966
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.01488973677624017,
          "train_loss": 0.007904130034148693,
          "validation_loss": 0.007655960973352194,
          "train_metric": 0.0007960750954225659,
          "validation_metric": 0.0007710803765803576,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.0658969169999182
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.010819402465131134,
          "train_loss": 0.008546552620828152,
          "validation_loss": 0.006857032421976328,
          "train_metric": 0.000860777567140758,
          "validation_metric": 0.0006906152120791376,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.0822664150000492
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006689575107884593,
          "train_loss": 0.004615181125700474,
          "validation_loss": 0.004366105422377586,
          "train_metric": 0.0004648241738323122,
          "validation_metric": 0.0004397381271701306,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.098546148999958
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0056382764923910145,
          "train_loss": 0.006820966489613056,
          "validation_loss": 0.0049172635190188885,
          "train_metric": 0.0006869828212074935,
          "validation_metric": 0.0004952488234266639,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.1148197779984912
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.010287283308571205,
          "train_loss": 0.005770284682512283,
          "validation_loss": 0.004310694057494402,
          "train_metric": 0.0005811619339510798,
          "validation_metric": 0.00043415723484940827,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.1320158269991225
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.007281918355147354,
          "train_loss": 0.004079360514879227,
          "validation_loss": 0.0032922043465077877,
          "train_metric": 0.00041085825068876147,
          "validation_metric": 0.0003315787762403488,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.1485051999989082
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.013370295717322733,
          "train_loss": 0.016408666968345642,
          "validation_loss": 0.013091842643916607,
          "train_metric": 0.001652620849199593,
          "validation_metric": 0.0013185623101890087,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.1653187929987325
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.012147121655289084,
          "train_loss": 0.009889648295938969,
          "validation_loss": 0.006711404304951429,
          "train_metric": 0.0009960491443052888,
          "validation_metric": 0.0006759479292668402,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.1816580209997483
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.008532549312803894,
          "train_loss": 0.012201963923871517,
          "validation_loss": 0.01128074899315834,
          "train_metric": 0.001228937180712819,
          "validation_metric": 0.0011361557990312576,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.1981600669987529
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.009187651146930875,
          "train_loss": 0.004321001004427671,
          "validation_loss": 0.0037306207232177258,
          "train_metric": 0.0004351954266894609,
          "validation_metric": 0.0003757344966288656,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.2148462810000638
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0052229498396627605,
          "train_loss": 0.005459694191813469,
          "validation_loss": 0.003148800227791071,
          "train_metric": 0.0005498803802765906,
          "validation_metric": 0.0003171356220263988,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.2327016279996315
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00755000038043363,
          "train_loss": 0.003514666110277176,
          "validation_loss": 0.003109073732048273,
          "train_metric": 0.00035398430190980434,
          "validation_metric": 0.00031313448562286794,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.2500113679998321
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0072291200631298125,
          "train_loss": 0.008164131082594395,
          "validation_loss": 0.008091752417385578,
          "train_metric": 0.0008222613832913339,
          "validation_metric": 0.0008149717468768358,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.2663398819986469
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005638060065393802,
          "train_loss": 0.015293244272470474,
          "validation_loss": 0.013811329379677773,
          "train_metric": 0.0015402794815599918,
          "validation_metric": 0.0013910264242440462,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.2822237330001371
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.012410079885739833,
          "train_loss": 0.005831478163599968,
          "validation_loss": 0.005804590880870819,
          "train_metric": 0.0005873250192962587,
          "validation_metric": 0.0005846170824952424,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.2982746249999764
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.007770354073727503,
          "train_loss": 0.0071770204231143,
          "validation_loss": 0.006104815751314163,
          "train_metric": 0.0007228432223200798,
          "validation_metric": 0.0006148546817712486,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.3143625049997354
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.011507475763210095,
          "train_loss": 0.007719400804489851,
          "validation_loss": 0.005554581061005592,
          "train_metric": 0.000777469773311168,
          "validation_metric": 0.0005594370886683464,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.3306344789998548
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.00520728959236294,
          "train_loss": 0.004582471679896116,
          "validation_loss": 0.004657945595681667,
          "train_metric": 0.00046152982395142317,
          "validation_metric": 0.00046913124970160425,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.3471746300001541
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0039461106425733306,
          "train_loss": 0.003331251908093691,
          "validation_loss": 0.002533932449296117,
          "train_metric": 0.00033551143133081496,
          "validation_metric": 0.00025520840426906943,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.36443202099872
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.006047001326805912,
          "train_loss": 0.003861012402921915,
          "validation_loss": 0.0033782129175961018,
          "train_metric": 0.000388867047149688,
          "validation_metric": 0.0003402412694413215,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.3811198069997772
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0038607403039350174,
          "train_loss": 0.004387541674077511,
          "validation_loss": 0.0039794547483325005,
          "train_metric": 0.00044189716572873294,
          "validation_metric": 0.00040079609607346356,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.398251339998751
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.004763638098665979,
          "train_loss": 0.0026209307834506035,
          "validation_loss": 0.0021505323238670826,
          "train_metric": 0.0002639705780893564,
          "validation_metric": 0.00021659376216121018,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.4158650949993898
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.003824164734396618,
          "train_loss": 0.004753844812512398,
          "validation_loss": 0.0040780967101454735,
          "train_metric": 0.0004787898506037891,
          "validation_metric": 0.0004107309796381742,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.4323311779990036
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.008297614811453968,
          "train_loss": 0.017588187009096146,
          "validation_loss": 0.014348255470395088,
          "train_metric": 0.001771418028511107,
          "validation_metric": 0.001445103669539094,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.454520273999151
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.022165453498018906,
          "train_loss": 0.011651214212179184,
          "validation_loss": 0.009735824540257454,
          "train_metric": 0.0011734676081687212,
          "validation_metric": 0.0009805565932765603,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.4820540309992793
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.015304860688047484,
          "train_loss": 0.008839533664286137,
          "validation_loss": 0.00849987007677555,
          "train_metric": 0.0008902854169718921,
          "validation_metric": 0.0008560757851228118,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.5051900609996665
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.008501485906890593,
          "train_loss": 0.007540834601968527,
          "validation_loss": 0.008740246295928955,
          "train_metric": 0.000759485294111073,
          "validation_metric": 0.000880285631865263,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.5216895869998552
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0056786502682371065,
          "train_loss": 0.00600406713783741,
          "validation_loss": 0.0043467930518090725,
          "train_metric": 0.0006047076312825084,
          "validation_metric": 0.00043779308907687664,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.538076580998677
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.00819231189234415,
          "train_loss": 0.012072235345840454,
          "validation_loss": 0.00812184065580368,
          "train_metric": 0.001215871307067573,
          "validation_metric": 0.0008180021541193128,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.5544717799984937
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.00904718523088377,
          "train_loss": 0.006723250728100538,
          "validation_loss": 0.005852183327078819,
          "train_metric": 0.0006771410699002445,
          "validation_metric": 0.0005894104251638055,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.570886765999603
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.005698387882148381,
          "train_loss": 0.004945941269397736,
          "validation_loss": 0.004040501546114683,
          "train_metric": 0.0004981370875611901,
          "validation_metric": 0.00040694448398426175,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.5877465229987138
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0067597094566735905,
          "train_loss": 0.023856552317738533,
          "validation_loss": 0.01870439387857914,
          "train_metric": 0.0024027444887906313,
          "validation_metric": 0.0018838379764929414,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.6039476879996073
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.009127945988439023,
          "train_loss": 0.003495903452858329,
          "validation_loss": 0.0034935101866722107,
          "train_metric": 0.0003520945901982486,
          "validation_metric": 0.00035185349406674504,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.6202898739993543
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004580363747663796,
          "train_loss": 0.004402665887027979,
          "validation_loss": 0.003863080870360136,
          "train_metric": 0.000443420372903347,
          "validation_metric": 0.00038907540147192776,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.636553300999367
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.003433535843214486,
          "train_loss": 0.002209350001066923,
          "validation_loss": 0.0015665796818211675,
          "train_metric": 0.00022251764312386513,
          "validation_metric": 0.00015778015949763358,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.6539906699999847
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.002979647382744588,
          "train_loss": 0.002050379291176796,
          "validation_loss": 0.0013632948976010084,
          "train_metric": 0.00020650666556321084,
          "validation_metric": 0.0001373060658806935,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.6711987210001098
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.006463776298915036,
          "train_loss": 0.018988341093063354,
          "validation_loss": 0.013512196950614452,
          "train_metric": 0.0019124364480376244,
          "validation_metric": 0.001360899186693132,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.6884519059985905
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.029347938136197627,
          "train_loss": 0.023144932463765144,
          "validation_loss": 0.02203613519668579,
          "train_metric": 0.002331072697415948,
          "validation_metric": 0.002219398971647024,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.7052809910001088
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.01818702364107594,
          "train_loss": 0.012489689514040947,
          "validation_loss": 0.013489692471921444,
          "train_metric": 0.001257915748283267,
          "validation_metric": 0.0013586323475465178,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.7215031479991012
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.007617757772095501,
          "train_loss": 0.010939270257949829,
          "validation_loss": 0.004890871699899435,
          "train_metric": 0.0011017632205039263,
          "validation_metric": 0.0004925906541757286,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.7380094499985717
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.006965954467887059,
          "train_loss": 0.004485698416829109,
          "validation_loss": 0.004439000505954027,
          "train_metric": 0.00045178309665061533,
          "validation_metric": 0.0004470798885449767,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.7542819329992199
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.004539015601039864,
          "train_loss": 0.004262355621904135,
          "validation_loss": 0.0031665386632084846,
          "train_metric": 0.0004292888334020972,
          "validation_metric": 0.00031892218976281583,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.7708768299999065
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.00725810659059789,
          "train_loss": 0.008403542451560497,
          "validation_loss": 0.007333074230700731,
          "train_metric": 0.0008463739650323987,
          "validation_metric": 0.0007385602802969515,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.7873572600001353
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.012389174953568727,
          "train_loss": 0.010053977370262146,
          "validation_loss": 0.008897265419363976,
          "train_metric": 0.0010125996777787805,
          "validation_metric": 0.0008961000712588429,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.8038659689991619
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.005710465455194935,
          "train_loss": 0.007609277497977018,
          "validation_loss": 0.004346129950135946,
          "train_metric": 0.0007663784781470895,
          "validation_metric": 0.00043772635399363935,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.8203546929998993
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0034564402085379697,
          "train_loss": 0.002612997544929385,
          "validation_loss": 0.0024581998586654663,
          "train_metric": 0.0002631715324241668,
          "validation_metric": 0.00024758087238296866,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.8374541140001384
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.002447650349495234,
          "train_loss": 0.010106563568115234,
          "validation_loss": 0.007336240261793137,
          "train_metric": 0.0010178961092606187,
          "validation_metric": 0.0007388792582787573,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.95354375199895
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.010072535849758424,
          "train_loss": 0.0037069644313305616,
          "validation_loss": 0.0032028802670538425,
          "train_metric": 0.0003733519115485251,
          "validation_metric": 0.0003225824038963765,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.971044277999681
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.00365701340342639,
          "train_loss": 0.003152288030833006,
          "validation_loss": 0.003359987400472164,
          "train_metric": 0.00031748690526001155,
          "validation_metric": 0.00033840560354292393,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 1.9877183889984735
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0037000329248257913,
          "train_loss": 0.0023708611261099577,
          "validation_loss": 0.0024130165111273527,
          "train_metric": 0.00023878445790614933,
          "validation_metric": 0.00024303018290083855,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.003909549999662
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.0026984021787939128,
          "train_loss": 0.0013689809711650014,
          "validation_loss": 0.0012193751754239202,
          "train_metric": 0.0001378787710564211,
          "validation_metric": 0.00012281100498512387,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.021262737998768
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.0035370289733691607,
          "train_loss": 0.010577745735645294,
          "validation_loss": 0.007510717958211899,
          "train_metric": 0.0010653516510501504,
          "validation_metric": 0.0007564519764855504,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.0376392289999785
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004780426577781327,
          "train_loss": 0.003921136260032654,
          "validation_loss": 0.00391165167093277,
          "train_metric": 0.0003949224774260074,
          "validation_metric": 0.0003939672315027565,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.0538339979993907
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.014111546050116885,
          "train_loss": 0.01995665393769741,
          "validation_loss": 0.014220423996448517,
          "train_metric": 0.0020099610555917025,
          "validation_metric": 0.001432228833436966,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.070627576998959
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.020882750512100756,
          "train_loss": 0.00943874754011631,
          "validation_loss": 0.009541790932416916,
          "train_metric": 0.0009506359929218888,
          "validation_metric": 0.0009610141860321164,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.0870752609989722
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.008654406439745799,
          "train_loss": 0.005306611768901348,
          "validation_loss": 0.005309447646141052,
          "train_metric": 0.0005344625096768141,
          "validation_metric": 0.0005347480764612556,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.1037387599990325
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.010432208349811845,
          "train_loss": 0.013176637701690197,
          "validation_loss": 0.010811160318553448,
          "train_metric": 0.0013271024217829108,
          "validation_metric": 0.0010888606775552034,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.1203620950000186
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.006840286900114734,
          "train_loss": 0.0055000558495521545,
          "validation_loss": 0.0036524024326354265,
          "train_metric": 0.0005539454286918044,
          "validation_metric": 0.000367856613593176,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.1377876119986468
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.004736230272101238,
          "train_loss": 0.005846857093274593,
          "validation_loss": 0.0026240688748657703,
          "train_metric": 0.0005888740415684879,
          "validation_metric": 0.00026428664568811655,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.1545419649992255
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.004318408842664212,
          "train_loss": 0.0024140786845237017,
          "validation_loss": 0.002599693601951003,
          "train_metric": 0.00024313718313351274,
          "validation_metric": 0.00026183162117376924,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.1712981889995717
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004896488171652891,
          "train_loss": 0.004789077211171389,
          "validation_loss": 0.0030997914727777243,
          "train_metric": 0.00048233833513222635,
          "validation_metric": 0.00031219961238093674,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.1869352929988963
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.009757885469298344,
          "train_loss": 0.010185400024056435,
          "validation_loss": 0.010342014953494072,
          "train_metric": 0.001025836099870503,
          "validation_metric": 0.0010416097939014435,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.2087629919988103
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.007231905041408027,
          "train_loss": 0.003225689521059394,
          "validation_loss": 0.0034216935746371746,
          "train_metric": 0.00032487959833815694,
          "validation_metric": 0.00034462043549865484,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.2278399579990946
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.004753537170472555,
          "train_loss": 0.002640755847096443,
          "validation_loss": 0.0023004368413239717,
          "train_metric": 0.00026596724637784064,
          "validation_metric": 0.00023169159248936921,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.2462366279996786
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0033418985549360514,
          "train_loss": 0.004325168672949076,
          "validation_loss": 0.002443412085995078,
          "train_metric": 0.000435615103924647,
          "validation_metric": 0.00024609154206700623,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.2639802749999944
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.003738133564183954,
          "train_loss": 0.0031395144760608673,
          "validation_loss": 0.0028043892234563828,
          "train_metric": 0.00031620042864233255,
          "validation_metric": 0.0002824478142429143,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.2815482579990203
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.008549317331926432,
          "train_loss": 0.004406583029776812,
          "validation_loss": 0.0036138477735221386,
          "train_metric": 0.0004438149044290185,
          "validation_metric": 0.00036397346411831677,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.299729292999473
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.006362757434544619,
          "train_loss": 0.009906893596053123,
          "validation_loss": 0.006951620802283287,
          "train_metric": 0.0009977860609069467,
          "validation_metric": 0.0007001417106948793,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.317290854998646
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.006150349086965434,
          "train_loss": 0.0076751867309212685,
          "validation_loss": 0.00497052725404501,
          "train_metric": 0.0007730167708359659,
          "validation_metric": 0.0005006132996641099,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.33497652799997
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.009499962165136822,
          "train_loss": 0.019416527822613716,
          "validation_loss": 0.017161814495921135,
          "train_metric": 0.0019555615726858377,
          "validation_metric": 0.0017284750938415527,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.35276548299953
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.01620232974528335,
          "train_loss": 0.010695550590753555,
          "validation_loss": 0.009720243513584137,
          "train_metric": 0.0010772165842354298,
          "validation_metric": 0.0009789873147383332,
          "lr_first": 0.03,
          "lr_last": 0.03,
          "seconds": 2.3703288799988513
        }
      ],
      "best": {
        "epoch": 111,
        "updates": 1776
      },
      "best_validation_metric": 0.00012281100498512387,
      "config": {
        "label": "LR 0.03",
        "batch": 64,
        "lr": 0.03,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.011082012677538493
    },
    "50c1636b4f2c_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 4,
          "samples_seen": 1024,
          "online_loss": 0.9806545972824097,
          "train_loss": 0.8940960764884949,
          "validation_loss": 0.6446229219436646,
          "train_metric": 0.09005008637905121,
          "validation_metric": 0.06492406129837036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.014324513000246952
        },
        {
          "epoch": 2,
          "updates": 8,
          "samples_seen": 2048,
          "online_loss": 0.8336821049451828,
          "train_loss": 0.717647910118103,
          "validation_loss": 0.528424859046936,
          "train_metric": 0.07227887213230133,
          "validation_metric": 0.053221020847558975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.021566891000475152
        },
        {
          "epoch": 3,
          "updates": 12,
          "samples_seen": 3072,
          "online_loss": 0.6288904398679733,
          "train_loss": 0.47919154167175293,
          "validation_loss": 0.3689866065979004,
          "train_metric": 0.04826241359114647,
          "validation_metric": 0.03716298192739487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.028617373000088264
        },
        {
          "epoch": 4,
          "updates": 16,
          "samples_seen": 4096,
          "online_loss": 0.3902045488357544,
          "train_loss": 0.2533669173717499,
          "validation_loss": 0.21113337576389313,
          "train_metric": 0.025518184527754784,
          "validation_metric": 0.021264581009745598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12574619200131565
        },
        {
          "epoch": 5,
          "updates": 20,
          "samples_seen": 5120,
          "online_loss": 0.21064967289566994,
          "train_loss": 0.18689057230949402,
          "validation_loss": 0.1506071239709854,
          "train_metric": 0.018822932615876198,
          "validation_metric": 0.015168598853051662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13329207300012058
        },
        {
          "epoch": 6,
          "updates": 24,
          "samples_seen": 6144,
          "online_loss": 0.1924927979707718,
          "train_loss": 0.1908048391342163,
          "validation_loss": 0.15706855058670044,
          "train_metric": 0.01921716518700123,
          "validation_metric": 0.01581936702132225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22285754699987592
        },
        {
          "epoch": 7,
          "updates": 28,
          "samples_seen": 7168,
          "online_loss": 0.1753976009786129,
          "train_loss": 0.14928551018238068,
          "validation_loss": 0.1295471042394638,
          "train_metric": 0.01503549050539732,
          "validation_metric": 0.013047510758042336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23013691500091227
        },
        {
          "epoch": 8,
          "updates": 32,
          "samples_seen": 8192,
          "online_loss": 0.13980857841670513,
          "train_loss": 0.12757600843906403,
          "validation_loss": 0.10588360577821732,
          "train_metric": 0.012848987244069576,
          "validation_metric": 0.010664209723472595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23721383600059198
        },
        {
          "epoch": 9,
          "updates": 36,
          "samples_seen": 9216,
          "online_loss": 0.12374548055231571,
          "train_loss": 0.11841696500778198,
          "validation_loss": 0.09804915636777878,
          "train_metric": 0.01192652340978384,
          "validation_metric": 0.009875152260065079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2439303959999961
        },
        {
          "epoch": 10,
          "updates": 40,
          "samples_seen": 10240,
          "online_loss": 0.11517713777720928,
          "train_loss": 0.11091237515211105,
          "validation_loss": 0.09270395338535309,
          "train_metric": 0.011170687153935432,
          "validation_metric": 0.00933680310845375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2511093049997726
        },
        {
          "epoch": 11,
          "updates": 44,
          "samples_seen": 11264,
          "online_loss": 0.10716254077851772,
          "train_loss": 0.09891185164451599,
          "validation_loss": 0.0820271372795105,
          "train_metric": 0.009962039068341255,
          "validation_metric": 0.008261472918093204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25770345000091766
        },
        {
          "epoch": 12,
          "updates": 48,
          "samples_seen": 12288,
          "online_loss": 0.09509728290140629,
          "train_loss": 0.09046357125043869,
          "validation_loss": 0.07596665620803833,
          "train_metric": 0.009111158549785614,
          "validation_metric": 0.007651083637028933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26489149000008183
        },
        {
          "epoch": 13,
          "updates": 52,
          "samples_seen": 13312,
          "online_loss": 0.08982400223612785,
          "train_loss": 0.08778328448534012,
          "validation_loss": 0.07448093593120575,
          "train_metric": 0.008841210044920444,
          "validation_metric": 0.007501447573304176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27145297000060964
        },
        {
          "epoch": 14,
          "updates": 56,
          "samples_seen": 14336,
          "online_loss": 0.08556756656616926,
          "train_loss": 0.08043214678764343,
          "validation_loss": 0.0668860375881195,
          "train_metric": 0.008100830018520355,
          "validation_metric": 0.006736516486853361,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2786581580003258
        },
        {
          "epoch": 15,
          "updates": 60,
          "samples_seen": 15360,
          "online_loss": 0.07813991792500019,
          "train_loss": 0.07464566081762314,
          "validation_loss": 0.06075477600097656,
          "train_metric": 0.007518037687987089,
          "validation_metric": 0.006118998862802982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28538072999981523
        },
        {
          "epoch": 16,
          "updates": 64,
          "samples_seen": 16384,
          "online_loss": 0.07296233251690865,
          "train_loss": 0.07022445648908615,
          "validation_loss": 0.05669572576880455,
          "train_metric": 0.007072750478982925,
          "validation_metric": 0.005710185971111059,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29201770100007707
        },
        {
          "epoch": 17,
          "updates": 68,
          "samples_seen": 17408,
          "online_loss": 0.06864664237946272,
          "train_loss": 0.06612277776002884,
          "validation_loss": 0.05306602269411087,
          "train_metric": 0.006659644190222025,
          "validation_metric": 0.00534461485221982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2986674050007423
        },
        {
          "epoch": 18,
          "updates": 72,
          "samples_seen": 18432,
          "online_loss": 0.06468546204268932,
          "train_loss": 0.0618831105530262,
          "validation_loss": 0.04991196095943451,
          "train_metric": 0.006232640706002712,
          "validation_metric": 0.0050269500352442265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3054370370009565
        },
        {
          "epoch": 19,
          "updates": 76,
          "samples_seen": 19456,
          "online_loss": 0.06100610550493002,
          "train_loss": 0.05926908552646637,
          "validation_loss": 0.04818829521536827,
          "train_metric": 0.005969365127384663,
          "validation_metric": 0.004853349179029465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31245945900082006
        },
        {
          "epoch": 20,
          "updates": 80,
          "samples_seen": 20480,
          "online_loss": 0.05827333591878414,
          "train_loss": 0.056519921869039536,
          "validation_loss": 0.04585028439760208,
          "train_metric": 0.00569247966632247,
          "validation_metric": 0.0046178726479411125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3193079930006206
        },
        {
          "epoch": 21,
          "updates": 84,
          "samples_seen": 21504,
          "online_loss": 0.05574245471507311,
          "train_loss": 0.054354675114154816,
          "validation_loss": 0.04393423721194267,
          "train_metric": 0.005474403966218233,
          "validation_metric": 0.0044248951599001884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3261774950005929
        },
        {
          "epoch": 22,
          "updates": 88,
          "samples_seen": 22528,
          "online_loss": 0.054028868675231934,
          "train_loss": 0.05264473333954811,
          "validation_loss": 0.042576633393764496,
          "train_metric": 0.005302185192704201,
          "validation_metric": 0.004288163501769304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33312315700095496
        },
        {
          "epoch": 23,
          "updates": 92,
          "samples_seen": 23552,
          "online_loss": 0.05223668646067381,
          "train_loss": 0.05124516040086746,
          "validation_loss": 0.04186812788248062,
          "train_metric": 0.005161224398761988,
          "validation_metric": 0.004216804634779692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33979915700001584
        },
        {
          "epoch": 24,
          "updates": 96,
          "samples_seen": 24576,
          "online_loss": 0.050707727670669556,
          "train_loss": 0.04981391131877899,
          "validation_loss": 0.040598198771476746,
          "train_metric": 0.0050170752219855785,
          "validation_metric": 0.004088901914656162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3466796169996087
        },
        {
          "epoch": 25,
          "updates": 100,
          "samples_seen": 25600,
          "online_loss": 0.04937855061143637,
          "train_loss": 0.04844607412815094,
          "validation_loss": 0.03981998935341835,
          "train_metric": 0.004879311192780733,
          "validation_metric": 0.004010523669421673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3535724450011912
        },
        {
          "epoch": 26,
          "updates": 104,
          "samples_seen": 26624,
          "online_loss": 0.04826956335455179,
          "train_loss": 0.0471600703895092,
          "validation_loss": 0.03878773748874664,
          "train_metric": 0.004749789368361235,
          "validation_metric": 0.003906558733433485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36038468700098747
        },
        {
          "epoch": 27,
          "updates": 108,
          "samples_seen": 27648,
          "online_loss": 0.04694915656000376,
          "train_loss": 0.04601278156042099,
          "validation_loss": 0.03812749683856964,
          "train_metric": 0.00463423877954483,
          "validation_metric": 0.00384006230160594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36728623300041363
        },
        {
          "epoch": 28,
          "updates": 112,
          "samples_seen": 28672,
          "online_loss": 0.04534175340086222,
          "train_loss": 0.04467248171567917,
          "validation_loss": 0.03685891628265381,
          "train_metric": 0.00449924822896719,
          "validation_metric": 0.0037122953217476606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3739797500002169
        },
        {
          "epoch": 29,
          "updates": 116,
          "samples_seen": 29696,
          "online_loss": 0.044646983966231346,
          "train_loss": 0.04326159507036209,
          "validation_loss": 0.03608102351427078,
          "train_metric": 0.004357149824500084,
          "validation_metric": 0.0036339485086500645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3810956380002608
        },
        {
          "epoch": 30,
          "updates": 120,
          "samples_seen": 30720,
          "online_loss": 0.043047722429037094,
          "train_loss": 0.04181276634335518,
          "validation_loss": 0.03477189317345619,
          "train_metric": 0.004211228806525469,
          "validation_metric": 0.00350209791213274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38790808799967635
        },
        {
          "epoch": 31,
          "updates": 124,
          "samples_seen": 31744,
          "online_loss": 0.04139538388699293,
          "train_loss": 0.040581997483968735,
          "validation_loss": 0.03395911306142807,
          "train_metric": 0.004087269771844149,
          "validation_metric": 0.0034202374517917633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3947132100001909
        },
        {
          "epoch": 32,
          "updates": 128,
          "samples_seen": 32768,
          "online_loss": 0.04034789092838764,
          "train_loss": 0.03924839571118355,
          "validation_loss": 0.032727036625146866,
          "train_metric": 0.003952954430133104,
          "validation_metric": 0.003296147333458066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4013476180007274
        },
        {
          "epoch": 33,
          "updates": 132,
          "samples_seen": 33792,
          "online_loss": 0.038975599221885204,
          "train_loss": 0.03782050311565399,
          "validation_loss": 0.0319119468331337,
          "train_metric": 0.0038091428577899933,
          "validation_metric": 0.0032140545081347227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4080596109997714
        },
        {
          "epoch": 34,
          "updates": 136,
          "samples_seen": 34816,
          "online_loss": 0.037856051698327065,
          "train_loss": 0.03675943613052368,
          "validation_loss": 0.031063836067914963,
          "train_metric": 0.0037022761534899473,
          "validation_metric": 0.003128635697066784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41520753400072863
        },
        {
          "epoch": 35,
          "updates": 140,
          "samples_seen": 35840,
          "online_loss": 0.035886808298528194,
          "train_loss": 0.036077238619327545,
          "validation_loss": 0.029999054968357086,
          "train_metric": 0.0036335671320557594,
          "validation_metric": 0.003021394833922386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42215862600096443
        },
        {
          "epoch": 36,
          "updates": 144,
          "samples_seen": 36864,
          "online_loss": 0.03577950783073902,
          "train_loss": 0.03417264297604561,
          "validation_loss": 0.02877442166209221,
          "train_metric": 0.0034417437855154276,
          "validation_metric": 0.0028980544302612543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42935702900103934
        },
        {
          "epoch": 37,
          "updates": 148,
          "samples_seen": 37888,
          "online_loss": 0.034034643322229385,
          "train_loss": 0.03313016518950462,
          "validation_loss": 0.02804451808333397,
          "train_metric": 0.003336748806759715,
          "validation_metric": 0.0028245411813259125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4364421760001278
        },
        {
          "epoch": 38,
          "updates": 152,
          "samples_seen": 38912,
          "online_loss": 0.032888269517570734,
          "train_loss": 0.032102081924676895,
          "validation_loss": 0.027109403163194656,
          "train_metric": 0.0032332041300833225,
          "validation_metric": 0.0027303597889840603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4432102680002572
        },
        {
          "epoch": 39,
          "updates": 156,
          "samples_seen": 39936,
          "online_loss": 0.03240619506686926,
          "train_loss": 0.031201623380184174,
          "validation_loss": 0.026284147053956985,
          "train_metric": 0.003142513334751129,
          "validation_metric": 0.002647243207320571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4504085410007974
        },
        {
          "epoch": 40,
          "updates": 160,
          "samples_seen": 40960,
          "online_loss": 0.031121395528316498,
          "train_loss": 0.030596835538744926,
          "validation_loss": 0.025827661156654358,
          "train_metric": 0.003081601345911622,
          "validation_metric": 0.0026012673042714596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45699151100052404
        },
        {
          "epoch": 41,
          "updates": 164,
          "samples_seen": 41984,
          "online_loss": 0.03018825128674507,
          "train_loss": 0.029666099697351456,
          "validation_loss": 0.024799330160021782,
          "train_metric": 0.0029878609348088503,
          "validation_metric": 0.002497697714716196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4637258530001418
        },
        {
          "epoch": 42,
          "updates": 168,
          "samples_seen": 43008,
          "online_loss": 0.029906265437602997,
          "train_loss": 0.02872026152908802,
          "validation_loss": 0.024129431694746017,
          "train_metric": 0.002892599208280444,
          "validation_metric": 0.0024302280507981777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47041283399994427
        },
        {
          "epoch": 43,
          "updates": 172,
          "samples_seen": 44032,
          "online_loss": 0.029051863588392735,
          "train_loss": 0.028523985296487808,
          "validation_loss": 0.023824302479624748,
          "train_metric": 0.0028728314209729433,
          "validation_metric": 0.00239949650131166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4771676350010239
        },
        {
          "epoch": 44,
          "updates": 176,
          "samples_seen": 45056,
          "online_loss": 0.02953867008909583,
          "train_loss": 0.027273638173937798,
          "validation_loss": 0.022809695452451706,
          "train_metric": 0.002746900776401162,
          "validation_metric": 0.002297308761626482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4939173729999311
        },
        {
          "epoch": 45,
          "updates": 180,
          "samples_seen": 46080,
          "online_loss": 0.028094878885895014,
          "train_loss": 0.026649948209524155,
          "validation_loss": 0.022346770390868187,
          "train_metric": 0.002684085164219141,
          "validation_metric": 0.0022506851237267256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5021874270005355
        },
        {
          "epoch": 46,
          "updates": 184,
          "samples_seen": 47104,
          "online_loss": 0.026579380501061678,
          "train_loss": 0.026354903355240822,
          "validation_loss": 0.02210228331387043,
          "train_metric": 0.0026543696876615286,
          "validation_metric": 0.0022260609548538923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5089728130005824
        },
        {
          "epoch": 47,
          "updates": 188,
          "samples_seen": 48128,
          "online_loss": 0.02635865332558751,
          "train_loss": 0.026332514360547066,
          "validation_loss": 0.02186340093612671,
          "train_metric": 0.002652114490047097,
          "validation_metric": 0.0022020016331225634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.516316289000315
        },
        {
          "epoch": 48,
          "updates": 192,
          "samples_seen": 49152,
          "online_loss": 0.02577700885012746,
          "train_loss": 0.025140123441815376,
          "validation_loss": 0.020984437316656113,
          "train_metric": 0.002532021375373006,
          "validation_metric": 0.00211347546428442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5230852420008887
        },
        {
          "epoch": 49,
          "updates": 196,
          "samples_seen": 50176,
          "online_loss": 0.02585179079324007,
          "train_loss": 0.024223214015364647,
          "validation_loss": 0.020324433222413063,
          "train_metric": 0.002439673524349928,
          "validation_metric": 0.002047002548351884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5350591220012575
        },
        {
          "epoch": 50,
          "updates": 200,
          "samples_seen": 51200,
          "online_loss": 0.02442727144807577,
          "train_loss": 0.023835426196455956,
          "validation_loss": 0.019870130345225334,
          "train_metric": 0.002400617115199566,
          "validation_metric": 0.0020012466702610254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5484532150003361
        },
        {
          "epoch": 51,
          "updates": 204,
          "samples_seen": 52224,
          "online_loss": 0.0245011686347425,
          "train_loss": 0.023165611550211906,
          "validation_loss": 0.019375061616301537,
          "train_metric": 0.002333155833184719,
          "validation_metric": 0.0019513856386765838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5558855640010734
        },
        {
          "epoch": 52,
          "updates": 208,
          "samples_seen": 53248,
          "online_loss": 0.024070339743047953,
          "train_loss": 0.022831246256828308,
          "validation_loss": 0.01898268796503544,
          "train_metric": 0.0022994796745479107,
          "validation_metric": 0.0019118668278679252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5720022030000109
        },
        {
          "epoch": 53,
          "updates": 212,
          "samples_seen": 54272,
          "online_loss": 0.023007486714050174,
          "train_loss": 0.02238297089934349,
          "validation_loss": 0.018817981705069542,
          "train_metric": 0.0022543310187757015,
          "validation_metric": 0.0018952779937535524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5862122440012172
        },
        {
          "epoch": 54,
          "updates": 216,
          "samples_seen": 55296,
          "online_loss": 0.02305908128619194,
          "train_loss": 0.022052761167287827,
          "validation_loss": 0.018328484147787094,
          "train_metric": 0.002221073256805539,
          "validation_metric": 0.0018459782004356384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6012891030004539
        },
        {
          "epoch": 55,
          "updates": 220,
          "samples_seen": 56320,
          "online_loss": 0.02179556619375944,
          "train_loss": 0.021361352875828743,
          "validation_loss": 0.017963480204343796,
          "train_metric": 0.002151437569409609,
          "validation_metric": 0.0018092163372784853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6102896630000032
        },
        {
          "epoch": 56,
          "updates": 224,
          "samples_seen": 57344,
          "online_loss": 0.022824440151453018,
          "train_loss": 0.020578429102897644,
          "validation_loss": 0.017286915332078934,
          "train_metric": 0.002072583884000778,
          "validation_metric": 0.0017410749569535255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6189232500000799
        },
        {
          "epoch": 57,
          "updates": 228,
          "samples_seen": 58368,
          "online_loss": 0.021623709704726934,
          "train_loss": 0.020176498219370842,
          "validation_loss": 0.017096061259508133,
          "train_metric": 0.002032103016972542,
          "validation_metric": 0.0017218526918441057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6325854540009459
        },
        {
          "epoch": 58,
          "updates": 232,
          "samples_seen": 59392,
          "online_loss": 0.022188946371898055,
          "train_loss": 0.019594835117459297,
          "validation_loss": 0.016539715230464935,
          "train_metric": 0.0019735198002308607,
          "validation_metric": 0.0016658195527270436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7531641750010749
        },
        {
          "epoch": 59,
          "updates": 236,
          "samples_seen": 60416,
          "online_loss": 0.02181227644905448,
          "train_loss": 0.01926298439502716,
          "validation_loss": 0.016160257160663605,
          "train_metric": 0.0019400971941649914,
          "validation_metric": 0.0016276017995551229,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7659086599996954
        },
        {
          "epoch": 60,
          "updates": 240,
          "samples_seen": 61440,
          "online_loss": 0.019953408744186163,
          "train_loss": 0.018792560324072838,
          "validation_loss": 0.015727292746305466,
          "train_metric": 0.0018927177879959345,
          "validation_metric": 0.001583995297551155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7731969400010712
        },
        {
          "epoch": 61,
          "updates": 244,
          "samples_seen": 62464,
          "online_loss": 0.018579754047095776,
          "train_loss": 0.018650224432349205,
          "validation_loss": 0.015783051028847694,
          "train_metric": 0.0018783825216814876,
          "validation_metric": 0.0015896110562607646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7818268540013378
        },
        {
          "epoch": 62,
          "updates": 248,
          "samples_seen": 63488,
          "online_loss": 0.01844689971767366,
          "train_loss": 0.01824452541768551,
          "validation_loss": 0.015111989341676235,
          "train_metric": 0.001837521675042808,
          "validation_metric": 0.0015220242785289884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7888683230012248
        },
        {
          "epoch": 63,
          "updates": 252,
          "samples_seen": 64512,
          "online_loss": 0.018537384225055575,
          "train_loss": 0.01802041567862034,
          "validation_loss": 0.01520947553217411,
          "train_metric": 0.001814950373955071,
          "validation_metric": 0.001531842746771872,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8059368610011006
        },
        {
          "epoch": 64,
          "updates": 256,
          "samples_seen": 65536,
          "online_loss": 0.017587951384484768,
          "train_loss": 0.017542364075779915,
          "validation_loss": 0.014541060663759708,
          "train_metric": 0.0017668025102466345,
          "validation_metric": 0.0014645223272964358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8182243470000685
        },
        {
          "epoch": 65,
          "updates": 260,
          "samples_seen": 66560,
          "online_loss": 0.0173424631357193,
          "train_loss": 0.016777707263827324,
          "validation_loss": 0.0141694825142622,
          "train_metric": 0.0016897892346605659,
          "validation_metric": 0.001427098293788731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8250609360002272
        },
        {
          "epoch": 66,
          "updates": 264,
          "samples_seen": 67584,
          "online_loss": 0.01698760804720223,
          "train_loss": 0.01637251488864422,
          "validation_loss": 0.013827148824930191,
          "train_metric": 0.0016489798435941339,
          "validation_metric": 0.0013926196843385696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8320277410002745
        },
        {
          "epoch": 67,
          "updates": 268,
          "samples_seen": 68608,
          "online_loss": 0.01684815506450832,
          "train_loss": 0.016265256330370903,
          "validation_loss": 0.013443517498672009,
          "train_metric": 0.0016381769673898816,
          "validation_metric": 0.0013539819046854973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8389563960008672
        },
        {
          "epoch": 68,
          "updates": 272,
          "samples_seen": 69632,
          "online_loss": 0.01708283694460988,
          "train_loss": 0.015905462205410004,
          "validation_loss": 0.01340516284108162,
          "train_metric": 0.0016019400209188461,
          "validation_metric": 0.001350119011476636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8485708250009338
        },
        {
          "epoch": 69,
          "updates": 276,
          "samples_seen": 70656,
          "online_loss": 0.016831737710162997,
          "train_loss": 0.015581070445477962,
          "validation_loss": 0.012949459254741669,
          "train_metric": 0.0015692682936787605,
          "validation_metric": 0.001304222154431045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.855491043001166
        },
        {
          "epoch": 70,
          "updates": 280,
          "samples_seen": 71680,
          "online_loss": 0.016493671806529164,
          "train_loss": 0.015100964345037937,
          "validation_loss": 0.01267625018954277,
          "train_metric": 0.001520913909189403,
          "validation_metric": 0.0012767055304720998,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8623511340010737
        },
        {
          "epoch": 71,
          "updates": 284,
          "samples_seen": 72704,
          "online_loss": 0.015238923719152808,
          "train_loss": 0.01482627633959055,
          "validation_loss": 0.01247537974268198,
          "train_metric": 0.0014932481572031975,
          "validation_metric": 0.0012564745265990496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8691739810001309
        },
        {
          "epoch": 72,
          "updates": 288,
          "samples_seen": 73728,
          "online_loss": 0.015084490180015564,
          "train_loss": 0.014699338003993034,
          "validation_loss": 0.012235949747264385,
          "train_metric": 0.0014804634265601635,
          "validation_metric": 0.0012323600240051746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8770894490007777
        },
        {
          "epoch": 73,
          "updates": 292,
          "samples_seen": 74752,
          "online_loss": 0.014928992837667465,
          "train_loss": 0.014552331529557705,
          "validation_loss": 0.01233485247939825,
          "train_metric": 0.0014656574930995703,
          "validation_metric": 0.0012423211010172963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8837443810007244
        },
        {
          "epoch": 74,
          "updates": 296,
          "samples_seen": 75776,
          "online_loss": 0.01512793987058103,
          "train_loss": 0.014268821105360985,
          "validation_loss": 0.01177399605512619,
          "train_metric": 0.0014371032593771815,
          "validation_metric": 0.0011858338257297873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8904916980009148
        },
        {
          "epoch": 75,
          "updates": 300,
          "samples_seen": 76800,
          "online_loss": 0.014339572982862592,
          "train_loss": 0.013847657479345798,
          "validation_loss": 0.01159699447453022,
          "train_metric": 0.001394685241393745,
          "validation_metric": 0.0011680066818371415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.897493939000924
        },
        {
          "epoch": 76,
          "updates": 304,
          "samples_seen": 77824,
          "online_loss": 0.014194600284099579,
          "train_loss": 0.013654490932822227,
          "validation_loss": 0.0115488450974226,
          "train_metric": 0.001375230378471315,
          "validation_metric": 0.0011631574016064405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9042923810011416
        },
        {
          "epoch": 77,
          "updates": 308,
          "samples_seen": 78848,
          "online_loss": 0.013700170209631324,
          "train_loss": 0.013835957273840904,
          "validation_loss": 0.011449619196355343,
          "train_metric": 0.0013935068855062127,
          "validation_metric": 0.0011531637283042073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9111903240009269
        },
        {
          "epoch": 78,
          "updates": 312,
          "samples_seen": 79872,
          "online_loss": 0.013872158480808139,
          "train_loss": 0.013609567657113075,
          "validation_loss": 0.011568338610231876,
          "train_metric": 0.0013707057805731893,
          "validation_metric": 0.00116512062959373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9175679800009675
        },
        {
          "epoch": 79,
          "updates": 316,
          "samples_seen": 80896,
          "online_loss": 0.01335234404541552,
          "train_loss": 0.013157082721590996,
          "validation_loss": 0.010928075760602951,
          "train_metric": 0.0013251331401988864,
          "validation_metric": 0.0011006357381120324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9244628289998218
        },
        {
          "epoch": 80,
          "updates": 320,
          "samples_seen": 81920,
          "online_loss": 0.013183231465518475,
          "train_loss": 0.013102409429848194,
          "validation_loss": 0.011168821714818478,
          "train_metric": 0.0013196266954764724,
          "validation_metric": 0.0011248827213421464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9309893450008531
        },
        {
          "epoch": 81,
          "updates": 324,
          "samples_seen": 82944,
          "online_loss": 0.013170052319765091,
          "train_loss": 0.012894312851130962,
          "validation_loss": 0.010726229287683964,
          "train_metric": 0.0012986678630113602,
          "validation_metric": 0.001080306596122682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9378664980013127
        },
        {
          "epoch": 82,
          "updates": 328,
          "samples_seen": 83968,
          "online_loss": 0.013541600434109569,
          "train_loss": 0.012880174443125725,
          "validation_loss": 0.01103147678077221,
          "train_metric": 0.0012972442200407386,
          "validation_metric": 0.0011110497871413827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9442142070001864
        },
        {
          "epoch": 83,
          "updates": 332,
          "samples_seen": 84992,
          "online_loss": 0.013578899204730988,
          "train_loss": 0.013063226826488972,
          "validation_loss": 0.010631739161908627,
          "train_metric": 0.0013156803324818611,
          "validation_metric": 0.0010707898763939738,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9509416079999937
        },
        {
          "epoch": 84,
          "updates": 336,
          "samples_seen": 86016,
          "online_loss": 0.013711947249248624,
          "train_loss": 0.012829947285354137,
          "validation_loss": 0.010987651534378529,
          "train_metric": 0.0012921856250613928,
          "validation_metric": 0.0011066359002143145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9571415610007534
        },
        {
          "epoch": 85,
          "updates": 340,
          "samples_seen": 87040,
          "online_loss": 0.012859286274760962,
          "train_loss": 0.01229657419025898,
          "validation_loss": 0.010399984195828438,
          "train_metric": 0.0012384661240503192,
          "validation_metric": 0.001047448255121708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9641517210002348
        },
        {
          "epoch": 86,
          "updates": 344,
          "samples_seen": 88064,
          "online_loss": 0.012358339736238122,
          "train_loss": 0.012291665188968182,
          "validation_loss": 0.010506649501621723,
          "train_metric": 0.0012379714753478765,
          "validation_metric": 0.0010581912938505411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9705448950007849
        },
        {
          "epoch": 87,
          "updates": 348,
          "samples_seen": 89088,
          "online_loss": 0.01186005724593997,
          "train_loss": 0.012559205293655396,
          "validation_loss": 0.010213209316134453,
          "train_metric": 0.0012649171985685825,
          "validation_metric": 0.0010286371689289808,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9775213280008757
        },
        {
          "epoch": 88,
          "updates": 352,
          "samples_seen": 90112,
          "online_loss": 0.012342147529125214,
          "train_loss": 0.011946284212172031,
          "validation_loss": 0.010342299938201904,
          "train_metric": 0.0012031861115247011,
          "validation_metric": 0.0010416386649012566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9837616069999058
        },
        {
          "epoch": 89,
          "updates": 356,
          "samples_seen": 91136,
          "online_loss": 0.013310885056853294,
          "train_loss": 0.011477994732558727,
          "validation_loss": 0.009852840565145016,
          "train_metric": 0.0011560216080397367,
          "validation_metric": 0.00099234189838171,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9904621730001963
        },
        {
          "epoch": 90,
          "updates": 360,
          "samples_seen": 92160,
          "online_loss": 0.012980739353224635,
          "train_loss": 0.01173469703644514,
          "validation_loss": 0.009662209078669548,
          "train_metric": 0.0011818758212029934,
          "validation_metric": 0.0009731423342600465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9974491330012825
        },
        {
          "epoch": 91,
          "updates": 364,
          "samples_seen": 93184,
          "online_loss": 0.012864765711128712,
          "train_loss": 0.01167253963649273,
          "validation_loss": 0.010080279782414436,
          "train_metric": 0.00117561558727175,
          "validation_metric": 0.001015248941257596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.003809423000348
        },
        {
          "epoch": 92,
          "updates": 368,
          "samples_seen": 94208,
          "online_loss": 0.012202005600556731,
          "train_loss": 0.012476594187319279,
          "validation_loss": 0.010192147456109524,
          "train_metric": 0.0012565969955176115,
          "validation_metric": 0.0010265156161040068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.010275755999828
        },
        {
          "epoch": 93,
          "updates": 372,
          "samples_seen": 95232,
          "online_loss": 0.014605716336518526,
          "train_loss": 0.012003298848867416,
          "validation_loss": 0.010272122919559479,
          "train_metric": 0.0012089284136891365,
          "validation_metric": 0.0010345705086365342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.01650375600002
        },
        {
          "epoch": 94,
          "updates": 376,
          "samples_seen": 96256,
          "online_loss": 0.011917078867554665,
          "train_loss": 0.011008678004145622,
          "validation_loss": 0.009242658503353596,
          "train_metric": 0.00110875372774899,
          "validation_metric": 0.0009308866574428976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0232618770005502
        },
        {
          "epoch": 95,
          "updates": 380,
          "samples_seen": 97280,
          "online_loss": 0.012748223962262273,
          "train_loss": 0.011334235779941082,
          "validation_loss": 0.009410209953784943,
          "train_metric": 0.0011415426852181554,
          "validation_metric": 0.0009477618150413036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0298011819995736
        },
        {
          "epoch": 96,
          "updates": 384,
          "samples_seen": 98304,
          "online_loss": 0.011365718673914671,
          "train_loss": 0.010697953402996063,
          "validation_loss": 0.00935949943959713,
          "train_metric": 0.001077458611689508,
          "validation_metric": 0.0009426545002497733,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.036105735000092
        },
        {
          "epoch": 97,
          "updates": 388,
          "samples_seen": 99328,
          "online_loss": 0.010768311331048608,
          "train_loss": 0.010592496953904629,
          "validation_loss": 0.009179176762700081,
          "train_metric": 0.0010668374598026276,
          "validation_metric": 0.0009244930115528405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0428636480010027
        },
        {
          "epoch": 98,
          "updates": 392,
          "samples_seen": 100352,
          "online_loss": 0.010830871993675828,
          "train_loss": 0.010635006241500378,
          "validation_loss": 0.0089234234765172,
          "train_metric": 0.0010711188660934567,
          "validation_metric": 0.0008987344335764647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0499819590004336
        },
        {
          "epoch": 99,
          "updates": 396,
          "samples_seen": 101376,
          "online_loss": 0.0123350212816149,
          "train_loss": 0.012094003148376942,
          "validation_loss": 0.010551251471042633,
          "train_metric": 0.0012180637568235397,
          "validation_metric": 0.001062683411873877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0563585020008759
        },
        {
          "epoch": 100,
          "updates": 400,
          "samples_seen": 102400,
          "online_loss": 0.011943624587729573,
          "train_loss": 0.011324156075716019,
          "validation_loss": 0.009220574051141739,
          "train_metric": 0.0011405275436118245,
          "validation_metric": 0.0009286624845117331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0652075039997726
        },
        {
          "epoch": 101,
          "updates": 404,
          "samples_seen": 103424,
          "online_loss": 0.010677092242985964,
          "train_loss": 0.010677406564354897,
          "validation_loss": 0.009249897673726082,
          "train_metric": 0.0010753892129287124,
          "validation_metric": 0.0009316158248111606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0715548340012901
        },
        {
          "epoch": 102,
          "updates": 408,
          "samples_seen": 104448,
          "online_loss": 0.011492251651361585,
          "train_loss": 0.011525267735123634,
          "validation_loss": 0.009417816996574402,
          "train_metric": 0.0011607827618718147,
          "validation_metric": 0.0009485280606895685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0780233519999456
        },
        {
          "epoch": 103,
          "updates": 412,
          "samples_seen": 105472,
          "online_loss": 0.011433448060415685,
          "train_loss": 0.01141116302460432,
          "validation_loss": 0.009994796477258205,
          "train_metric": 0.0011492905905470252,
          "validation_metric": 0.0010066393297165632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.084249464000095
        },
        {
          "epoch": 104,
          "updates": 416,
          "samples_seen": 106496,
          "online_loss": 0.013884158805012703,
          "train_loss": 0.01602601446211338,
          "validation_loss": 0.012048078700900078,
          "train_metric": 0.0016140814404934645,
          "validation_metric": 0.0012134384596720338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0905269310005679
        },
        {
          "epoch": 105,
          "updates": 420,
          "samples_seen": 107520,
          "online_loss": 0.015585960121825337,
          "train_loss": 0.015901101753115654,
          "validation_loss": 0.013339366763830185,
          "train_metric": 0.001601500785909593,
          "validation_metric": 0.0013434923021122813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0980074590006552
        },
        {
          "epoch": 106,
          "updates": 424,
          "samples_seen": 108544,
          "online_loss": 0.013450263300910592,
          "train_loss": 0.013010220602154732,
          "validation_loss": 0.009994863532483578,
          "train_metric": 0.0013103416422381997,
          "validation_metric": 0.0010066460818052292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1043058350005595
        },
        {
          "epoch": 107,
          "updates": 428,
          "samples_seen": 109568,
          "online_loss": 0.012856036657467484,
          "train_loss": 0.011302098631858826,
          "validation_loss": 0.0096169114112854,
          "train_metric": 0.0011383059900254011,
          "validation_metric": 0.0009685800177976489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1148444829996151
        },
        {
          "epoch": 108,
          "updates": 432,
          "samples_seen": 110592,
          "online_loss": 0.01448991079814732,
          "train_loss": 0.011067896150052547,
          "validation_loss": 0.00902693159878254,
          "train_metric": 0.0011147179175168276,
          "validation_metric": 0.0009091593674384058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1654418880007142
        },
        {
          "epoch": 109,
          "updates": 436,
          "samples_seen": 111616,
          "online_loss": 0.013184086885303259,
          "train_loss": 0.009466356597840786,
          "validation_loss": 0.008135948330163956,
          "train_metric": 0.000953416689299047,
          "validation_metric": 0.0008194227702915668,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2571456710011262
        },
        {
          "epoch": 110,
          "updates": 440,
          "samples_seen": 112640,
          "online_loss": 0.011551156640052795,
          "train_loss": 0.009416744112968445,
          "validation_loss": 0.00812205858528614,
          "train_metric": 0.000948420085478574,
          "validation_metric": 0.0008180239819921553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3517288920011197
        },
        {
          "epoch": 111,
          "updates": 444,
          "samples_seen": 113664,
          "online_loss": 0.011012779548764229,
          "train_loss": 0.009925923310220242,
          "validation_loss": 0.0081639988347888,
          "train_metric": 0.0009997024899348617,
          "validation_metric": 0.0008222479955293238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4235878660001617
        },
        {
          "epoch": 112,
          "updates": 448,
          "samples_seen": 114688,
          "online_loss": 0.011505345348268747,
          "train_loss": 0.009733986109495163,
          "validation_loss": 0.008503604680299759,
          "train_metric": 0.0009803713764995337,
          "validation_metric": 0.0008564520394429564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4901578139997582
        },
        {
          "epoch": 113,
          "updates": 452,
          "samples_seen": 115712,
          "online_loss": 0.009817319340072572,
          "train_loss": 0.0102904187515378,
          "validation_loss": 0.00849820300936699,
          "train_metric": 0.0010364132467657328,
          "validation_metric": 0.0008559078560210764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5634018939999805
        },
        {
          "epoch": 114,
          "updates": 456,
          "samples_seen": 116736,
          "online_loss": 0.010039204265922308,
          "train_loss": 0.009876638650894165,
          "validation_loss": 0.00867165345698595,
          "train_metric": 0.0009947388898581266,
          "validation_metric": 0.0008733771974220872,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.644255017999967
        },
        {
          "epoch": 115,
          "updates": 460,
          "samples_seen": 117760,
          "online_loss": 0.010788581450469792,
          "train_loss": 0.009058278985321522,
          "validation_loss": 0.007766430731862783,
          "train_metric": 0.000912316667381674,
          "validation_metric": 0.0007822063635103405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6867286699998658
        },
        {
          "epoch": 116,
          "updates": 464,
          "samples_seen": 118784,
          "online_loss": 0.009217860409989953,
          "train_loss": 0.009065975435078144,
          "validation_loss": 0.007953179068863392,
          "train_metric": 0.0009130918188020587,
          "validation_metric": 0.000801015121396631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7446849010011647
        },
        {
          "epoch": 117,
          "updates": 468,
          "samples_seen": 119808,
          "online_loss": 0.008780061267316341,
          "train_loss": 0.009019536897540092,
          "validation_loss": 0.007749817334115505,
          "train_metric": 0.0009084146586246789,
          "validation_metric": 0.0007805332425050437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8246414019995427
        },
        {
          "epoch": 118,
          "updates": 472,
          "samples_seen": 120832,
          "online_loss": 0.008856099680997431,
          "train_loss": 0.008942189626395702,
          "validation_loss": 0.007965302094817162,
          "train_metric": 0.0009006245527416468,
          "validation_metric": 0.0008022361434996128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.895072651001101
        },
        {
          "epoch": 119,
          "updates": 476,
          "samples_seen": 121856,
          "online_loss": 0.009643364464864135,
          "train_loss": 0.009479436092078686,
          "validation_loss": 0.007829059846699238,
          "train_metric": 0.0009547340450808406,
          "validation_metric": 0.0007885143859311938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9882347780003329
        },
        {
          "epoch": 120,
          "updates": 480,
          "samples_seen": 122880,
          "online_loss": 0.009901220095343888,
          "train_loss": 0.008752468973398209,
          "validation_loss": 0.007807506248354912,
          "train_metric": 0.0008815167238935828,
          "validation_metric": 0.0007863434148021042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0642683700007183
        },
        {
          "epoch": 121,
          "updates": 484,
          "samples_seen": 123904,
          "online_loss": 0.008822341333143413,
          "train_loss": 0.008394516073167324,
          "validation_loss": 0.0073541635647416115,
          "train_metric": 0.000845464994199574,
          "validation_metric": 0.0007406844524666667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1019521050002368
        },
        {
          "epoch": 122,
          "updates": 488,
          "samples_seen": 124928,
          "online_loss": 0.009220734238624573,
          "train_loss": 0.00832668598741293,
          "validation_loss": 0.0072213029488921165,
          "train_metric": 0.0008386332192458212,
          "validation_metric": 0.0007273031515069306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.163620396000624
        },
        {
          "epoch": 123,
          "updates": 492,
          "samples_seen": 125952,
          "online_loss": 0.009187388001009822,
          "train_loss": 0.008267062716186047,
          "validation_loss": 0.007265729364007711,
          "train_metric": 0.0008326282259076834,
          "validation_metric": 0.0007317776908166707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2014880890001223
        },
        {
          "epoch": 124,
          "updates": 496,
          "samples_seen": 126976,
          "online_loss": 0.009668856160715222,
          "train_loss": 0.008249476552009583,
          "validation_loss": 0.007331956177949905,
          "train_metric": 0.0008308570832014084,
          "validation_metric": 0.0007384478230960667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2555486380006187
        },
        {
          "epoch": 125,
          "updates": 500,
          "samples_seen": 128000,
          "online_loss": 0.00842579931486398,
          "train_loss": 0.008085557259619236,
          "validation_loss": 0.006985497660934925,
          "train_metric": 0.0008143478189595044,
          "validation_metric": 0.0007035537855699658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.308094306001294
        },
        {
          "epoch": 126,
          "updates": 504,
          "samples_seen": 129024,
          "online_loss": 0.008113076677545905,
          "train_loss": 0.008146866224706173,
          "validation_loss": 0.007153032813221216,
          "train_metric": 0.000820522487629205,
          "validation_metric": 0.0007204273133538663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.601124462000371
        },
        {
          "epoch": 127,
          "updates": 508,
          "samples_seen": 130048,
          "online_loss": 0.008126242784783244,
          "train_loss": 0.007920407690107822,
          "validation_loss": 0.006912947632372379,
          "train_metric": 0.000797714339569211,
          "validation_metric": 0.0006962466868571937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6373432990003494
        },
        {
          "epoch": 128,
          "updates": 512,
          "samples_seen": 131072,
          "online_loss": 0.008043643669225276,
          "train_loss": 0.007881473749876022,
          "validation_loss": 0.006861462723463774,
          "train_metric": 0.0007937931804917753,
          "validation_metric": 0.0006910613738000393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6756480770000053
        },
        {
          "epoch": 129,
          "updates": 516,
          "samples_seen": 132096,
          "online_loss": 0.007939411094412208,
          "train_loss": 0.00802565272897482,
          "validation_loss": 0.007142952177673578,
          "train_metric": 0.0008083143620751798,
          "validation_metric": 0.0007194119389168918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.708922908999739
        },
        {
          "epoch": 130,
          "updates": 520,
          "samples_seen": 133120,
          "online_loss": 0.007947169709950686,
          "train_loss": 0.007737984415143728,
          "validation_loss": 0.006691594608128071,
          "train_metric": 0.0007793414406478405,
          "validation_metric": 0.000673952919896692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.745243957000639
        },
        {
          "epoch": 131,
          "updates": 524,
          "samples_seen": 134144,
          "online_loss": 0.007808164926245809,
          "train_loss": 0.007600081618875265,
          "validation_loss": 0.006692728493362665,
          "train_metric": 0.0007654523360542953,
          "validation_metric": 0.0006740671233274043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7790069530001347
        },
        {
          "epoch": 132,
          "updates": 528,
          "samples_seen": 135168,
          "online_loss": 0.00816407147794962,
          "train_loss": 0.007703413255512714,
          "validation_loss": 0.006863360293209553,
          "train_metric": 0.0007758595165796578,
          "validation_metric": 0.00069125258596614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8133437220003543
        },
        {
          "epoch": 133,
          "updates": 532,
          "samples_seen": 136192,
          "online_loss": 0.007738560903817415,
          "train_loss": 0.007458015810698271,
          "validation_loss": 0.006449965760111809,
          "train_metric": 0.0007511440780945122,
          "validation_metric": 0.0006496169953607023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8491116559998773
        },
        {
          "epoch": 134,
          "updates": 536,
          "samples_seen": 137216,
          "online_loss": 0.007549898815341294,
          "train_loss": 0.007396166678518057,
          "validation_loss": 0.006485652178525925,
          "train_metric": 0.0007449148688465357,
          "validation_metric": 0.0006532111437991261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.883381609000935
        },
        {
          "epoch": 135,
          "updates": 540,
          "samples_seen": 138240,
          "online_loss": 0.007509160786867142,
          "train_loss": 0.007530219852924347,
          "validation_loss": 0.006490312982350588,
          "train_metric": 0.0007584161357954144,
          "validation_metric": 0.0006536805303767323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9166095100008533
        },
        {
          "epoch": 136,
          "updates": 544,
          "samples_seen": 139264,
          "online_loss": 0.007658005808480084,
          "train_loss": 0.007221882231533527,
          "validation_loss": 0.0062768105417490005,
          "train_metric": 0.0007273615919984877,
          "validation_metric": 0.0006321773398667574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9519965710005636
        },
        {
          "epoch": 137,
          "updates": 548,
          "samples_seen": 140288,
          "online_loss": 0.007493449957109988,
          "train_loss": 0.007151769008487463,
          "validation_loss": 0.006257520988583565,
          "train_metric": 0.0007203000714071095,
          "validation_metric": 0.0006302346009761095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9887211450004543
        },
        {
          "epoch": 138,
          "updates": 552,
          "samples_seen": 141312,
          "online_loss": 0.007304032566025853,
          "train_loss": 0.007187902927398682,
          "validation_loss": 0.006208083592355251,
          "train_metric": 0.0007239392725750804,
          "validation_metric": 0.0006252555176615715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0261525620007887
        },
        {
          "epoch": 139,
          "updates": 556,
          "samples_seen": 142336,
          "online_loss": 0.007407858269289136,
          "train_loss": 0.0074580661021173,
          "validation_loss": 0.00661054439842701,
          "train_metric": 0.0007511490257456899,
          "validation_metric": 0.0006657898775301874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0613726980009233
        },
        {
          "epoch": 140,
          "updates": 560,
          "samples_seen": 143360,
          "online_loss": 0.007848967565223575,
          "train_loss": 0.0069804079830646515,
          "validation_loss": 0.0059867617674171925,
          "train_metric": 0.0007030410342849791,
          "validation_metric": 0.0006029647774994373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.098092125999756
        },
        {
          "epoch": 141,
          "updates": 564,
          "samples_seen": 144384,
          "online_loss": 0.006949851172976196,
          "train_loss": 0.007131040561944246,
          "validation_loss": 0.006295507308095694,
          "train_metric": 0.0007182122208178043,
          "validation_metric": 0.000634060415904969,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1314159410012508
        },
        {
          "epoch": 142,
          "updates": 568,
          "samples_seen": 145408,
          "online_loss": 0.007022133097052574,
          "train_loss": 0.00697439257055521,
          "validation_loss": 0.005939807277172804,
          "train_metric": 0.0007024351507425308,
          "validation_metric": 0.0005982356378808618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.167724671000542
        },
        {
          "epoch": 143,
          "updates": 572,
          "samples_seen": 146432,
          "online_loss": 0.0070856367237865925,
          "train_loss": 0.006769709289073944,
          "validation_loss": 0.005939200986176729,
          "train_metric": 0.0006818202673457563,
          "validation_metric": 0.0005981746362522244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.203456234999976
        },
        {
          "epoch": 144,
          "updates": 576,
          "samples_seen": 147456,
          "online_loss": 0.00679561251308769,
          "train_loss": 0.006763513199985027,
          "validation_loss": 0.005761445965617895,
          "train_metric": 0.0006811963394284248,
          "validation_metric": 0.0005802717641927302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2395464340006583
        },
        {
          "epoch": 145,
          "updates": 580,
          "samples_seen": 148480,
          "online_loss": 0.007329074782319367,
          "train_loss": 0.006620769854635,
          "validation_loss": 0.00571850873529911,
          "train_metric": 0.0006668196874670684,
          "validation_metric": 0.0005759472260251641,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2760511200012843
        },
        {
          "epoch": 146,
          "updates": 584,
          "samples_seen": 149504,
          "online_loss": 0.007134231505915523,
          "train_loss": 0.0066349389962852,
          "validation_loss": 0.0059576393105089664,
          "train_metric": 0.000668246706482023,
          "validation_metric": 0.0006000315770506859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3101298210003733
        },
        {
          "epoch": 147,
          "updates": 588,
          "samples_seen": 150528,
          "online_loss": 0.006819091620855033,
          "train_loss": 0.006609054747968912,
          "validation_loss": 0.0056187184527516365,
          "train_metric": 0.0006656398181803524,
          "validation_metric": 0.0005658968584612012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3464863839999452
        },
        {
          "epoch": 148,
          "updates": 592,
          "samples_seen": 151552,
          "online_loss": 0.006713250302709639,
          "train_loss": 0.006465293001383543,
          "validation_loss": 0.0054846894927322865,
          "train_metric": 0.0006511606043204665,
          "validation_metric": 0.0005523978034034371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3833634080001502
        },
        {
          "epoch": 149,
          "updates": 596,
          "samples_seen": 152576,
          "online_loss": 0.0064540376188233495,
          "train_loss": 0.006415774114429951,
          "validation_loss": 0.005661151371896267,
          "train_metric": 0.0006461733137257397,
          "validation_metric": 0.0005701704067178071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.41759395400004
        },
        {
          "epoch": 150,
          "updates": 600,
          "samples_seen": 153600,
          "online_loss": 0.006451577879488468,
          "train_loss": 0.0063376217149198055,
          "validation_loss": 0.005407122895121574,
          "train_metric": 0.0006383020663633943,
          "validation_metric": 0.0005445856368169188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4544010530007654
        },
        {
          "epoch": 151,
          "updates": 604,
          "samples_seen": 154624,
          "online_loss": 0.006540414760820568,
          "train_loss": 0.006263326853513718,
          "validation_loss": 0.005426486954092979,
          "train_metric": 0.000630819471552968,
          "validation_metric": 0.0005465359427034855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5199696060008137
        },
        {
          "epoch": 152,
          "updates": 608,
          "samples_seen": 155648,
          "online_loss": 0.006291868863627315,
          "train_loss": 0.006421277299523354,
          "validation_loss": 0.005405659321695566,
          "train_metric": 0.0006467276834882796,
          "validation_metric": 0.000544438196811825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.560469551001006
        },
        {
          "epoch": 153,
          "updates": 612,
          "samples_seen": 156672,
          "online_loss": 0.006607759976759553,
          "train_loss": 0.006255362648516893,
          "validation_loss": 0.005326671525835991,
          "train_metric": 0.0006300173117779195,
          "validation_metric": 0.0005364828393794596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.605294617000254
        },
        {
          "epoch": 154,
          "updates": 616,
          "samples_seen": 157696,
          "online_loss": 0.00633845804259181,
          "train_loss": 0.006341890897601843,
          "validation_loss": 0.0055468459613621235,
          "train_metric": 0.0006387319881469011,
          "validation_metric": 0.0005586580373346806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6742682399999467
        },
        {
          "epoch": 155,
          "updates": 620,
          "samples_seen": 158720,
          "online_loss": 0.006515808287076652,
          "train_loss": 0.006034699734300375,
          "validation_loss": 0.005144215188920498,
          "train_metric": 0.0006077928119339049,
          "validation_metric": 0.0005181065062060952,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7107756240002345
        },
        {
          "epoch": 156,
          "updates": 624,
          "samples_seen": 159744,
          "online_loss": 0.006629477487877011,
          "train_loss": 0.006404663436114788,
          "validation_loss": 0.005273404531180859,
          "train_metric": 0.0006450542714446783,
          "validation_metric": 0.0005311179556883872,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7447205480002594
        },
        {
          "epoch": 157,
          "updates": 628,
          "samples_seen": 160768,
          "online_loss": 0.006995784700848162,
          "train_loss": 0.008764564990997314,
          "validation_loss": 0.007306552492082119,
          "train_metric": 0.000882734777405858,
          "validation_metric": 0.0007358891889452934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7785815250008454
        },
        {
          "epoch": 158,
          "updates": 632,
          "samples_seen": 161792,
          "online_loss": 0.008240191731601954,
          "train_loss": 0.008250937797129154,
          "validation_loss": 0.006117409560829401,
          "train_metric": 0.0008310042321681976,
          "validation_metric": 0.000616123026702553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8136221610002394
        },
        {
          "epoch": 159,
          "updates": 636,
          "samples_seen": 162816,
          "online_loss": 0.007837836630642414,
          "train_loss": 0.006188474595546722,
          "validation_loss": 0.005429375451058149,
          "train_metric": 0.0006232804735191166,
          "validation_metric": 0.000546826864592731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.846738047001054
        },
        {
          "epoch": 160,
          "updates": 640,
          "samples_seen": 163840,
          "online_loss": 0.0063129502814263105,
          "train_loss": 0.005792725831270218,
          "validation_loss": 0.004923227708786726,
          "train_metric": 0.0005834221374243498,
          "validation_metric": 0.0004958495264872909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8824972239999624
        },
        {
          "epoch": 161,
          "updates": 644,
          "samples_seen": 164864,
          "online_loss": 0.00610811822116375,
          "train_loss": 0.005854789167642593,
          "validation_loss": 0.004939508158713579,
          "train_metric": 0.0005896729417145252,
          "validation_metric": 0.0004974891198799014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9160406890005106
        },
        {
          "epoch": 162,
          "updates": 648,
          "samples_seen": 165888,
          "online_loss": 0.005902223638258874,
          "train_loss": 0.005696673411875963,
          "validation_loss": 0.004870845470577478,
          "train_metric": 0.0005737480241805315,
          "validation_metric": 0.0004905738169327378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9518406610004604
        },
        {
          "epoch": 163,
          "updates": 652,
          "samples_seen": 166912,
          "online_loss": 0.006119434256106615,
          "train_loss": 0.005734091624617577,
          "validation_loss": 0.004833959974348545,
          "train_metric": 0.0005775167956016958,
          "validation_metric": 0.0004868588293902576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.988811348001036
        },
        {
          "epoch": 164,
          "updates": 656,
          "samples_seen": 167936,
          "online_loss": 0.006356395431794226,
          "train_loss": 0.005936636123806238,
          "validation_loss": 0.005238578654825687,
          "train_metric": 0.0005979162524454296,
          "validation_metric": 0.0005276104784570634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.023496330000853
        },
        {
          "epoch": 165,
          "updates": 660,
          "samples_seen": 168960,
          "online_loss": 0.006274535553529859,
          "train_loss": 0.007161206565797329,
          "validation_loss": 0.00533342594280839,
          "train_metric": 0.0007212506025098264,
          "validation_metric": 0.0005371631123125553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.057900175999748
        },
        {
          "epoch": 166,
          "updates": 664,
          "samples_seen": 169984,
          "online_loss": 0.008600012166425586,
          "train_loss": 0.009845152497291565,
          "validation_loss": 0.008329850621521473,
          "train_metric": 0.000991567736491561,
          "validation_metric": 0.0008389520226046443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.091497819001233
        },
        {
          "epoch": 167,
          "updates": 668,
          "samples_seen": 171008,
          "online_loss": 0.008408086723648012,
          "train_loss": 0.00943647138774395,
          "validation_loss": 0.006652262061834335,
          "train_metric": 0.0009504067711532116,
          "validation_metric": 0.0006699913647025824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.125671231000524
        },
        {
          "epoch": 168,
          "updates": 672,
          "samples_seen": 172032,
          "online_loss": 0.0085602100007236,
          "train_loss": 0.009680913761258125,
          "validation_loss": 0.00766993872821331,
          "train_metric": 0.0009750261669978499,
          "validation_metric": 0.0007724880706518888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.159567009999591
        },
        {
          "epoch": 169,
          "updates": 676,
          "samples_seen": 173056,
          "online_loss": 0.00826445803977549,
          "train_loss": 0.009849182330071926,
          "validation_loss": 0.007117017172276974,
          "train_metric": 0.0009919734438881278,
          "validation_metric": 0.0007167997537180781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.193145797000398
        },
        {
          "epoch": 170,
          "updates": 680,
          "samples_seen": 174080,
          "online_loss": 0.007260025478899479,
          "train_loss": 0.008649499155580997,
          "validation_loss": 0.007112556602805853,
          "train_metric": 0.0008711458649486303,
          "validation_metric": 0.0007163506234064698,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.226708182000948
        },
        {
          "epoch": 171,
          "updates": 684,
          "samples_seen": 175104,
          "online_loss": 0.006968381232582033,
          "train_loss": 0.007729665841907263,
          "validation_loss": 0.005835152231156826,
          "train_metric": 0.000778503599576652,
          "validation_metric": 0.0005876952782273293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.260670970001229
        },
        {
          "epoch": 172,
          "updates": 688,
          "samples_seen": 176128,
          "online_loss": 0.006827189004980028,
          "train_loss": 0.005754518788307905,
          "validation_loss": 0.005062179174274206,
          "train_metric": 0.0005795740289613605,
          "validation_metric": 0.0005098441615700722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.294929556999705
        },
        {
          "epoch": 173,
          "updates": 692,
          "samples_seen": 177152,
          "online_loss": 0.005838778568431735,
          "train_loss": 0.0055382433347404,
          "validation_loss": 0.004564330447465181,
          "train_metric": 0.0005577916163019836,
          "validation_metric": 0.0004597026272676885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.3309086589997605
        },
        {
          "epoch": 174,
          "updates": 696,
          "samples_seen": 178176,
          "online_loss": 0.0064271793235093355,
          "train_loss": 0.005321048200130463,
          "validation_loss": 0.0044740610755980015,
          "train_metric": 0.0005359164206311107,
          "validation_metric": 0.0004506110562942922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.366901143001087
        },
        {
          "epoch": 175,
          "updates": 700,
          "samples_seen": 179200,
          "online_loss": 0.005881970748305321,
          "train_loss": 0.0052443197928369045,
          "validation_loss": 0.004447183571755886,
          "train_metric": 0.0005281887133605778,
          "validation_metric": 0.00044790407991968095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.403288726000028
        },
        {
          "epoch": 176,
          "updates": 704,
          "samples_seen": 180224,
          "online_loss": 0.005492116091772914,
          "train_loss": 0.005243706051260233,
          "validation_loss": 0.004370907787233591,
          "train_metric": 0.0005281268968246877,
          "validation_metric": 0.00044022180372849107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.442424762000883
        },
        {
          "epoch": 177,
          "updates": 708,
          "samples_seen": 181248,
          "online_loss": 0.0062004722421988845,
          "train_loss": 0.005162710789591074,
          "validation_loss": 0.004505534190684557,
          "train_metric": 0.000519969267770648,
          "validation_metric": 0.00045378089998848736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.664043112999934
        },
        {
          "epoch": 178,
          "updates": 712,
          "samples_seen": 182272,
          "online_loss": 0.00514081388246268,
          "train_loss": 0.005145032424479723,
          "validation_loss": 0.004399300552904606,
          "train_metric": 0.0005181888118386269,
          "validation_metric": 0.00044308145879767835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.697569296000438
        },
        {
          "epoch": 179,
          "updates": 716,
          "samples_seen": 183296,
          "online_loss": 0.005799281294457614,
          "train_loss": 0.005651326850056648,
          "validation_loss": 0.004447731655091047,
          "train_metric": 0.000569180934689939,
          "validation_metric": 0.0004479592898860574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.730677300000025
        },
        {
          "epoch": 180,
          "updates": 720,
          "samples_seen": 184320,
          "online_loss": 0.005620143609121442,
          "train_loss": 0.00500295776873827,
          "validation_loss": 0.004295173101127148,
          "train_metric": 0.0005038796225562692,
          "validation_metric": 0.000432594126323238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.766815525999846
        },
        {
          "epoch": 181,
          "updates": 724,
          "samples_seen": 185344,
          "online_loss": 0.005256966687738895,
          "train_loss": 0.005198282189667225,
          "validation_loss": 0.004260700196027756,
          "train_metric": 0.0005235520075075328,
          "validation_metric": 0.0004291220975574106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.802976478000346
        },
        {
          "epoch": 182,
          "updates": 728,
          "samples_seen": 186368,
          "online_loss": 0.005268436740152538,
          "train_loss": 0.004981015808880329,
          "validation_loss": 0.004387909546494484,
          "train_metric": 0.0005016695940867066,
          "validation_metric": 0.0004419341858010739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.836682814000596
        },
        {
          "epoch": 183,
          "updates": 732,
          "samples_seen": 187392,
          "online_loss": 0.005184005363844335,
          "train_loss": 0.004903919529169798,
          "validation_loss": 0.004215721972286701,
          "train_metric": 0.000493904750328511,
          "validation_metric": 0.00042459211545065045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.872862520000126
        },
        {
          "epoch": 184,
          "updates": 736,
          "samples_seen": 188416,
          "online_loss": 0.005011343746446073,
          "train_loss": 0.005173346493393183,
          "validation_loss": 0.004161322955042124,
          "train_metric": 0.0005210405215620995,
          "validation_metric": 0.0004191132029518485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.909145355000874
        },
        {
          "epoch": 185,
          "updates": 740,
          "samples_seen": 189440,
          "online_loss": 0.005075042019598186,
          "train_loss": 0.004967402201145887,
          "validation_loss": 0.004372223746031523,
          "train_metric": 0.000500298454426229,
          "validation_metric": 0.00044035434257239103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.942964462999953
        },
        {
          "epoch": 186,
          "updates": 744,
          "samples_seen": 190464,
          "online_loss": 0.005136806168593466,
          "train_loss": 0.0048463428393006325,
          "validation_loss": 0.003978007473051548,
          "train_metric": 0.00048810598673298955,
          "validation_metric": 0.00040065040229819715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.978872245001185
        },
        {
          "epoch": 187,
          "updates": 748,
          "samples_seen": 191488,
          "online_loss": 0.005345261306501925,
          "train_loss": 0.0047306278720498085,
          "validation_loss": 0.003987557254731655,
          "train_metric": 0.00047645156155340374,
          "validation_metric": 0.00040161219658330083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.013662993000253
        },
        {
          "epoch": 188,
          "updates": 752,
          "samples_seen": 192512,
          "online_loss": 0.0058598428731784225,
          "train_loss": 0.005099005065858364,
          "validation_loss": 0.0044996775686740875,
          "train_metric": 0.0005135531537234783,
          "validation_metric": 0.0004531910235527903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.0469459790001565
        },
        {
          "epoch": 189,
          "updates": 756,
          "samples_seen": 193536,
          "online_loss": 0.005117974535096437,
          "train_loss": 0.0052747237496078014,
          "validation_loss": 0.004073804710060358,
          "train_metric": 0.0005312509019859135,
          "validation_metric": 0.0004102987004444003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.080486844999541
        },
        {
          "epoch": 190,
          "updates": 760,
          "samples_seen": 194560,
          "online_loss": 0.0052022491581737995,
          "train_loss": 0.0046471478417515755,
          "validation_loss": 0.004058862570673227,
          "train_metric": 0.0004680437268689275,
          "validation_metric": 0.000408793828682974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.1140345740004705
        },
        {
          "epoch": 191,
          "updates": 764,
          "samples_seen": 195584,
          "online_loss": 0.004855574225075543,
          "train_loss": 0.0045766993425786495,
          "validation_loss": 0.00393061526119709,
          "train_metric": 0.00046094838762655854,
          "validation_metric": 0.0003958771703764796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.148999236000236
        },
        {
          "epoch": 192,
          "updates": 768,
          "samples_seen": 196608,
          "online_loss": 0.004789818078279495,
          "train_loss": 0.004717552103102207,
          "validation_loss": 0.0037822765298187733,
          "train_metric": 0.0004751345841214061,
          "validation_metric": 0.0003809370391536504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.184621313999742
        },
        {
          "epoch": 193,
          "updates": 772,
          "samples_seen": 197632,
          "online_loss": 0.004950490081682801,
          "train_loss": 0.004530223552137613,
          "validation_loss": 0.003936830908060074,
          "train_metric": 0.0004562674730550498,
          "validation_metric": 0.00039650319376960397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.219694900000832
        },
        {
          "epoch": 194,
          "updates": 776,
          "samples_seen": 198656,
          "online_loss": 0.004827813128940761,
          "train_loss": 0.004827161319553852,
          "validation_loss": 0.004226321820169687,
          "train_metric": 0.00048617407446727157,
          "validation_metric": 0.00042565964395180345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.2532154970012925
        },
        {
          "epoch": 195,
          "updates": 780,
          "samples_seen": 199680,
          "online_loss": 0.004998187941964716,
          "train_loss": 0.005355621688067913,
          "validation_loss": 0.004086405970156193,
          "train_metric": 0.0005393986357375979,
          "validation_metric": 0.00041156786028295755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.2868172060007055
        },
        {
          "epoch": 196,
          "updates": 784,
          "samples_seen": 200704,
          "online_loss": 0.0056822969345375896,
          "train_loss": 0.004424688871949911,
          "validation_loss": 0.0037610565777868032,
          "train_metric": 0.0004456384340301156,
          "validation_metric": 0.0003787998284678906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.323186915000406
        },
        {
          "epoch": 197,
          "updates": 788,
          "samples_seen": 201728,
          "online_loss": 0.0045434279018081725,
          "train_loss": 0.0043828124180436134,
          "validation_loss": 0.003769872710108757,
          "train_metric": 0.0004414208233356476,
          "validation_metric": 0.0003796877572312951,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.35727626500011
        },
        {
          "epoch": 198,
          "updates": 792,
          "samples_seen": 202752,
          "online_loss": 0.004422282159794122,
          "train_loss": 0.004444758407771587,
          "validation_loss": 0.0036791167221963406,
          "train_metric": 0.00044765978236682713,
          "validation_metric": 0.0003705472045112401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.393616859000758
        },
        {
          "epoch": 199,
          "updates": 796,
          "samples_seen": 203776,
          "online_loss": 0.004822818678803742,
          "train_loss": 0.004860087763518095,
          "validation_loss": 0.00427228631451726,
          "train_metric": 0.0004894902231171727,
          "validation_metric": 0.00043028901563957334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.429024099999879
        },
        {
          "epoch": 200,
          "updates": 800,
          "samples_seen": 204800,
          "online_loss": 0.0046162495273165405,
          "train_loss": 0.004604123532772064,
          "validation_loss": 0.003616696922108531,
          "train_metric": 0.00046371042844839394,
          "validation_metric": 0.00036426051519811153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.464562508999734
        },
        {
          "epoch": 201,
          "updates": 804,
          "samples_seen": 205824,
          "online_loss": 0.004309505398850888,
          "train_loss": 0.0045021711848676205,
          "validation_loss": 0.003958823159337044,
          "train_metric": 0.00045344216050580144,
          "validation_metric": 0.00039871811168268323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.498494723000476
        },
        {
          "epoch": 202,
          "updates": 808,
          "samples_seen": 206848,
          "online_loss": 0.004451628483366221,
          "train_loss": 0.004229309502989054,
          "validation_loss": 0.0034852270036935806,
          "train_metric": 0.0004259605484548956,
          "validation_metric": 0.0003510192909743637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.534273806000783
        },
        {
          "epoch": 203,
          "updates": 812,
          "samples_seen": 207872,
          "online_loss": 0.004389816022012383,
          "train_loss": 0.004282766487449408,
          "validation_loss": 0.003740210086107254,
          "train_metric": 0.0004313445242587477,
          "validation_metric": 0.0003767001908272505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.567861819999962
        },
        {
          "epoch": 204,
          "updates": 816,
          "samples_seen": 208896,
          "online_loss": 0.004512218467425555,
          "train_loss": 0.0041525717824697495,
          "validation_loss": 0.003567455103620887,
          "train_metric": 0.0004182318225502968,
          "validation_metric": 0.00035930104786530137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.601474663000772
        },
        {
          "epoch": 205,
          "updates": 820,
          "samples_seen": 209920,
          "online_loss": 0.004225142591167241,
          "train_loss": 0.004165322985500097,
          "validation_loss": 0.003461267100647092,
          "train_metric": 0.00041951608727686107,
          "validation_metric": 0.00034860611776821315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.638166742999601
        },
        {
          "epoch": 206,
          "updates": 824,
          "samples_seen": 210944,
          "online_loss": 0.004178076051175594,
          "train_loss": 0.004073341377079487,
          "validation_loss": 0.0035255420953035355,
          "train_metric": 0.0004102520179003477,
          "validation_metric": 0.000355079653672874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.67247817200041
        },
        {
          "epoch": 207,
          "updates": 828,
          "samples_seen": 211968,
          "online_loss": 0.004170366213656962,
          "train_loss": 0.004213447216898203,
          "validation_loss": 0.00342695158906281,
          "train_metric": 0.00042436295188963413,
          "validation_metric": 0.000345150037901476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.707890744999531
        },
        {
          "epoch": 208,
          "updates": 832,
          "samples_seen": 212992,
          "online_loss": 0.004123570863157511,
          "train_loss": 0.004253590013831854,
          "validation_loss": 0.003764943452551961,
          "train_metric": 0.00042840599780902267,
          "validation_metric": 0.0003791913331951946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.742084019000686
        },
        {
          "epoch": 209,
          "updates": 836,
          "samples_seen": 214016,
          "online_loss": 0.004300035710912198,
          "train_loss": 0.0048277066089212894,
          "validation_loss": 0.003672892227768898,
          "train_metric": 0.0004862289351876825,
          "validation_metric": 0.00036992024979554117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.7750319030001265
        },
        {
          "epoch": 210,
          "updates": 840,
          "samples_seen": 215040,
          "online_loss": 0.004419962759129703,
          "train_loss": 0.0040053995326161385,
          "validation_loss": 0.00353936105966568,
          "train_metric": 0.00040340921259485185,
          "validation_metric": 0.0003564715152606368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.809118475001014
        },
        {
          "epoch": 211,
          "updates": 844,
          "samples_seen": 216064,
          "online_loss": 0.003948063938878477,
          "train_loss": 0.003985250368714333,
          "validation_loss": 0.0032258876599371433,
          "train_metric": 0.00040137977339327335,
          "validation_metric": 0.00032489956356585026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.845575416000429
        },
        {
          "epoch": 212,
          "updates": 848,
          "samples_seen": 217088,
          "online_loss": 0.004004420537967235,
          "train_loss": 0.003877284936606884,
          "validation_loss": 0.0033041329588741064,
          "train_metric": 0.00039050591294653714,
          "validation_metric": 0.0003327801823616028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.879471803000342
        },
        {
          "epoch": 213,
          "updates": 852,
          "samples_seen": 218112,
          "online_loss": 0.003925953875295818,
          "train_loss": 0.0040566748939454556,
          "validation_loss": 0.003543236292898655,
          "train_metric": 0.0004085734544787556,
          "validation_metric": 0.00035686182673089206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.913681085999997
        },
        {
          "epoch": 214,
          "updates": 856,
          "samples_seen": 219136,
          "online_loss": 0.004098422359675169,
          "train_loss": 0.004818212706595659,
          "validation_loss": 0.003693744307383895,
          "train_metric": 0.0004852727288380265,
          "validation_metric": 0.00037202041130512953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.948044082000706
        },
        {
          "epoch": 215,
          "updates": 860,
          "samples_seen": 220160,
          "online_loss": 0.004810441751033068,
          "train_loss": 0.0039483290165662766,
          "validation_loss": 0.0034208449069410563,
          "train_metric": 0.0003976612351834774,
          "validation_metric": 0.0003445349575486034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.982099693999771
        },
        {
          "epoch": 216,
          "updates": 864,
          "samples_seen": 221184,
          "online_loss": 0.003979547473136336,
          "train_loss": 0.0038348373491317034,
          "validation_loss": 0.0030732888262718916,
          "train_metric": 0.00038623082218691707,
          "validation_metric": 0.0003095304127782583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.017788199000279
        },
        {
          "epoch": 217,
          "updates": 868,
          "samples_seen": 222208,
          "online_loss": 0.003950073732994497,
          "train_loss": 0.003740770509466529,
          "validation_loss": 0.003098241752013564,
          "train_metric": 0.00037675671046599746,
          "validation_metric": 0.0003120434994343668,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.051774290001049
        },
        {
          "epoch": 218,
          "updates": 872,
          "samples_seen": 223232,
          "online_loss": 0.004270628502126783,
          "train_loss": 0.004889920353889465,
          "validation_loss": 0.004304172936826944,
          "train_metric": 0.000492494844365865,
          "validation_metric": 0.000433500565122813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.084275250001156
        },
        {
          "epoch": 219,
          "updates": 876,
          "samples_seen": 224256,
          "online_loss": 0.004363962507341057,
          "train_loss": 0.005544198676943779,
          "validation_loss": 0.004136689007282257,
          "train_metric": 0.0005583914462476969,
          "validation_metric": 0.0004166322178207338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.117826902000161
        },
        {
          "epoch": 220,
          "updates": 880,
          "samples_seen": 225280,
          "online_loss": 0.004355081589892507,
          "train_loss": 0.004594212863594294,
          "validation_loss": 0.004001360386610031,
          "train_metric": 0.0004627122834790498,
          "validation_metric": 0.00040300231194123626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.152006092001102
        },
        {
          "epoch": 221,
          "updates": 884,
          "samples_seen": 226304,
          "online_loss": 0.004164917569141835,
          "train_loss": 0.0039096553809940815,
          "validation_loss": 0.0030632975976914167,
          "train_metric": 0.0003937661531381309,
          "validation_metric": 0.00030852408963255584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.188440398000239
        },
        {
          "epoch": 222,
          "updates": 888,
          "samples_seen": 227328,
          "online_loss": 0.003655898501165211,
          "train_loss": 0.0036628753878176212,
          "validation_loss": 0.003195817582309246,
          "train_metric": 0.0003689114237204194,
          "validation_metric": 0.00032187101896852255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.222608993000904
        },
        {
          "epoch": 223,
          "updates": 892,
          "samples_seen": 228352,
          "online_loss": 0.0038181909476406872,
          "train_loss": 0.004038048908114433,
          "validation_loss": 0.0031055540312081575,
          "train_metric": 0.00040669747977517545,
          "validation_metric": 0.0003127800300717354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.257283050001206
        },
        {
          "epoch": 224,
          "updates": 896,
          "samples_seen": 229376,
          "online_loss": 0.004233073501382023,
          "train_loss": 0.00351416552439332,
          "validation_loss": 0.002983933547511697,
          "train_metric": 0.00035393383586779237,
          "validation_metric": 0.0003005308681167662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.292968468000254
        },
        {
          "epoch": 225,
          "updates": 900,
          "samples_seen": 230400,
          "online_loss": 0.004314176563639194,
          "train_loss": 0.00439703231677413,
          "validation_loss": 0.0038453415036201477,
          "train_metric": 0.000442852993728593,
          "validation_metric": 0.00038728874642401934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.358650775000569
        },
        {
          "epoch": 226,
          "updates": 904,
          "samples_seen": 231424,
          "online_loss": 0.005277529708109796,
          "train_loss": 0.00450705923140049,
          "validation_loss": 0.0032734896522015333,
          "train_metric": 0.0004539345682132989,
          "validation_metric": 0.00032969392486847937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.384113639000134
        },
        {
          "epoch": 227,
          "updates": 908,
          "samples_seen": 232448,
          "online_loss": 0.004495802684687078,
          "train_loss": 0.0035022743977606297,
          "validation_loss": 0.002993556670844555,
          "train_metric": 0.0003527362714521587,
          "validation_metric": 0.00030150002567097545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.398585623001054
        },
        {
          "epoch": 228,
          "updates": 912,
          "samples_seen": 233472,
          "online_loss": 0.0034204035182483494,
          "train_loss": 0.0036218143068253994,
          "validation_loss": 0.0028961964417248964,
          "train_metric": 0.0003647758858278394,
          "validation_metric": 0.0002916943049058318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.4164852000012615
        },
        {
          "epoch": 229,
          "updates": 916,
          "samples_seen": 234496,
          "online_loss": 0.0037375325337052345,
          "train_loss": 0.0034256128128618,
          "validation_loss": 0.002776659792289138,
          "train_metric": 0.00034501519985497,
          "validation_metric": 0.00027965501067228615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.436707230999673
        },
        {
          "epoch": 230,
          "updates": 920,
          "samples_seen": 235520,
          "online_loss": 0.003567790612578392,
          "train_loss": 0.003798827761784196,
          "validation_loss": 0.0034352485090494156,
          "train_metric": 0.00038260401925072074,
          "validation_metric": 0.0003459856379777193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.451004164000551
        },
        {
          "epoch": 231,
          "updates": 924,
          "samples_seen": 236544,
          "online_loss": 0.0037791167269460857,
          "train_loss": 0.0036776813212782145,
          "validation_loss": 0.002943763742223382,
          "train_metric": 0.00037040261668153107,
          "validation_metric": 0.00029648508643731475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.466244799999913
        },
        {
          "epoch": 232,
          "updates": 928,
          "samples_seen": 237568,
          "online_loss": 0.0036581321619451046,
          "train_loss": 0.0036244732327759266,
          "validation_loss": 0.003209517104551196,
          "train_metric": 0.0003650436410680413,
          "validation_metric": 0.00032325080246664584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.477047940001285
        },
        {
          "epoch": 233,
          "updates": 932,
          "samples_seen": 238592,
          "online_loss": 0.003431596211157739,
          "train_loss": 0.0033285771496593952,
          "validation_loss": 0.0026963562704622746,
          "train_metric": 0.00033524210448376834,
          "validation_metric": 0.0002715671434998512,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.492937996001274
        },
        {
          "epoch": 234,
          "updates": 936,
          "samples_seen": 239616,
          "online_loss": 0.00406141096027568,
          "train_loss": 0.004353757482022047,
          "validation_loss": 0.003041456453502178,
          "train_metric": 0.0004384945204947144,
          "validation_metric": 0.00030632439302280545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.515305646000343
        },
        {
          "epoch": 235,
          "updates": 940,
          "samples_seen": 240640,
          "online_loss": 0.003924466611351818,
          "train_loss": 0.0032164014410227537,
          "validation_loss": 0.0027292072772979736,
          "train_metric": 0.0003239441430196166,
          "validation_metric": 0.0002748757542576641,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.526958282000123
        },
        {
          "epoch": 236,
          "updates": 944,
          "samples_seen": 241664,
          "online_loss": 0.003829049994237721,
          "train_loss": 0.0033992493990808725,
          "validation_loss": 0.0027557355351746082,
          "train_metric": 0.0003423600282985717,
          "validation_metric": 0.0002775475732050836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.541572357000405
        },
        {
          "epoch": 237,
          "updates": 948,
          "samples_seen": 242688,
          "online_loss": 0.00344923761440441,
          "train_loss": 0.003174837678670883,
          "validation_loss": 0.0026958927046507597,
          "train_metric": 0.0003197579935658723,
          "validation_metric": 0.0002715204609557986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.55697893100114
        },
        {
          "epoch": 238,
          "updates": 952,
          "samples_seen": 243712,
          "online_loss": 0.0031801769509911537,
          "train_loss": 0.003241788363084197,
          "validation_loss": 0.0025808445643633604,
          "train_metric": 0.00032650106004439294,
          "validation_metric": 0.00025993320741690695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.572724990999632
        },
        {
          "epoch": 239,
          "updates": 956,
          "samples_seen": 244736,
          "online_loss": 0.0034214603947475553,
          "train_loss": 0.0031832088716328144,
          "validation_loss": 0.0025971292052417994,
          "train_metric": 0.0003206011315342039,
          "validation_metric": 0.0002615733537822962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.589088882999931
        },
        {
          "epoch": 240,
          "updates": 960,
          "samples_seen": 245760,
          "online_loss": 0.0037572612054646015,
          "train_loss": 0.00387355824932456,
          "validation_loss": 0.003471246687695384,
          "train_metric": 0.00039013061905279756,
          "validation_metric": 0.0003496112476568669,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.603861211000549
        }
      ],
      "best": {
        "epoch": 238,
        "updates": 952
      },
      "best_validation_metric": 0.00025993320741690695,
      "config": {
        "label": "Batch 256",
        "batch": 256,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.0161224442134841
    },
    "449a997d9d64_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.946105096489191,
          "train_loss": 0.9352960586547852,
          "validation_loss": 0.6764616370201111,
          "train_metric": 0.09419960528612137,
          "validation_metric": 0.0681307464838028,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.012980808000065736
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.9253312796354294,
          "train_loss": 0.9140766263008118,
          "validation_loss": 0.660655677318573,
          "train_metric": 0.09206245094537735,
          "validation_metric": 0.06653882563114166,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.025754062000487465
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.9034468568861485,
          "train_loss": 0.8917216062545776,
          "validation_loss": 0.644091010093689,
          "train_metric": 0.08981093764305115,
          "validation_metric": 0.06487048417329788,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.03832339700056764
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.880322877317667,
          "train_loss": 0.8672615885734558,
          "validation_loss": 0.6258240938186646,
          "train_metric": 0.08734741806983948,
          "validation_metric": 0.06303071230649948,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.05100523799956136
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.8550155833363533,
          "train_loss": 0.8407058715820312,
          "validation_loss": 0.6061688661575317,
          "train_metric": 0.08467280864715576,
          "validation_metric": 0.061051104217767715,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.06593382600112818
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.8271960057318211,
          "train_loss": 0.8117698431015015,
          "validation_loss": 0.584735631942749,
          "train_metric": 0.08175848424434662,
          "validation_metric": 0.05889243260025978,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.07999930100049824
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.7968348003923893,
          "train_loss": 0.7799336910247803,
          "validation_loss": 0.5611906051635742,
          "train_metric": 0.07855206727981567,
          "validation_metric": 0.056521061807870865,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.09263008900052228
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.7633976601064205,
          "train_loss": 0.745169997215271,
          "validation_loss": 0.5356494188308716,
          "train_metric": 0.07505078613758087,
          "validation_metric": 0.053948648273944855,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.1054695510010788
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.7273965962231159,
          "train_loss": 0.7069858312606812,
          "validation_loss": 0.5075681805610657,
          "train_metric": 0.0712050274014473,
          "validation_metric": 0.05112041160464287,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.11782945100094366
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.6881908923387527,
          "train_loss": 0.6661955118179321,
          "validation_loss": 0.47744274139404297,
          "train_metric": 0.0670967698097229,
          "validation_metric": 0.04808628559112549,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.13097853699946427
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.6462142802774906,
          "train_loss": 0.6231247186660767,
          "validation_loss": 0.4456397294998169,
          "train_metric": 0.06275884062051773,
          "validation_metric": 0.044883206486701965,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.14333949600040796
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.6018033735454082,
          "train_loss": 0.5787760615348816,
          "validation_loss": 0.41301390528678894,
          "train_metric": 0.058292206376791,
          "validation_metric": 0.04159724339842796,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.15599596300125995
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.5570604894310236,
          "train_loss": 0.5322027802467346,
          "validation_loss": 0.3786316215991974,
          "train_metric": 0.05360151454806328,
          "validation_metric": 0.038134388625621796,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.1678902100011328
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.5106495060026646,
          "train_loss": 0.4861220419406891,
          "validation_loss": 0.3446585536003113,
          "train_metric": 0.04896043241024017,
          "validation_metric": 0.03471274673938751,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.18044166400068207
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.4648212846368551,
          "train_loss": 0.44131216406822205,
          "validation_loss": 0.3116416335105896,
          "train_metric": 0.04444734752178192,
          "validation_metric": 0.03138739988207817,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.19257491300049878
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.4209738355129957,
          "train_loss": 0.3984934389591217,
          "validation_loss": 0.2800578773021698,
          "train_metric": 0.04013479873538017,
          "validation_metric": 0.02820640429854393,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.20489517500027432
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.3799182791262865,
          "train_loss": 0.35889139771461487,
          "validation_loss": 0.25103121995925903,
          "train_metric": 0.03614622354507446,
          "validation_metric": 0.02528294548392296,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.21733399600088887
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.3419335074722767,
          "train_loss": 0.324411004781723,
          "validation_loss": 0.2256629765033722,
          "train_metric": 0.032673489302396774,
          "validation_metric": 0.022727951407432556,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.23006476900081907
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.3094523632898927,
          "train_loss": 0.2940099239349365,
          "validation_loss": 0.20370469987392426,
          "train_metric": 0.029611604288220406,
          "validation_metric": 0.020516391843557358,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.24287745100082248
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.2812332771718502,
          "train_loss": 0.26847177743911743,
          "validation_loss": 0.18569433689117432,
          "train_metric": 0.02703949436545372,
          "validation_metric": 0.01870245486497879,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.2557845980009006
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.2586441105231643,
          "train_loss": 0.24650034308433533,
          "validation_loss": 0.17074669897556305,
          "train_metric": 0.02482661046087742,
          "validation_metric": 0.017196983098983765,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.2710020070007886
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.2387241991236806,
          "train_loss": 0.2296541929244995,
          "validation_loss": 0.15967348217964172,
          "train_metric": 0.02312992885708809,
          "validation_metric": 0.016081729903817177,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.2843764210010704
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.22300768736749887,
          "train_loss": 0.21635118126869202,
          "validation_loss": 0.15146298706531525,
          "train_metric": 0.021790098398923874,
          "validation_metric": 0.015254797413945198,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.2968663720002951
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.21063609002158046,
          "train_loss": 0.20532751083374023,
          "validation_loss": 0.14526841044425964,
          "train_metric": 0.020679831504821777,
          "validation_metric": 0.014630901627242565,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.3092897639999137
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.2005100236274302,
          "train_loss": 0.1962990015745163,
          "validation_loss": 0.14052341878414154,
          "train_metric": 0.019770517945289612,
          "validation_metric": 0.014153003692626953,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.3216745629997604
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.1924244426190853,
          "train_loss": 0.18857772648334503,
          "validation_loss": 0.13688480854034424,
          "train_metric": 0.01899285800755024,
          "validation_metric": 0.013786538504064083,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.33396389900008217
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.18554858956485987,
          "train_loss": 0.18221957981586456,
          "validation_loss": 0.13405370712280273,
          "train_metric": 0.018352489918470383,
          "validation_metric": 0.013501398265361786,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.34652478899988637
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.17976062931120396,
          "train_loss": 0.17687740921974182,
          "validation_loss": 0.13192963600158691,
          "train_metric": 0.01781444624066353,
          "validation_metric": 0.013287471607327461,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.37233695600116334
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.1747593334876001,
          "train_loss": 0.17230486869812012,
          "validation_loss": 0.1300993710756302,
          "train_metric": 0.01735391467809677,
          "validation_metric": 0.013103132136166096,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.3848113820004073
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.1704508033581078,
          "train_loss": 0.1681842803955078,
          "validation_loss": 0.12842236459255219,
          "train_metric": 0.01693890616297722,
          "validation_metric": 0.012934230268001556,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.3972221920012089
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.16647262731567025,
          "train_loss": 0.16441485285758972,
          "validation_loss": 0.12666639685630798,
          "train_metric": 0.016559261828660965,
          "validation_metric": 0.01275737676769495,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.4095194680012355
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.1628264202736318,
          "train_loss": 0.1609763354063034,
          "validation_loss": 0.12491939216852188,
          "train_metric": 0.0162129458039999,
          "validation_metric": 0.012581422924995422,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.42212164600096
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.1595021546818316,
          "train_loss": 0.15773004293441772,
          "validation_loss": 0.1233525276184082,
          "train_metric": 0.015885991975665092,
          "validation_metric": 0.012423614040017128,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.43417731000045023
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.15640347730368376,
          "train_loss": 0.1547771841287613,
          "validation_loss": 0.1220046654343605,
          "train_metric": 0.015588589012622833,
          "validation_metric": 0.012287862598896027,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.44646823199946084
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.15353687619790435,
          "train_loss": 0.15195676684379578,
          "validation_loss": 0.12061530351638794,
          "train_metric": 0.01530452910810709,
          "validation_metric": 0.012147931382060051,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.4585597469995264
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.15078453812748194,
          "train_loss": 0.14939436316490173,
          "validation_loss": 0.11926774680614471,
          "train_metric": 0.015046454966068268,
          "validation_metric": 0.012012209743261337,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.4712185120006325
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.14828788908198476,
          "train_loss": 0.14691829681396484,
          "validation_loss": 0.11793196946382523,
          "train_metric": 0.014797072857618332,
          "validation_metric": 0.011877675540745258,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.48420259700105817
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.1458951886743307,
          "train_loss": 0.14460037648677826,
          "validation_loss": 0.11654242873191833,
          "train_metric": 0.01456362009048462,
          "validation_metric": 0.011737726628780365,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.4969653690004634
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.1436219154857099,
          "train_loss": 0.1424161046743393,
          "validation_loss": 0.11533178389072418,
          "train_metric": 0.014343627728521824,
          "validation_metric": 0.011615796014666557,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.509256314000595
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.14149455865845084,
          "train_loss": 0.14038583636283875,
          "validation_loss": 0.1140904352068901,
          "train_metric": 0.014139147475361824,
          "validation_metric": 0.011490771546959877,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.5979519789998449
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.13949813973158598,
          "train_loss": 0.1384384036064148,
          "validation_loss": 0.11286703497171402,
          "train_metric": 0.013943011872470379,
          "validation_metric": 0.011367554776370525,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.610556928000733
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.13764331047423184,
          "train_loss": 0.13655918836593628,
          "validation_loss": 0.11175912618637085,
          "train_metric": 0.013753741979598999,
          "validation_metric": 0.011255969293415546,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6228710840005078
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.1357967322692275,
          "train_loss": 0.13486027717590332,
          "validation_loss": 0.11090188473463058,
          "train_metric": 0.013582633808255196,
          "validation_metric": 0.011169633828103542,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6353129239996633
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.1340988678857684,
          "train_loss": 0.13317056000232697,
          "validation_loss": 0.10969143360853195,
          "train_metric": 0.01341245137155056,
          "validation_metric": 0.011047720909118652,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6475784040012513
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.13242072984576225,
          "train_loss": 0.13156260550022125,
          "validation_loss": 0.10869282484054565,
          "train_metric": 0.013250503689050674,
          "validation_metric": 0.010947142727673054,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6597215109995886
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.13084430247545242,
          "train_loss": 0.13001355528831482,
          "validation_loss": 0.10764186829328537,
          "train_metric": 0.013094489462673664,
          "validation_metric": 0.010841293260455132,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6730449030001182
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.12935680616647005,
          "train_loss": 0.12851965427398682,
          "validation_loss": 0.10660363733768463,
          "train_metric": 0.01294402964413166,
          "validation_metric": 0.010736726224422455,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6853965650007012
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.12790643400512636,
          "train_loss": 0.12711435556411743,
          "validation_loss": 0.10559078305959702,
          "train_metric": 0.012802490964531898,
          "validation_metric": 0.010634717531502247,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6982854070010944
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.1265087763313204,
          "train_loss": 0.12577316164970398,
          "validation_loss": 0.10472969710826874,
          "train_metric": 0.012667413800954819,
          "validation_metric": 0.010547991842031479,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7104938049997145
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.12515744380652905,
          "train_loss": 0.12439782917499542,
          "validation_loss": 0.1036882996559143,
          "train_metric": 0.01252889446914196,
          "validation_metric": 0.010443105362355709,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7228282209998724
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.12377710617147386,
          "train_loss": 0.12314017117023468,
          "validation_loss": 0.10277373343706131,
          "train_metric": 0.012402227148413658,
          "validation_metric": 0.010350994765758514,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7346649510000134
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.12255483912304044,
          "train_loss": 0.12186127156019211,
          "validation_loss": 0.10177859663963318,
          "train_metric": 0.012273420579731464,
          "validation_metric": 0.01025076862424612,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.746530017000623
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.12128749792464077,
          "train_loss": 0.12063225358724594,
          "validation_loss": 0.10086491703987122,
          "train_metric": 0.012149639427661896,
          "validation_metric": 0.010158744640648365,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7581740260011429
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.12009601667523384,
          "train_loss": 0.11944391578435898,
          "validation_loss": 0.10000042617321014,
          "train_metric": 0.012029952369630337,
          "validation_metric": 0.010071677155792713,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7702475340011006
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.11903398064896464,
          "train_loss": 0.11821748316287994,
          "validation_loss": 0.09884102642536163,
          "train_metric": 0.01190643198788166,
          "validation_metric": 0.009954906068742275,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7823060840000835
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.11770546925254166,
          "train_loss": 0.1170528382062912,
          "validation_loss": 0.09802445769309998,
          "train_metric": 0.011789131909608841,
          "validation_metric": 0.009872663766145706,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7942501770012314
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.1165726208128035,
          "train_loss": 0.1159338653087616,
          "validation_loss": 0.09718569368124008,
          "train_metric": 0.011676433496177197,
          "validation_metric": 0.00978818628937006,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8129961630002072
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.11547909560613334,
          "train_loss": 0.11480632424354553,
          "validation_loss": 0.09618398547172546,
          "train_metric": 0.011562872678041458,
          "validation_metric": 0.00968729704618454,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8252905820008891
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.11436443170532584,
          "train_loss": 0.11369379609823227,
          "validation_loss": 0.09530041366815567,
          "train_metric": 0.01145082339644432,
          "validation_metric": 0.00959830917418003,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8375206559994695
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.11322734132409096,
          "train_loss": 0.11262568831443787,
          "validation_loss": 0.09440384060144424,
          "train_metric": 0.011343245394527912,
          "validation_metric": 0.009508009068667889,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8499394709997432
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.11214133026078343,
          "train_loss": 0.11153344064950943,
          "validation_loss": 0.09349074959754944,
          "train_metric": 0.011233238503336906,
          "validation_metric": 0.009416045621037483,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8629684469997301
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.11105814948678017,
          "train_loss": 0.11046973615884781,
          "validation_loss": 0.0925757959485054,
          "train_metric": 0.011126106604933739,
          "validation_metric": 0.009323895908892155,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.875918993000596
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.11001077806577086,
          "train_loss": 0.10940254479646683,
          "validation_loss": 0.09163275361061096,
          "train_metric": 0.011018624529242516,
          "validation_metric": 0.009228914976119995,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8892913950003276
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.10894224047660828,
          "train_loss": 0.10835419595241547,
          "validation_loss": 0.0907750353217125,
          "train_metric": 0.01091303862631321,
          "validation_metric": 0.009142527356743813,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9020153130004473
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.10792326764203608,
          "train_loss": 0.1072804182767868,
          "validation_loss": 0.08981337398290634,
          "train_metric": 0.010804890654981136,
          "validation_metric": 0.009045673534274101,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9143277969997143
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.10682977945543826,
          "train_loss": 0.10627035051584244,
          "validation_loss": 0.08891620486974716,
          "train_metric": 0.010703161358833313,
          "validation_metric": 0.008955313824117184,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9265211010006169
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.10583128919824958,
          "train_loss": 0.10523271560668945,
          "validation_loss": 0.08806435018777847,
          "train_metric": 0.010598654858767986,
          "validation_metric": 0.008869519457221031,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9393587840004329
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.10483254096470773,
          "train_loss": 0.10421008616685867,
          "validation_loss": 0.08715417981147766,
          "train_metric": 0.010495658963918686,
          "validation_metric": 0.00877784937620163,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9517787149998185
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.10397868370637298,
          "train_loss": 0.1031460165977478,
          "validation_loss": 0.08610253781080246,
          "train_metric": 0.010388487949967384,
          "validation_metric": 0.008671932853758335,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9645155769994744
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.10269632865674794,
          "train_loss": 0.10213808715343475,
          "validation_loss": 0.08525387197732925,
          "train_metric": 0.010286974720656872,
          "validation_metric": 0.008586456999182701,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9773175330010417
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.10170382191427052,
          "train_loss": 0.10114112496376038,
          "validation_loss": 0.08438218384981155,
          "train_metric": 0.010186564177274704,
          "validation_metric": 0.008498664945363998,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9903008279998176
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.10069943103007972,
          "train_loss": 0.10014300048351288,
          "validation_loss": 0.08350662887096405,
          "train_metric": 0.01008603535592556,
          "validation_metric": 0.008410482667386532,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0024984259998746
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.09973698994144797,
          "train_loss": 0.09912193566560745,
          "validation_loss": 0.08257240802049637,
          "train_metric": 0.009983197785913944,
          "validation_metric": 0.008316390216350555,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0145150270000158
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.09873620141297579,
          "train_loss": 0.09814515709877014,
          "validation_loss": 0.08171972632408142,
          "train_metric": 0.009884821251034737,
          "validation_metric": 0.008230511099100113,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0271281670011376
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.09776388946920633,
          "train_loss": 0.09712322056293488,
          "validation_loss": 0.08074228465557098,
          "train_metric": 0.009781893342733383,
          "validation_metric": 0.008132066577672958,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.039400695999575
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.09671122138388455,
          "train_loss": 0.09616592526435852,
          "validation_loss": 0.07991526275873184,
          "train_metric": 0.009685480035841465,
          "validation_metric": 0.008048771880567074,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.051876710000215
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.09576002252288163,
          "train_loss": 0.09521771967411041,
          "validation_loss": 0.0790753960609436,
          "train_metric": 0.009589980356395245,
          "validation_metric": 0.007964183576405048,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0641054219995567
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.09479762404225767,
          "train_loss": 0.0942293033003807,
          "validation_loss": 0.07814589887857437,
          "train_metric": 0.009490430355072021,
          "validation_metric": 0.00787056889384985,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0760861540002225
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.09381649363785982,
          "train_loss": 0.09328161925077438,
          "validation_loss": 0.07729966938495636,
          "train_metric": 0.00939498282968998,
          "validation_metric": 0.007785339839756489,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.091235632000462
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.09290410205721855,
          "train_loss": 0.09233670681715012,
          "validation_loss": 0.0764271691441536,
          "train_metric": 0.009299816563725471,
          "validation_metric": 0.007697464432567358,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1034643620005227
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.09194667707197368,
          "train_loss": 0.09140560775995255,
          "validation_loss": 0.07563844323158264,
          "train_metric": 0.00920603796839714,
          "validation_metric": 0.007618027273565531,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.116763019001155
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.09100990393199027,
          "train_loss": 0.09046974778175354,
          "validation_loss": 0.07476503401994705,
          "train_metric": 0.009111782535910606,
          "validation_metric": 0.0075300601311028,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1293404380012362
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.09007471334189177,
          "train_loss": 0.08954597264528275,
          "validation_loss": 0.07392662763595581,
          "train_metric": 0.009018742479383945,
          "validation_metric": 0.007445618510246277,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1419455380000727
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.08915867703035474,
          "train_loss": 0.08863513171672821,
          "validation_loss": 0.07312441617250443,
          "train_metric": 0.008927006274461746,
          "validation_metric": 0.00736482348293066,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1545590400000947
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.08833083137869835,
          "train_loss": 0.08775748312473297,
          "validation_loss": 0.07232673466205597,
          "train_metric": 0.008838609792292118,
          "validation_metric": 0.007284483406692743,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1669144580009743
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.08742315997369587,
          "train_loss": 0.08690022677183151,
          "validation_loss": 0.07159111648797989,
          "train_metric": 0.00875227339565754,
          "validation_metric": 0.0072103943675756454,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.179198918000111
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.08652598003391176,
          "train_loss": 0.0860295295715332,
          "validation_loss": 0.07082974165678024,
          "train_metric": 0.008664579130709171,
          "validation_metric": 0.007133712992072105,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.191223521000211
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.08571912022307515,
          "train_loss": 0.08516069501638412,
          "validation_loss": 0.0700128972530365,
          "train_metric": 0.008577072992920876,
          "validation_metric": 0.007051443215459585,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2037342020012147
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.08482099371030927,
          "train_loss": 0.08434662967920303,
          "validation_loss": 0.06931575387716293,
          "train_metric": 0.008495083078742027,
          "validation_metric": 0.006981228478252888,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2162920470000245
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.08407376380637288,
          "train_loss": 0.08350052684545517,
          "validation_loss": 0.06851791590452194,
          "train_metric": 0.008409867994487286,
          "validation_metric": 0.006900873500853777,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2286475939999946
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.08317274949513376,
          "train_loss": 0.08268940448760986,
          "validation_loss": 0.0678214579820633,
          "train_metric": 0.00832817330956459,
          "validation_metric": 0.00683072954416275,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2408231170011277
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.08238106570206583,
          "train_loss": 0.08190880715847015,
          "validation_loss": 0.06711762398481369,
          "train_metric": 0.008249555714428425,
          "validation_metric": 0.006759841926395893,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2532745970001997
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0816012704744935,
          "train_loss": 0.08118052035570145,
          "validation_loss": 0.06652076542377472,
          "train_metric": 0.00817620474845171,
          "validation_metric": 0.006699728313833475,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.266112018000058
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.08084044000133872,
          "train_loss": 0.08043859899044037,
          "validation_loss": 0.06584271788597107,
          "train_metric": 0.008101481013000011,
          "validation_metric": 0.006631437689065933,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2813571710012184
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0801180531270802,
          "train_loss": 0.07968392968177795,
          "validation_loss": 0.06518911570310593,
          "train_metric": 0.008025473915040493,
          "validation_metric": 0.006565609015524387,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.293786206000732
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.07946192636154592,
          "train_loss": 0.07897809892892838,
          "validation_loss": 0.06453653424978256,
          "train_metric": 0.007954385131597519,
          "validation_metric": 0.006499883718788624,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.306123746000594
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.07871628855355084,
          "train_loss": 0.07831528037786484,
          "validation_loss": 0.06397704780101776,
          "train_metric": 0.007887627929449081,
          "validation_metric": 0.006443534977734089,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.319952691999788
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.07800122257322073,
          "train_loss": 0.07763585448265076,
          "validation_loss": 0.0633859634399414,
          "train_metric": 0.007819199934601784,
          "validation_metric": 0.006384002044796944,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3323211139995692
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.07734275166876614,
          "train_loss": 0.07698151469230652,
          "validation_loss": 0.06282014399766922,
          "train_metric": 0.007753295823931694,
          "validation_metric": 0.006327015347778797,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3449952430000849
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.07681320118717849,
          "train_loss": 0.07632710784673691,
          "validation_loss": 0.062265798449516296,
          "train_metric": 0.007687386590987444,
          "validation_metric": 0.006271182559430599,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3572573300007207
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.07606801006477326,
          "train_loss": 0.07572828233242035,
          "validation_loss": 0.06172575801610947,
          "train_metric": 0.0076270755380392075,
          "validation_metric": 0.006216791924089193,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.369188275999477
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.07550310390070081,
          "train_loss": 0.07514006644487381,
          "validation_loss": 0.061258621513843536,
          "train_metric": 0.007567832246422768,
          "validation_metric": 0.006169744301587343,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3817242400000396
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.07493271515704691,
          "train_loss": 0.07457247376441956,
          "validation_loss": 0.06078370660543442,
          "train_metric": 0.007510667201131582,
          "validation_metric": 0.006121912505477667,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.393963533000715
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0744134047999978,
          "train_loss": 0.07402420043945312,
          "validation_loss": 0.06032115966081619,
          "train_metric": 0.007455445826053619,
          "validation_metric": 0.006075326818972826,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4063500740012387
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.07383275323081762,
          "train_loss": 0.0735306367278099,
          "validation_loss": 0.059917718172073364,
          "train_metric": 0.007405735552310944,
          "validation_metric": 0.006034693215042353,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4193057110005611
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0733522892696783,
          "train_loss": 0.07300805300474167,
          "validation_loss": 0.059488940984010696,
          "train_metric": 0.007353104185312986,
          "validation_metric": 0.00599150825291872,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4318888400011929
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.07281509775202721,
          "train_loss": 0.0725410133600235,
          "validation_loss": 0.059086039662361145,
          "train_metric": 0.0073060644790530205,
          "validation_metric": 0.005950930528342724,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4443375499995454
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.07235001330263913,
          "train_loss": 0.07206297665834427,
          "validation_loss": 0.05867169052362442,
          "train_metric": 0.007257918827235699,
          "validation_metric": 0.005909197963774204,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4569627769997169
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.07191070925910026,
          "train_loss": 0.07160866260528564,
          "validation_loss": 0.05832282453775406,
          "train_metric": 0.007212162017822266,
          "validation_metric": 0.005874061491340399,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4720204309996916
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.07147087552584708,
          "train_loss": 0.0711890384554863,
          "validation_loss": 0.057919010519981384,
          "train_metric": 0.00716989953070879,
          "validation_metric": 0.005833391100168228,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.575406424999528
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.07104702363722026,
          "train_loss": 0.07077883929014206,
          "validation_loss": 0.05766988545656204,
          "train_metric": 0.007128585129976273,
          "validation_metric": 0.0058083008043468,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.5998424600002181
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.07062760111875832,
          "train_loss": 0.07037486881017685,
          "validation_loss": 0.057336580008268356,
          "train_metric": 0.007087899371981621,
          "validation_metric": 0.005774730816483498,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6142174579999846
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0702271448681131,
          "train_loss": 0.06998693943023682,
          "validation_loss": 0.05700527876615524,
          "train_metric": 0.007048828061670065,
          "validation_metric": 0.005741363391280174,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.626699008000287
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.069874252891168,
          "train_loss": 0.069632887840271,
          "validation_loss": 0.05674433708190918,
          "train_metric": 0.007013170048594475,
          "validation_metric": 0.0057150823995471,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6405709160007973
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.06948047201149166,
          "train_loss": 0.06927688419818878,
          "validation_loss": 0.05643337592482567,
          "train_metric": 0.0069773150607943535,
          "validation_metric": 0.005683762487024069,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6583156180004153
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.06918690435122699,
          "train_loss": 0.06892332434654236,
          "validation_loss": 0.0561542883515358,
          "train_metric": 0.006941705942153931,
          "validation_metric": 0.005655654240399599,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6732095410006878
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.06887579604517668,
          "train_loss": 0.06859791278839111,
          "validation_loss": 0.05591359734535217,
          "train_metric": 0.006908930838108063,
          "validation_metric": 0.005631412845104933,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6880431990011857
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.06848312169313431,
          "train_loss": 0.06830054521560669,
          "validation_loss": 0.055736713111400604,
          "train_metric": 0.0068789818324148655,
          "validation_metric": 0.005613597109913826,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7007256899996719
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.06819812441244721,
          "train_loss": 0.06797449290752411,
          "validation_loss": 0.05548000708222389,
          "train_metric": 0.006846142001450062,
          "validation_metric": 0.005587743129581213,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7138026900011027
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.06788154353853315,
          "train_loss": 0.06768669933080673,
          "validation_loss": 0.05524030327796936,
          "train_metric": 0.006817156914621592,
          "validation_metric": 0.005563600920140743,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7264065990002564
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.06760768056847155,
          "train_loss": 0.067418172955513,
          "validation_loss": 0.05501139536499977,
          "train_metric": 0.006790111307054758,
          "validation_metric": 0.005540546029806137,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.73848165399977
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.06732908019330353,
          "train_loss": 0.06712609529495239,
          "validation_loss": 0.05482743680477142,
          "train_metric": 0.006760695017874241,
          "validation_metric": 0.005522019229829311,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7505348440008675
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.06724195065908134,
          "train_loss": 0.06685289740562439,
          "validation_loss": 0.05456693843007088,
          "train_metric": 0.006733178626745939,
          "validation_metric": 0.005495782475918531,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7629145629998675
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.06681277765892446,
          "train_loss": 0.06660061329603195,
          "validation_loss": 0.054448552429676056,
          "train_metric": 0.006707770749926567,
          "validation_metric": 0.005483859218657017,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7749284990004526
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.06653313129208982,
          "train_loss": 0.0663495808839798,
          "validation_loss": 0.054278016090393066,
          "train_metric": 0.006682485342025757,
          "validation_metric": 0.0054666828364133835,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7870547490001627
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.06642242288216949,
          "train_loss": 0.06613551080226898,
          "validation_loss": 0.054112061858177185,
          "train_metric": 0.006660927087068558,
          "validation_metric": 0.0054499683901667595,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.799141375000545
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.06617018685210496,
          "train_loss": 0.06587109714746475,
          "validation_loss": 0.05380482226610184,
          "train_metric": 0.0066342949867248535,
          "validation_metric": 0.005419024731963873,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.812584199000412
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0657809159019962,
          "train_loss": 0.06564411520957947,
          "validation_loss": 0.05368620902299881,
          "train_metric": 0.006611434742808342,
          "validation_metric": 0.00540707865729928,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8249111770001036
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.06559656630270183,
          "train_loss": 0.06542057543992996,
          "validation_loss": 0.05352872237563133,
          "train_metric": 0.0065889195539057255,
          "validation_metric": 0.005391217302531004,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8371122910011763
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.06537043361458927,
          "train_loss": 0.06518369913101196,
          "validation_loss": 0.05338975042104721,
          "train_metric": 0.006565063260495663,
          "validation_metric": 0.005377220455557108,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.849466902000131
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.06515136477537453,
          "train_loss": 0.0649709403514862,
          "validation_loss": 0.0532553531229496,
          "train_metric": 0.00654363539069891,
          "validation_metric": 0.0053636846132576466,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8615578019998793
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.06498384918086231,
          "train_loss": 0.06476891040802002,
          "validation_loss": 0.05313107371330261,
          "train_metric": 0.0065232873894274235,
          "validation_metric": 0.00535116670653224,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8977462040002138
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.06482536206021905,
          "train_loss": 0.06452950090169907,
          "validation_loss": 0.05282706022262573,
          "train_metric": 0.006499174050986767,
          "validation_metric": 0.0053205471485853195,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.910985229000289
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.06454466644208878,
          "train_loss": 0.0643238052725792,
          "validation_loss": 0.05272211879491806,
          "train_metric": 0.006478458642959595,
          "validation_metric": 0.00530997896566987,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9236965960008092
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.06425991328433156,
          "train_loss": 0.06411585956811905,
          "validation_loss": 0.05265204980969429,
          "train_metric": 0.0064575145952403545,
          "validation_metric": 0.005302920937538147,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9374137289996725
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.06409513321705163,
          "train_loss": 0.06391415745019913,
          "validation_loss": 0.05240627005696297,
          "train_metric": 0.006437199655920267,
          "validation_metric": 0.0052781677804887295,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9539407860011124
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.06383403320796788,
          "train_loss": 0.06369558721780777,
          "validation_loss": 0.05229288712143898,
          "train_metric": 0.006415185984224081,
          "validation_metric": 0.0052667479030787945,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9675062030000845
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.06363105704076588,
          "train_loss": 0.06349141895771027,
          "validation_loss": 0.05216612666845322,
          "train_metric": 0.006394623778760433,
          "validation_metric": 0.005253981798887253,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9805186289995618
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.06355332094244659,
          "train_loss": 0.06330831348896027,
          "validation_loss": 0.052065759897232056,
          "train_metric": 0.0063761817291378975,
          "validation_metric": 0.005243872292339802,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9928671199995733
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0633000488160178,
          "train_loss": 0.0630909726023674,
          "validation_loss": 0.051818013191223145,
          "train_metric": 0.006354291923344135,
          "validation_metric": 0.005218920763581991,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0061478879997594
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.06309695180971175,
          "train_loss": 0.06289026886224747,
          "validation_loss": 0.05162446200847626,
          "train_metric": 0.006334078032523394,
          "validation_metric": 0.00519942631945014,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0195223669998086
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.06285888073034585,
          "train_loss": 0.06269511580467224,
          "validation_loss": 0.05156675726175308,
          "train_metric": 0.006314422469586134,
          "validation_metric": 0.005193615332245827,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.032728254000176
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.06265345122665167,
          "train_loss": 0.06248779594898224,
          "validation_loss": 0.05138937383890152,
          "train_metric": 0.006293542217463255,
          "validation_metric": 0.005175749305635691,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0470212929994887
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.06251469382550567,
          "train_loss": 0.06232024356722832,
          "validation_loss": 0.051293812692165375,
          "train_metric": 0.006276666186749935,
          "validation_metric": 0.005166125018149614,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.060245605000091
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.06227767572272569,
          "train_loss": 0.06211775541305542,
          "validation_loss": 0.051156774163246155,
          "train_metric": 0.00625627301633358,
          "validation_metric": 0.005152322817593813,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0739285020008538
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.062192426063120365,
          "train_loss": 0.06191454082727432,
          "validation_loss": 0.050965745002031326,
          "train_metric": 0.00623580627143383,
          "validation_metric": 0.005133083555847406,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.087778150000304
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.06189025181811303,
          "train_loss": 0.061725061386823654,
          "validation_loss": 0.050762422382831573,
          "train_metric": 0.00621672160923481,
          "validation_metric": 0.005112605635076761,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.1023482800010242
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.06174471171107143,
          "train_loss": 0.061526454985141754,
          "validation_loss": 0.05070308595895767,
          "train_metric": 0.006196719128638506,
          "validation_metric": 0.005106628872454166,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.1159818200012523
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.06149722612462938,
          "train_loss": 0.061341747641563416,
          "validation_loss": 0.0505189523100853,
          "train_metric": 0.006178115960210562,
          "validation_metric": 0.005088083446025848,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.128742326000065
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0612808798905462,
          "train_loss": 0.06114331632852554,
          "validation_loss": 0.0503658801317215,
          "train_metric": 0.0061581311747431755,
          "validation_metric": 0.005072667263448238,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.141312621999532
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.06112142838537693,
          "train_loss": 0.06095170974731445,
          "validation_loss": 0.05028834939002991,
          "train_metric": 0.0061388337053358555,
          "validation_metric": 0.005064858589321375,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.155078236999543
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.06090781441889703,
          "train_loss": 0.06076136231422424,
          "validation_loss": 0.050124555826187134,
          "train_metric": 0.0061196619644761086,
          "validation_metric": 0.005048362072557211,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.168166855999516
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.06073561531957239,
          "train_loss": 0.06056975945830345,
          "validation_loss": 0.04995313659310341,
          "train_metric": 0.006100364029407501,
          "validation_metric": 0.0050310976803302765,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.181148857000153
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.06053819111548364,
          "train_loss": 0.0603727325797081,
          "validation_loss": 0.04980842396616936,
          "train_metric": 0.006080521270632744,
          "validation_metric": 0.005016522016376257,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.1937275340005726
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.06031982752028853,
          "train_loss": 0.060174182057380676,
          "validation_loss": 0.049685705453157425,
          "train_metric": 0.006060522980988026,
          "validation_metric": 0.005004162434488535,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.2065292250008497
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.060120540438219905,
          "train_loss": 0.05998610705137253,
          "validation_loss": 0.04954712837934494,
          "train_metric": 0.006041581276804209,
          "validation_metric": 0.004990205634385347,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.219866378000006
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.059967138222418725,
          "train_loss": 0.059791021049022675,
          "validation_loss": 0.04940802976489067,
          "train_metric": 0.006021932233124971,
          "validation_metric": 0.004976195283234119,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.233037594000052
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.05982582469005138,
          "train_loss": 0.05963677167892456,
          "validation_loss": 0.049337971955537796,
          "train_metric": 0.006006397306919098,
          "validation_metric": 0.00496914004907012,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.2465884880002704
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.05959089857060462,
          "train_loss": 0.05941391363739967,
          "validation_loss": 0.049083080142736435,
          "train_metric": 0.005983950570225716,
          "validation_metric": 0.004943467676639557,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.2614580690005823
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.059424839448183775,
          "train_loss": 0.059244513511657715,
          "validation_loss": 0.04894592985510826,
          "train_metric": 0.005966890603303909,
          "validation_metric": 0.004929654765874147,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.283121939000921
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.059310548356734216,
          "train_loss": 0.05904615670442581,
          "validation_loss": 0.048746563494205475,
          "train_metric": 0.005946912337094545,
          "validation_metric": 0.004909575916826725,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.296021346000998
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.058956354041583836,
          "train_loss": 0.05883881077170372,
          "validation_loss": 0.048662710934877396,
          "train_metric": 0.0059260292910039425,
          "validation_metric": 0.0049011302180588245,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.3084820229996694
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.058829075656831264,
          "train_loss": 0.05864764377474785,
          "validation_loss": 0.048494335263967514,
          "train_metric": 0.005906775593757629,
          "validation_metric": 0.004884171765297651,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.3523891540007753
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.05866361001972109,
          "train_loss": 0.05846532806754112,
          "validation_loss": 0.04842544347047806,
          "train_metric": 0.005888413637876511,
          "validation_metric": 0.004877233877778053,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.3746769910012517
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.05840241978876293,
          "train_loss": 0.05827050656080246,
          "validation_loss": 0.04826924949884415,
          "train_metric": 0.005868792533874512,
          "validation_metric": 0.004861502442508936,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.3928930050005874
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.058212265372276306,
          "train_loss": 0.058072127401828766,
          "validation_loss": 0.048067063093185425,
          "train_metric": 0.005848811939358711,
          "validation_metric": 0.004841138608753681,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.410941170001024
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.05802057485561818,
          "train_loss": 0.05788261815905571,
          "validation_loss": 0.04796930029988289,
          "train_metric": 0.005829725414514542,
          "validation_metric": 0.004831292666494846,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4287995110007614
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0578514663502574,
          "train_loss": 0.05768474191427231,
          "validation_loss": 0.04780501127243042,
          "train_metric": 0.005809795577079058,
          "validation_metric": 0.004814745858311653,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4469217300011223
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.05767955118790269,
          "train_loss": 0.0574917197227478,
          "validation_loss": 0.04758609086275101,
          "train_metric": 0.005790355149656534,
          "validation_metric": 0.004792696330696344,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.464890610001021
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.057440573000349104,
          "train_loss": 0.0572940967977047,
          "validation_loss": 0.04747513681650162,
          "train_metric": 0.0057704513892531395,
          "validation_metric": 0.0047815218567848206,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4830698080004368
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.057313659228384495,
          "train_loss": 0.05710215121507645,
          "validation_loss": 0.04737577214837074,
          "train_metric": 0.005751119926571846,
          "validation_metric": 0.004771513864398003,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.500852938001117
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.05714831920340657,
          "train_loss": 0.05690862983465195,
          "validation_loss": 0.0471915602684021,
          "train_metric": 0.005731629207730293,
          "validation_metric": 0.004752961918711662,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.519377340000574
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.056862091179937124,
          "train_loss": 0.0567082017660141,
          "validation_loss": 0.04702680557966232,
          "train_metric": 0.0057114423252642155,
          "validation_metric": 0.004736367613077164,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5376081539998268
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.056939444155432284,
          "train_loss": 0.05655307695269585,
          "validation_loss": 0.04700165614485741,
          "train_metric": 0.005695818457752466,
          "validation_metric": 0.004733834881335497,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5564302849998057
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.05671792314387858,
          "train_loss": 0.056371018290519714,
          "validation_loss": 0.04673611745238304,
          "train_metric": 0.005677482578903437,
          "validation_metric": 0.0047070905566215515,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5753730119995453
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.05630923714488745,
          "train_loss": 0.056127727031707764,
          "validation_loss": 0.04659704118967056,
          "train_metric": 0.005652978550642729,
          "validation_metric": 0.004693082999438047,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.593829673000073
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.056178198428824544,
          "train_loss": 0.055956024676561356,
          "validation_loss": 0.04648667573928833,
          "train_metric": 0.005635685287415981,
          "validation_metric": 0.004681968130171299,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.6124105689996213
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.055903569562360644,
          "train_loss": 0.0557381734251976,
          "validation_loss": 0.046230945736169815,
          "train_metric": 0.0056137447245419025,
          "validation_metric": 0.004656211473047733,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.6308663399995567
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.05567497666925192,
          "train_loss": 0.055535439401865005,
          "validation_loss": 0.046140339225530624,
          "train_metric": 0.005593326408416033,
          "validation_metric": 0.004647086374461651,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.649052835000475
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.05552417645230889,
          "train_loss": 0.05534502863883972,
          "validation_loss": 0.04603986814618111,
          "train_metric": 0.005574148613959551,
          "validation_metric": 0.004636966623365879,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.667426568999872
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.05530351388733834,
          "train_loss": 0.055145010352134705,
          "validation_loss": 0.04588012397289276,
          "train_metric": 0.005554002709686756,
          "validation_metric": 0.004620878491550684,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.685495360999994
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.055144741432741284,
          "train_loss": 0.05494064837694168,
          "validation_loss": 0.045672424137592316,
          "train_metric": 0.005533420480787754,
          "validation_metric": 0.004599959589540958,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.703539983000155
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.05488811817485839,
          "train_loss": 0.05474632978439331,
          "validation_loss": 0.04552158713340759,
          "train_metric": 0.005513849202543497,
          "validation_metric": 0.004584767390042543,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7214528910008084
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.05469042924232781,
          "train_loss": 0.05455005168914795,
          "validation_loss": 0.04534173756837845,
          "train_metric": 0.005494081415235996,
          "validation_metric": 0.004566654097288847,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.739250319000348
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.054525920655578375,
          "train_loss": 0.05434998497366905,
          "validation_loss": 0.045223940163850784,
          "train_metric": 0.005473931785672903,
          "validation_metric": 0.00455478997901082,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7572374989995296
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.05433612666092813,
          "train_loss": 0.05414760112762451,
          "validation_loss": 0.04509210214018822,
          "train_metric": 0.005453548394143581,
          "validation_metric": 0.00454151164740324,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7749240360008116
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.05412962508853525,
          "train_loss": 0.053951360285282135,
          "validation_loss": 0.04492536187171936,
          "train_metric": 0.005433783400803804,
          "validation_metric": 0.004524717573076487,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7928721709995443
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.05393870605621487,
          "train_loss": 0.053752630949020386,
          "validation_loss": 0.044675737619400024,
          "train_metric": 0.005413768347352743,
          "validation_metric": 0.004499576520174742,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.811625260001165
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.053754089400172234,
          "train_loss": 0.053542137145996094,
          "validation_loss": 0.044548001140356064,
          "train_metric": 0.005392568185925484,
          "validation_metric": 0.004486711695790291,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.831089069000882
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.05349531432148069,
          "train_loss": 0.053354356437921524,
          "validation_loss": 0.04453001171350479,
          "train_metric": 0.005373655818402767,
          "validation_metric": 0.004484900273382664,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.849634035001145
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.05347634351346642,
          "train_loss": 0.05317100137472153,
          "validation_loss": 0.04419400542974472,
          "train_metric": 0.005355188623070717,
          "validation_metric": 0.004451058804988861,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.867766864999794
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.05335378262680024,
          "train_loss": 0.052991386502981186,
          "validation_loss": 0.04427609592676163,
          "train_metric": 0.0053370981477200985,
          "validation_metric": 0.004459326155483723,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.885230992000288
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.053131150430999696,
          "train_loss": 0.05276736617088318,
          "validation_loss": 0.04391725733876228,
          "train_metric": 0.005314536392688751,
          "validation_metric": 0.004423184785991907,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.903420233000361
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.05270988564006984,
          "train_loss": 0.05253376066684723,
          "validation_loss": 0.04381268471479416,
          "train_metric": 0.0052910069935023785,
          "validation_metric": 0.004412653390318155,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.921577879000324
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.05250507383607328,
          "train_loss": 0.05232570692896843,
          "validation_loss": 0.043651483952999115,
          "train_metric": 0.005270053166896105,
          "validation_metric": 0.004396417643874884,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.940177759001017
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.0523126944899559,
          "train_loss": 0.05213093012571335,
          "validation_loss": 0.0435531809926033,
          "train_metric": 0.005250437185168266,
          "validation_metric": 0.004386517219245434,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.958024662000753
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.052109943120740354,
          "train_loss": 0.051930174231529236,
          "validation_loss": 0.04325598478317261,
          "train_metric": 0.00523021724075079,
          "validation_metric": 0.004356584511697292,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.9759233059994585
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.052014106535352767,
          "train_loss": 0.05172639340162277,
          "validation_loss": 0.04324677959084511,
          "train_metric": 0.005209692753851414,
          "validation_metric": 0.004355657380074263,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.9938023110007634
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.05165067862253636,
          "train_loss": 0.051518045365810394,
          "validation_loss": 0.04300319403409958,
          "train_metric": 0.005188709124922752,
          "validation_metric": 0.004331124480813742,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.021109270001034
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.051532529410906136,
          "train_loss": 0.051325783133506775,
          "validation_loss": 0.042736321687698364,
          "train_metric": 0.00516934460029006,
          "validation_metric": 0.004304245579987764,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0450151420009206
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.05131666688248515,
          "train_loss": 0.05111737921833992,
          "validation_loss": 0.042742419987916946,
          "train_metric": 0.005148355383425951,
          "validation_metric": 0.004304860252887011,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.090598560000217
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0510772115085274,
          "train_loss": 0.0509142130613327,
          "validation_loss": 0.04254331812262535,
          "train_metric": 0.005127893295139074,
          "validation_metric": 0.004284807480871677,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1295896940009698
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.050833410350605845,
          "train_loss": 0.05069780349731445,
          "validation_loss": 0.04233817383646965,
          "train_metric": 0.005106097087264061,
          "validation_metric": 0.004264145623892546,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.149750787000812
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.050653367885388434,
          "train_loss": 0.05049629509449005,
          "validation_loss": 0.042181346565485,
          "train_metric": 0.005085802171379328,
          "validation_metric": 0.004248350393027067,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1682335220011737
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.05050300096627325,
          "train_loss": 0.050288718193769455,
          "validation_loss": 0.04198429360985756,
          "train_metric": 0.0050648958422243595,
          "validation_metric": 0.004228503908962011,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.186351366000963
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.05021543079055846,
          "train_loss": 0.05008775740861893,
          "validation_loss": 0.04191955551505089,
          "train_metric": 0.005044654943048954,
          "validation_metric": 0.004221983719617128,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.2052451910003583
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.05006960453465581,
          "train_loss": 0.049872107803821564,
          "validation_loss": 0.04171266406774521,
          "train_metric": 0.0050229355692863464,
          "validation_metric": 0.004201147239655256,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.22445747600068
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.049824043293483555,
          "train_loss": 0.04966757446527481,
          "validation_loss": 0.04154486209154129,
          "train_metric": 0.005002336110919714,
          "validation_metric": 0.004184246063232422,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.2428965350009094
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.04961815383285284,
          "train_loss": 0.049471791833639145,
          "validation_loss": 0.041446734219789505,
          "train_metric": 0.004982617683708668,
          "validation_metric": 0.004174362402409315,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.2606230239998695
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.04940986051224172,
          "train_loss": 0.049257464706897736,
          "validation_loss": 0.041228827089071274,
          "train_metric": 0.00496103148907423,
          "validation_metric": 0.004152416717261076,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.384159035000266
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.04935232421848923,
          "train_loss": 0.049057234078645706,
          "validation_loss": 0.040984172374010086,
          "train_metric": 0.004940865095704794,
          "validation_metric": 0.0041277757845819,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.40317441300067
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.049019643862266093,
          "train_loss": 0.048845965415239334,
          "validation_loss": 0.04090970754623413,
          "train_metric": 0.004919586703181267,
          "validation_metric": 0.0041202763095498085,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.422432191000553
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.04879665107000619,
          "train_loss": 0.048635825514793396,
          "validation_loss": 0.04075305163860321,
          "train_metric": 0.004898421932011843,
          "validation_metric": 0.004104497842490673,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.441726073000609
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.048587781260721385,
          "train_loss": 0.04842523857951164,
          "validation_loss": 0.04059528931975365,
          "train_metric": 0.0048772129230201244,
          "validation_metric": 0.004088608548045158,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4601178869997966
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.04843206820078194,
          "train_loss": 0.04821924492716789,
          "validation_loss": 0.04039283096790314,
          "train_metric": 0.004856465384364128,
          "validation_metric": 0.004068218171596527,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4787650119997124
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.04818384116515517,
          "train_loss": 0.04801226779818535,
          "validation_loss": 0.04016993194818497,
          "train_metric": 0.004835620056837797,
          "validation_metric": 0.004045769106596708,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4980566080012068
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.04796679865103215,
          "train_loss": 0.047813139855861664,
          "validation_loss": 0.040164560079574585,
          "train_metric": 0.00481556449085474,
          "validation_metric": 0.004045227076858282,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5163403620008467
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.04775884246919304,
          "train_loss": 0.047593213617801666,
          "validation_loss": 0.03989424556493759,
          "train_metric": 0.004793413914740086,
          "validation_metric": 0.004018002655357122,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5352950470005453
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.047612760798074305,
          "train_loss": 0.047383882105350494,
          "validation_loss": 0.03970606252551079,
          "train_metric": 0.004772331099957228,
          "validation_metric": 0.003999049309641123,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5547948330004147
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.04736100637819618,
          "train_loss": 0.04719110205769539,
          "validation_loss": 0.03957851231098175,
          "train_metric": 0.004752914421260357,
          "validation_metric": 0.003986202646046877,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5735065680000844
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.04732775443699211,
          "train_loss": 0.04697800427675247,
          "validation_loss": 0.03937031328678131,
          "train_metric": 0.004731452092528343,
          "validation_metric": 0.003965233452618122,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5920061100005114
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.04697895469143987,
          "train_loss": 0.04677201807498932,
          "validation_loss": 0.03926176205277443,
          "train_metric": 0.004710706416517496,
          "validation_metric": 0.003954301122575998,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.6114542830000573
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.04673343792092055,
          "train_loss": 0.04658852145075798,
          "validation_loss": 0.03907905891537666,
          "train_metric": 0.0046922252513468266,
          "validation_metric": 0.003935900051146746,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.636352089999491
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.04661906533874571,
          "train_loss": 0.04637235775589943,
          "validation_loss": 0.03894822299480438,
          "train_metric": 0.0046704537235200405,
          "validation_metric": 0.003922722302377224,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.657671090000804
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.04637956619262695,
          "train_loss": 0.04616275802254677,
          "validation_loss": 0.038703687489032745,
          "train_metric": 0.004649343900382519,
          "validation_metric": 0.0038980937097221613,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.682118103999528
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.04616415756754577,
          "train_loss": 0.045980606228113174,
          "validation_loss": 0.03860202431678772,
          "train_metric": 0.0046309977769851685,
          "validation_metric": 0.0038878542836755514,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.704993610999736
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.046014148625545204,
          "train_loss": 0.04576319828629494,
          "validation_loss": 0.03842276334762573,
          "train_metric": 0.004609101451933384,
          "validation_metric": 0.0038697998970746994,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7271840480007086
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0457851798273623,
          "train_loss": 0.04557831585407257,
          "validation_loss": 0.03829306364059448,
          "train_metric": 0.004590480588376522,
          "validation_metric": 0.0038567373994737864,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.751847812000051
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.045514031779021025,
          "train_loss": 0.045362360775470734,
          "validation_loss": 0.03808065503835678,
          "train_metric": 0.0045687309466302395,
          "validation_metric": 0.003835343988612294,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.774090509999951
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.045323372702114284,
          "train_loss": 0.04515823721885681,
          "validation_loss": 0.03786328434944153,
          "train_metric": 0.004548172000795603,
          "validation_metric": 0.003813451621681452,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8002694509996218
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.04512827331200242,
          "train_loss": 0.044959161430597305,
          "validation_loss": 0.037795476615428925,
          "train_metric": 0.004528121557086706,
          "validation_metric": 0.0038066222332417965,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.822590568999658
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.044907253119163215,
          "train_loss": 0.044756583869457245,
          "validation_loss": 0.03756903111934662,
          "train_metric": 0.004507719539105892,
          "validation_metric": 0.0037838153075426817,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8522022390006896
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.04481696302536875,
          "train_loss": 0.04457203671336174,
          "validation_loss": 0.03739645704627037,
          "train_metric": 0.0044891322031617165,
          "validation_metric": 0.0037664342671632767,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8755383690004237
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.044590892270207405,
          "train_loss": 0.044377606362104416,
          "validation_loss": 0.03730221465229988,
          "train_metric": 0.004469549749046564,
          "validation_metric": 0.003756942693144083,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.902799452000181
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.044389075599610806,
          "train_loss": 0.04416286572813988,
          "validation_loss": 0.037086121737957,
          "train_metric": 0.0044479225762188435,
          "validation_metric": 0.0037351783830672503,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9263610920006613
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.04414274403825402,
          "train_loss": 0.04396449774503708,
          "validation_loss": 0.03691425174474716,
          "train_metric": 0.00442794244736433,
          "validation_metric": 0.0037178685888648033,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9534536999999546
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.04394651378970593,
          "train_loss": 0.04380390793085098,
          "validation_loss": 0.036815740168094635,
          "train_metric": 0.004411769099533558,
          "validation_metric": 0.00370794627815485,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.977502286001254
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.043943264172412455,
          "train_loss": 0.04359918087720871,
          "validation_loss": 0.03657089173793793,
          "train_metric": 0.0043911500833928585,
          "validation_metric": 0.0036832864861935377,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.005587052000919
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.04353397153317928,
          "train_loss": 0.04337698221206665,
          "validation_loss": 0.03644911199808121,
          "train_metric": 0.0043687704019248486,
          "validation_metric": 0.0036710205022245646,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.028830896000727
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.0435607829131186,
          "train_loss": 0.043226465582847595,
          "validation_loss": 0.03633886203169823,
          "train_metric": 0.00435361173003912,
          "validation_metric": 0.0036599168088287115,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.057940468999732
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 3840
      },
      "best_validation_metric": 0.0036599168088287115,
      "config": {
        "label": "LR 0.0001",
        "batch": 64,
        "lr": 0.0001,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.0604972462912876
    },
    "32d8a0bb9dc3_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7750606071203947,
          "train_loss": 0.38536715507507324,
          "validation_loss": 0.30186760425567627,
          "train_metric": 0.03881276398897171,
          "validation_metric": 0.030403001233935356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.022844549999717856
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.2060750015079975,
          "train_loss": 0.12688414752483368,
          "validation_loss": 0.1095275804400444,
          "train_metric": 0.01277930662035942,
          "validation_metric": 0.011031216941773891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04602210200027912
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.1240247250534594,
          "train_loss": 0.09904789179563522,
          "validation_loss": 0.08297248184680939,
          "train_metric": 0.009975740686058998,
          "validation_metric": 0.008356684818863869,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05847085200002766
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09276640950702131,
          "train_loss": 0.0849667638540268,
          "validation_loss": 0.06827860325574875,
          "train_metric": 0.00855754129588604,
          "validation_metric": 0.006876770872622728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0711577070014755
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07504639495164156,
          "train_loss": 0.06760113686323166,
          "validation_loss": 0.055644359439611435,
          "train_metric": 0.006808537989854813,
          "validation_metric": 0.005604295991361141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08355269500134455
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06591126136481762,
          "train_loss": 0.0652155801653862,
          "validation_loss": 0.052817028015851974,
          "train_metric": 0.0065682739950716496,
          "validation_metric": 0.005319537594914436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09611831900110701
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06305562076158822,
          "train_loss": 0.06514137238264084,
          "validation_loss": 0.05300118774175644,
          "train_metric": 0.00656080013141036,
          "validation_metric": 0.005338085815310478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10805701100071019
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06273406581021845,
          "train_loss": 0.05920339375734329,
          "validation_loss": 0.04777815192937851,
          "train_metric": 0.005962749011814594,
          "validation_metric": 0.0048120408318936825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12025022800116858
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05564656085334718,
          "train_loss": 0.052560701966285706,
          "validation_loss": 0.0434177927672863,
          "train_metric": 0.005293721333146095,
          "validation_metric": 0.004372881259769201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13243190400135063
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05131560959853232,
          "train_loss": 0.051526989787817,
          "validation_loss": 0.04277803748846054,
          "train_metric": 0.0051896097138524055,
          "validation_metric": 0.004308447707444429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14808082800118427
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04965876357164234,
          "train_loss": 0.0480007603764534,
          "validation_loss": 0.03882240504026413,
          "train_metric": 0.004834461025893688,
          "validation_metric": 0.003910050727427006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17330320099972596
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04525484796613455,
          "train_loss": 0.041656482964754105,
          "validation_loss": 0.034643612802028656,
          "train_metric": 0.004195488058030605,
          "validation_metric": 0.0034891776740550995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1869157399996766
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.0406425753608346,
          "train_loss": 0.03904619812965393,
          "validation_loss": 0.032580845057964325,
          "train_metric": 0.003932590130716562,
          "validation_metric": 0.003281423356384039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19922176500040223
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03730706369969994,
          "train_loss": 0.03512808308005333,
          "validation_loss": 0.029410352930426598,
          "train_metric": 0.003537971992045641,
          "validation_metric": 0.0029621028807014227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2116017180014751
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03614106320310384,
          "train_loss": 0.03985850140452385,
          "validation_loss": 0.033791474997997284,
          "train_metric": 0.004014401696622372,
          "validation_metric": 0.0034033532720059156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22280760800094868
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.034006054629571736,
          "train_loss": 0.03577851876616478,
          "validation_loss": 0.03025796450674534,
          "train_metric": 0.0036034814547747374,
          "validation_metric": 0.0030474714003503323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2340152169999783
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.030239578511100262,
          "train_loss": 0.02710607275366783,
          "validation_loss": 0.023179994896054268,
          "train_metric": 0.0027300245128571987,
          "validation_metric": 0.00233460427261889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.246660172000702
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029969822440762073,
          "train_loss": 0.03654274344444275,
          "validation_loss": 0.028627360239624977,
          "train_metric": 0.003680451540276408,
          "validation_metric": 0.0028832426760345697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2587022900006559
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.031163847947027534,
          "train_loss": 0.02538416162133217,
          "validation_loss": 0.019166603684425354,
          "train_metric": 0.002556599909439683,
          "validation_metric": 0.0019303900189697742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.271378225999797
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.023735167924314737,
          "train_loss": 0.02233070135116577,
          "validation_loss": 0.018863607197999954,
          "train_metric": 0.0022490667179226875,
          "validation_metric": 0.001899873255752027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28370116200130724
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023142108839238063,
          "train_loss": 0.01990034058690071,
          "validation_loss": 0.01686880923807621,
          "train_metric": 0.002004289533942938,
          "validation_metric": 0.0016989647410809994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29635380800027633
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02312045422149822,
          "train_loss": 0.020990870893001556,
          "validation_loss": 0.016307419165968895,
          "train_metric": 0.002114123897626996,
          "validation_metric": 0.0016424234490841627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3144892960008292
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.022139596985653043,
          "train_loss": 0.020363621413707733,
          "validation_loss": 0.01820952072739601,
          "train_metric": 0.0020509492605924606,
          "validation_metric": 0.0018339961534366012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3433234590011125
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.018391056975815445,
          "train_loss": 0.01689496450126171,
          "validation_loss": 0.01443193294107914,
          "train_metric": 0.001701599103398621,
          "validation_metric": 0.001453531440347433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3715355180011102
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019256686326116323,
          "train_loss": 0.019413551315665245,
          "validation_loss": 0.014614056795835495,
          "train_metric": 0.0019552616868168116,
          "validation_metric": 0.0014718743041157722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4029409780014248
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.018628374149557203,
          "train_loss": 0.021552782505750656,
          "validation_loss": 0.02029510959982872,
          "train_metric": 0.002170717343688011,
          "validation_metric": 0.0020440490916371346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4287909690010565
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.017534870916279033,
          "train_loss": 0.015243593603372574,
          "validation_loss": 0.013040601275861263,
          "train_metric": 0.001535278744995594,
          "validation_metric": 0.001313401386141777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4592652050014294
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015402383840410039,
          "train_loss": 0.014646788127720356,
          "validation_loss": 0.012762422673404217,
          "train_metric": 0.0014751707203686237,
          "validation_metric": 0.0012853844091296196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4895106380008656
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016872672975296155,
          "train_loss": 0.018696356564760208,
          "validation_loss": 0.015021326020359993,
          "train_metric": 0.0018830286571756005,
          "validation_metric": 0.001512893009930849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5168939870000031
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.016743629763368517,
          "train_loss": 0.01551545225083828,
          "validation_loss": 0.013348198495805264,
          "train_metric": 0.0015626593958586454,
          "validation_metric": 0.001344381715171039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5438024930008396
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.021813345549162477,
          "train_loss": 0.014776387251913548,
          "validation_loss": 0.011951805092394352,
          "train_metric": 0.001488223671913147,
          "validation_metric": 0.0012037419946864247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5562532280000596
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.018929659272544086,
          "train_loss": 0.014764338731765747,
          "validation_loss": 0.010697023011744022,
          "train_metric": 0.0014870099257677794,
          "validation_metric": 0.001077365013770759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5690664570010995
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.0153573855641298,
          "train_loss": 0.012026515789330006,
          "validation_loss": 0.009608716703951359,
          "train_metric": 0.0012112664990127087,
          "validation_metric": 0.0009677548077888787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5813832720014034
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013202551752328873,
          "train_loss": 0.01557985134422779,
          "validation_loss": 0.010827160440385342,
          "train_metric": 0.0015691454755142331,
          "validation_metric": 0.0010904718656092882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5936492670007283
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014358282889588736,
          "train_loss": 0.015606962144374847,
          "validation_loss": 0.014100464060902596,
          "train_metric": 0.0015718759968876839,
          "validation_metric": 0.001420147018507123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6051429460003419
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.016984775924356654,
          "train_loss": 0.02222522161900997,
          "validation_loss": 0.01624390110373497,
          "train_metric": 0.002238442888483405,
          "validation_metric": 0.001636026194319129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6164646560009714
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.012299630441702902,
          "train_loss": 0.011116322129964828,
          "validation_loss": 0.009583269245922565,
          "train_metric": 0.0011195952538400888,
          "validation_metric": 0.0009651916334405541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6288518310011568
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011669085710309446,
          "train_loss": 0.009766273200511932,
          "validation_loss": 0.007932213135063648,
          "train_metric": 0.0009836232056841254,
          "validation_metric": 0.0007989035220816731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6412939760011795
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01077001288649626,
          "train_loss": 0.009837810881435871,
          "validation_loss": 0.00763594638556242,
          "train_metric": 0.0009908282663673162,
          "validation_metric": 0.0007690644706599414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.653495507000116
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012661474363994785,
          "train_loss": 0.012773655354976654,
          "validation_loss": 0.009138710796833038,
          "train_metric": 0.0012865157332271338,
          "validation_metric": 0.0009204175439663231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7428281730008166
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010784950369270518,
          "train_loss": 0.009931527078151703,
          "validation_loss": 0.007522317115217447,
          "train_metric": 0.0010002671042457223,
          "validation_metric": 0.0007576202624477446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7553219870005705
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010854612351977266,
          "train_loss": 0.009563778527081013,
          "validation_loss": 0.007020589429885149,
          "train_metric": 0.0009632286964915693,
          "validation_metric": 0.0007070880965329707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7677749459999177
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011339242788380943,
          "train_loss": 0.0087605444714427,
          "validation_loss": 0.007057554554194212,
          "train_metric": 0.0008823300013318658,
          "validation_metric": 0.0007108110003173351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7798063800000818
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.00978471252892632,
          "train_loss": 0.00902597513049841,
          "validation_loss": 0.0073502580635249615,
          "train_metric": 0.0009090631501749158,
          "validation_metric": 0.0007402910850942135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.791204421000657
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013840032886946574,
          "train_loss": 0.010268403217196465,
          "validation_loss": 0.008520981296896935,
          "train_metric": 0.0010341958841308951,
          "validation_metric": 0.0008582019945606589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8026873110011365
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.012792111374437809,
          "train_loss": 0.010196718387305737,
          "validation_loss": 0.007601565681397915,
          "train_metric": 0.0010269759222865105,
          "validation_metric": 0.0007656019297428429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8139903930004948
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011586189008085057,
          "train_loss": 0.012726811692118645,
          "validation_loss": 0.011708073318004608,
          "train_metric": 0.0012817978858947754,
          "validation_metric": 0.00117919419426471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8253195790002792
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.009888891261653043,
          "train_loss": 0.014078420586884022,
          "validation_loss": 0.013561593368649483,
          "train_metric": 0.0014179269783198833,
          "validation_metric": 0.0013658740790560842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8365089690014429
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.01211502478690818,
          "train_loss": 0.010559041053056717,
          "validation_loss": 0.008403655141592026,
          "train_metric": 0.001063467818312347,
          "validation_metric": 0.0008463853155262768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8479923410013726
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.011117999012640212,
          "train_loss": 0.00816005002707243,
          "validation_loss": 0.006046057213097811,
          "train_metric": 0.0008218504954129457,
          "validation_metric": 0.0006089368253014982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.860032766000586
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.00928575212310534,
          "train_loss": 0.012129377573728561,
          "validation_loss": 0.01141818892210722,
          "train_metric": 0.0012216264149174094,
          "validation_metric": 0.0011499980464577675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.871916959000373
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008625807182397693,
          "train_loss": 0.00884421169757843,
          "validation_loss": 0.0063227578066289425,
          "train_metric": 0.0008907566079869866,
          "validation_metric": 0.0006368050235323608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8834383210014494
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008779473515460268,
          "train_loss": 0.010252740234136581,
          "validation_loss": 0.00750874076038599,
          "train_metric": 0.0010326184565201402,
          "validation_metric": 0.0007562528480775654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8953103189996909
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0091602051834343,
          "train_loss": 0.01068117842078209,
          "validation_loss": 0.010125876404345036,
          "train_metric": 0.0010757691925391555,
          "validation_metric": 0.0010198411764577031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9073519860012311
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.013468224395182915,
          "train_loss": 0.00999869592487812,
          "validation_loss": 0.007742003537714481,
          "train_metric": 0.0010070321150124073,
          "validation_metric": 0.0007797462749294937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9186928830004035
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.010853008774574846,
          "train_loss": 0.008161451667547226,
          "validation_loss": 0.0058862557634711266,
          "train_metric": 0.0008219915325753391,
          "validation_metric": 0.0005928422324359417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9316707950001728
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008787122336798348,
          "train_loss": 0.007462658453732729,
          "validation_loss": 0.006129884626716375,
          "train_metric": 0.0007516115438193083,
          "validation_metric": 0.0006173795554786921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.943603746000008
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007215693418402225,
          "train_loss": 0.006528432480990887,
          "validation_loss": 0.005115540232509375,
          "train_metric": 0.0006575197330676019,
          "validation_metric": 0.0005152184749022126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9563752250014659
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007845858577638865,
          "train_loss": 0.0069840820506215096,
          "validation_loss": 0.006143107544630766,
          "train_metric": 0.0007034111185930669,
          "validation_metric": 0.0006187112885527313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9679100480007037
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007135726671549492,
          "train_loss": 0.006284323520958424,
          "validation_loss": 0.005035600624978542,
          "train_metric": 0.0006329340394586325,
          "validation_metric": 0.0005071673076599836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9803269540007022
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.009110017359489575,
          "train_loss": 0.006696862168610096,
          "validation_loss": 0.005621404852718115,
          "train_metric": 0.0006744833663105965,
          "validation_metric": 0.0005661673494614661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9920468660002371
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007818591664545238,
          "train_loss": 0.006637724116444588,
          "validation_loss": 0.005256288219243288,
          "train_metric": 0.0006685271509923041,
          "validation_metric": 0.0005293941358104348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0040672430004633
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006970037764403969,
          "train_loss": 0.006174634676426649,
          "validation_loss": 0.004937969613820314,
          "train_metric": 0.0006218865746632218,
          "validation_metric": 0.0004973342875018716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0160357840013603
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007135511885280721,
          "train_loss": 0.009442456066608429,
          "validation_loss": 0.008755910210311413,
          "train_metric": 0.0009510095696896315,
          "validation_metric": 0.0008818632340990007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0282810990011058
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008260824164608493,
          "train_loss": 0.008309222757816315,
          "validation_loss": 0.0070204706862568855,
          "train_metric": 0.0008368745329789817,
          "validation_metric": 0.0007070761639624834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0398662469997362
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0069822732475586236,
          "train_loss": 0.0067450315691530704,
          "validation_loss": 0.004807147663086653,
          "train_metric": 0.000679334974847734,
          "validation_metric": 0.0004841582849621773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0520567410003423
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008227432139392477,
          "train_loss": 0.00827042106539011,
          "validation_loss": 0.005731015466153622,
          "train_metric": 0.0008329665288329124,
          "validation_metric": 0.0005772068398073316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0638072350011498
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00791827542707324,
          "train_loss": 0.006597149185836315,
          "validation_loss": 0.00453611183911562,
          "train_metric": 0.000664440740365535,
          "validation_metric": 0.00045686058001592755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0761028080014512
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0060867796710226685,
          "train_loss": 0.006203494966030121,
          "validation_loss": 0.005195828154683113,
          "train_metric": 0.0006247933488339186,
          "validation_metric": 0.0005233047413639724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0879902720007522
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006417378521291539,
          "train_loss": 0.005572378169745207,
          "validation_loss": 0.0043088579550385475,
          "train_metric": 0.0005612294771708548,
          "validation_metric": 0.000433972425526008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.100093861001369
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.007080948373186402,
          "train_loss": 0.006128726992756128,
          "validation_loss": 0.00446541141718626,
          "train_metric": 0.0006172629073262215,
          "validation_metric": 0.0004497398913372308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.111940009999671
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006899416650412604,
          "train_loss": 0.005498890299350023,
          "validation_loss": 0.0047007291577756405,
          "train_metric": 0.0005538280820474029,
          "validation_metric": 0.00047344021731987596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1238794450000569
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.010936821228824556,
          "train_loss": 0.01833980344235897,
          "validation_loss": 0.014932270161807537,
          "train_metric": 0.0018471179064363241,
          "validation_metric": 0.0015039234422147274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.135260464001476
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.010608561802655458,
          "train_loss": 0.006225930526852608,
          "validation_loss": 0.005035202018916607,
          "train_metric": 0.0006270530284382403,
          "validation_metric": 0.0005071270861662924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1470366550001927
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006480617972556502,
          "train_loss": 0.006028239615261555,
          "validation_loss": 0.004369248636066914,
          "train_metric": 0.0006071422249078751,
          "validation_metric": 0.0004400547768454999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1585277890008001
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.009216629012371413,
          "train_loss": 0.006511844228953123,
          "validation_loss": 0.005704014096409082,
          "train_metric": 0.0006558491731993854,
          "validation_metric": 0.0005744873778894544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1698737560000154
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007850948211853392,
          "train_loss": 0.009311886504292488,
          "validation_loss": 0.007155830971896648,
          "train_metric": 0.0009378591785207391,
          "validation_metric": 0.0007207090966403484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1815461920014059
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.008934477751608938,
          "train_loss": 0.012272932566702366,
          "validation_loss": 0.01282534096390009,
          "train_metric": 0.0012360847322270274,
          "validation_metric": 0.001291721360757947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1936300200013648
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00737155377282761,
          "train_loss": 0.005852990318089724,
          "validation_loss": 0.004597283899784088,
          "train_metric": 0.0005894917994737625,
          "validation_metric": 0.0004630216280929744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2119557389996771
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.006133655791927595,
          "train_loss": 0.005515925120562315,
          "validation_loss": 0.004237538203597069,
          "train_metric": 0.0005555438692681491,
          "validation_metric": 0.0004267893673386425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2245920590012247
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005041230389906559,
          "train_loss": 0.00474447151646018,
          "validation_loss": 0.0038033362943679094,
          "train_metric": 0.000477845809655264,
          "validation_metric": 0.00038305812631733716,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.2374806190000527
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0049037838107324205,
          "train_loss": 0.00500243017449975,
          "validation_loss": 0.004411051049828529,
          "train_metric": 0.0005038264207541943,
          "validation_metric": 0.0004442649078555405,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.2493052760000865
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.00502309557487024,
          "train_loss": 0.004997940734028816,
          "validation_loss": 0.00362793798558414,
          "train_metric": 0.0005033742636442184,
          "validation_metric": 0.000365392625099048,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.2619173390012293
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.005047214348451234,
          "train_loss": 0.00574933085590601,
          "validation_loss": 0.004080896731466055,
          "train_metric": 0.0005790514987893403,
          "validation_metric": 0.00041101296665146947,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.2736429150008917
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005391317376052029,
          "train_loss": 0.004832236561924219,
          "validation_loss": 0.003644507611170411,
          "train_metric": 0.00048668516683392227,
          "validation_metric": 0.0003670614678412676,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.2852876630004175
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005059303737652954,
          "train_loss": 0.005107297096401453,
          "validation_loss": 0.004661521874368191,
          "train_metric": 0.0005143882590346038,
          "validation_metric": 0.0004694914387073368,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.2970805570002995
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.004988682885596063,
          "train_loss": 0.004644440952688456,
          "validation_loss": 0.003761210013180971,
          "train_metric": 0.00046777111128903925,
          "validation_metric": 0.00037881528260186315,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.308723418000227
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.004758000206493307,
          "train_loss": 0.004567118361592293,
          "validation_loss": 0.0036226236261427402,
          "train_metric": 0.00045998336281627417,
          "validation_metric": 0.0003648574056569487,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.3212532920006197
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004673292685765773,
          "train_loss": 0.0046128565445542336,
          "validation_loss": 0.003866031765937805,
          "train_metric": 0.0004645901208277792,
          "validation_metric": 0.000389372551580891,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.3332390300001862
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0047656327005825005,
          "train_loss": 0.004617845173925161,
          "validation_loss": 0.003953693900257349,
          "train_metric": 0.00046509248204529285,
          "validation_metric": 0.00039820163510739803,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.3450517370001762
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004744009354908485,
          "train_loss": 0.0046650744043290615,
          "validation_loss": 0.004029775038361549,
          "train_metric": 0.00046984918299131095,
          "validation_metric": 0.0004058641497977078,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.3564034600003652
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004689709268859588,
          "train_loss": 0.004572537261992693,
          "validation_loss": 0.0038634873926639557,
          "train_metric": 0.00046052917605265975,
          "validation_metric": 0.00038911632145754993,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.3682632489999378
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004750270054501016,
          "train_loss": 0.0054732137359678745,
          "validation_loss": 0.0038010224234312773,
          "train_metric": 0.0005512420902960002,
          "validation_metric": 0.0003828250919468701,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.3799616169999354
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004965516571246553,
          "train_loss": 0.00456768274307251,
          "validation_loss": 0.003558696713298559,
          "train_metric": 0.0004600402317009866,
          "validation_metric": 0.0003584188816603273,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.3920810430008714
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004894592158962041,
          "train_loss": 0.005821473430842161,
          "validation_loss": 0.0054839784279465675,
          "train_metric": 0.0005863175028935075,
          "validation_metric": 0.0005523262079805136,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.4038707630006684
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004820666079467628,
          "train_loss": 0.004593316465616226,
          "validation_loss": 0.0035561295226216316,
          "train_metric": 0.00046262203250080347,
          "validation_metric": 0.00035816029412671924,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.4163302570013911
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005032763830968179,
          "train_loss": 0.004780857358127832,
          "validation_loss": 0.003458757419139147,
          "train_metric": 0.00048151047667488456,
          "validation_metric": 0.00034835332189686596,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.442900070000178
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004818686829821672,
          "train_loss": 0.0046686516143381596,
          "validation_loss": 0.003572339890524745,
          "train_metric": 0.0004702094884123653,
          "validation_metric": 0.000359793019015342,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.466717396000604
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.004780120187206194,
          "train_loss": 0.004374153912067413,
          "validation_loss": 0.003546791849657893,
          "train_metric": 0.00044054878526367247,
          "validation_metric": 0.0003572198620531708,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.4811580870009493
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004911688512947876,
          "train_loss": 0.004610151052474976,
          "validation_loss": 0.004077969118952751,
          "train_metric": 0.0004643175343517214,
          "validation_metric": 0.0004107181157451123,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.5018655960011529
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.004972768809238914,
          "train_loss": 0.005286091472953558,
          "validation_loss": 0.005002881400287151,
          "train_metric": 0.0005323957884684205,
          "validation_metric": 0.0005038718227297068,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.5197847170002206
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.005148311160155572,
          "train_loss": 0.004461069591343403,
          "validation_loss": 0.0035439017228782177,
          "train_metric": 0.0004493025771807879,
          "validation_metric": 0.000356928794644773,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.5322867770009907
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0046347979514393955,
          "train_loss": 0.005202305968850851,
          "validation_loss": 0.0036487651523202658,
          "train_metric": 0.0005239571910351515,
          "validation_metric": 0.0003674902836792171,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.5466983940004866
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004608947419910692,
          "train_loss": 0.004270557314157486,
          "validation_loss": 0.0034450197126716375,
          "train_metric": 0.0004301149456296116,
          "validation_metric": 0.00034696978400461376,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.5598449080007413
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004518281493801624,
          "train_loss": 0.004451180808246136,
          "validation_loss": 0.0038834507577121258,
          "train_metric": 0.00044830661499872804,
          "validation_metric": 0.0003911269595846534,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.5743311229998653
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.005192222051846329,
          "train_loss": 0.004291570745408535,
          "validation_loss": 0.0033969725482165813,
          "train_metric": 0.00043223125976510346,
          "validation_metric": 0.00034213069011457264,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.5876931929997227
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004425687802722678,
          "train_loss": 0.0044556655921041965,
          "validation_loss": 0.0032884764950722456,
          "train_metric": 0.0004487582773435861,
          "validation_metric": 0.0003312033077236265,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.600994055001138
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0045606026105815545,
          "train_loss": 0.004408468026667833,
          "validation_loss": 0.0032490172889083624,
          "train_metric": 0.00044400474871508777,
          "validation_metric": 0.0003272291214670986,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.614189752999664
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.004437955787579995,
          "train_loss": 0.004214728716760874,
          "validation_loss": 0.0035251695662736893,
          "train_metric": 0.00042449202737770975,
          "validation_metric": 0.0003550421679392457,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.6263295389999257
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.00461914534389507,
          "train_loss": 0.004273613449186087,
          "validation_loss": 0.003256484167650342,
          "train_metric": 0.0004304226895328611,
          "validation_metric": 0.0003279811062384397,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.726356802000737
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.00456532074167626,
          "train_loss": 0.005155818536877632,
          "validation_loss": 0.004795063752681017,
          "train_metric": 0.0005192751414142549,
          "validation_metric": 0.00048294125008396804,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.741552377001426
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004577920066367369,
          "train_loss": 0.004272719845175743,
          "validation_loss": 0.0034206868149340153,
          "train_metric": 0.00043033267138525844,
          "validation_metric": 0.00034451906685717404,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.7576627830003417
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004375878321297932,
          "train_loss": 0.004375044256448746,
          "validation_loss": 0.0038480227813124657,
          "train_metric": 0.00044063845416530967,
          "validation_metric": 0.00038755880086682737,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.7714145000009012
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004514911044680048,
          "train_loss": 0.005181567277759314,
          "validation_loss": 0.004918737802654505,
          "train_metric": 0.0005218685255385935,
          "validation_metric": 0.0004953972529619932,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.7835427690006327
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004541810565569904,
          "train_loss": 0.004566565155982971,
          "validation_loss": 0.0032164859585464,
          "train_metric": 0.00045992774539627135,
          "validation_metric": 0.0003239526995457709,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.8002089300007356
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.004362832500191871,
          "train_loss": 0.00423412537202239,
          "validation_loss": 0.003751708660274744,
          "train_metric": 0.000426445621997118,
          "validation_metric": 0.00037785840686410666,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.813175006000165
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.004390971269458532,
          "train_loss": 0.004596564453095198,
          "validation_loss": 0.003390977159142494,
          "train_metric": 0.0004629490722436458,
          "validation_metric": 0.00034152675652876496,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.8247765800006164
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.004452181121450849,
          "train_loss": 0.0040939548052847385,
          "validation_loss": 0.003107423195615411,
          "train_metric": 0.0004123280814383179,
          "validation_metric": 0.00031296827364712954,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.8371712360003585
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0047511310476693325,
          "train_loss": 0.005367332138121128,
          "validation_loss": 0.003637331770732999,
          "train_metric": 0.0005405779811553657,
          "validation_metric": 0.00036633870331570506,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.849113080999814
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0044725688712787814,
          "train_loss": 0.004054759629070759,
          "validation_loss": 0.0033847191371023655,
          "train_metric": 0.0004083805251866579,
          "validation_metric": 0.0003408965130802244,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.8608271920002153
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004062514184624888,
          "train_loss": 0.003960333298891783,
          "validation_loss": 0.0031137524638324976,
          "train_metric": 0.00039887032471597195,
          "validation_metric": 0.0003136057057417929,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.8723503689998324
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0043595946408458985,
          "train_loss": 0.003903955454006791,
          "validation_loss": 0.0032426121179014444,
          "train_metric": 0.0003931921091862023,
          "validation_metric": 0.00032658400596119463,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.8837825990012789
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.004126244682993274,
          "train_loss": 0.004515955690294504,
          "validation_loss": 0.0032216543331742287,
          "train_metric": 0.00045483047142624855,
          "validation_metric": 0.0003244732797611505,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.896363198000472
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.004257282053004019,
          "train_loss": 0.004165589343756437,
          "validation_loss": 0.003076025517657399,
          "train_metric": 0.00041954295011237264,
          "validation_metric": 0.0003098060260526836,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9094221030009066
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.004225035350827966,
          "train_loss": 0.006396124139428139,
          "validation_loss": 0.006279343273490667,
          "train_metric": 0.0006441941950470209,
          "validation_metric": 0.0006324324058368802,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.921304470000905
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.005322974888258614,
          "train_loss": 0.005139094777405262,
          "validation_loss": 0.005010196939110756,
          "train_metric": 0.0005175907863304019,
          "validation_metric": 0.0005046086735092103,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9332501950011647
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.004643073429178912,
          "train_loss": 0.004105349536985159,
          "validation_loss": 0.00340811675414443,
          "train_metric": 0.00041347582009620965,
          "validation_metric": 0.0003432530502323061,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9452285440002015
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.003999155363999307,
          "train_loss": 0.0038795610889792442,
          "validation_loss": 0.003059673123061657,
          "train_metric": 0.00039073516381904483,
          "validation_metric": 0.00030815909849479795,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9582047769999917
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.00399555479089031,
          "train_loss": 0.003797795157879591,
          "validation_loss": 0.0031672737095505,
          "train_metric": 0.0003825000603683293,
          "validation_metric": 0.0003189962008036673,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9706803190001665
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.004025205220386852,
          "train_loss": 0.0038598733954131603,
          "validation_loss": 0.002941880375146866,
          "train_metric": 0.0003887522907461971,
          "validation_metric": 0.00029629538767039776,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9827836399999796
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.003964473788073519,
          "train_loss": 0.004671039059758186,
          "validation_loss": 0.0032209435012191534,
          "train_metric": 0.0004704499151557684,
          "validation_metric": 0.00032440165523439646,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9947674850009207
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.004388618828670587,
          "train_loss": 0.003939889371395111,
          "validation_loss": 0.002905053086578846,
          "train_metric": 0.0003968111996073276,
          "validation_metric": 0.0002925863373093307,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.007177261000834
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0038543412229046226,
          "train_loss": 0.0037817831616848707,
          "validation_loss": 0.0031304750591516495,
          "train_metric": 0.00038088735891506076,
          "validation_metric": 0.000315290002617985,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.018989154001247
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.00380409962599515,
          "train_loss": 0.004265932831913233,
          "validation_loss": 0.003082460258156061,
          "train_metric": 0.0004296491388231516,
          "validation_metric": 0.00031045411014929414,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0306872870005463
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.00442943480447866,
          "train_loss": 0.004667850211262703,
          "validation_loss": 0.003212481504306197,
          "train_metric": 0.0004701287252828479,
          "validation_metric": 0.0003235492913518101,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0422309920013504
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004003091744380072,
          "train_loss": 0.003723218571394682,
          "validation_loss": 0.0030849867034703493,
          "train_metric": 0.00037498894380405545,
          "validation_metric": 0.0003107085940428078,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0539767270001903
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003929132406483404,
          "train_loss": 0.004161885008215904,
          "validation_loss": 0.003878184827044606,
          "train_metric": 0.00041916986810974777,
          "validation_metric": 0.0003905965422745794,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0654384350000328
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.004154608126555104,
          "train_loss": 0.004161234479397535,
          "validation_loss": 0.0038705195765942335,
          "train_metric": 0.00041910429717972875,
          "validation_metric": 0.0003898245340678841,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.077020665999953
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0046171616850188,
          "train_loss": 0.005875219590961933,
          "validation_loss": 0.005803531501442194,
          "train_metric": 0.0005917306407354772,
          "validation_metric": 0.0005845104460604489,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.088332553001237
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.004612062075466383,
          "train_loss": 0.005373936612159014,
          "validation_loss": 0.003656137967482209,
          "train_metric": 0.0005412431783042848,
          "validation_metric": 0.00036823280970565975,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.099631926999791
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.004854528946452774,
          "train_loss": 0.005887926556169987,
          "validation_loss": 0.003886669874191284,
          "train_metric": 0.0005930103943683207,
          "validation_metric": 0.00039145120535977185,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1111214520005888
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.00415214836539235,
          "train_loss": 0.0036135113332420588,
          "validation_loss": 0.0029300423339009285,
          "train_metric": 0.00036393964546732605,
          "validation_metric": 0.0002951031783595681,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.123027608000484
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0038432163346442394,
          "train_loss": 0.003684197086840868,
          "validation_loss": 0.0030791284516453743,
          "train_metric": 0.0003710588498506695,
          "validation_metric": 0.000310118542984128,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.137086820001059
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.004020751701318659,
          "train_loss": 0.0035994870122522116,
          "validation_loss": 0.003077118191868067,
          "train_metric": 0.0003625271492637694,
          "validation_metric": 0.00030991609673947096,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1490382310003042
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.003833814953395631,
          "train_loss": 0.003576416987925768,
          "validation_loss": 0.00294123450294137,
          "train_metric": 0.00036020364495925605,
          "validation_metric": 0.00029623034060932696,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.161215506001099
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.004615240060957149,
          "train_loss": 0.0037269913591444492,
          "validation_loss": 0.003283236874267459,
          "train_metric": 0.00037536892341449857,
          "validation_metric": 0.00033067556796595454,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.173006189001171
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.004220501679810695,
          "train_loss": 0.004076576326042414,
          "validation_loss": 0.0037421691231429577,
          "train_metric": 0.00041057783528231084,
          "validation_metric": 0.0003768975439015776,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.184707653001169
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.004329258263169322,
          "train_loss": 0.004544458817690611,
          "validation_loss": 0.0031697754748165607,
          "train_metric": 0.00045770121505483985,
          "validation_metric": 0.00031924815266393125,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1969670159996895
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0047735689440742135,
          "train_loss": 0.0035836626775562763,
          "validation_loss": 0.002900409745052457,
          "train_metric": 0.0003609333361964673,
          "validation_metric": 0.000292118638753891,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.209202718000597
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0053493724699364975,
          "train_loss": 0.003881285199895501,
          "validation_loss": 0.003135723527520895,
          "train_metric": 0.0003909087972715497,
          "validation_metric": 0.00031581861549057066,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.22083984100027
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.004835138701309916,
          "train_loss": 0.003501425962895155,
          "validation_loss": 0.0028126698452979326,
          "train_metric": 0.0003526508226059377,
          "validation_metric": 0.0002832818136084825,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2328434780010866
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0042148210923187435,
          "train_loss": 0.0034732420463114977,
          "validation_loss": 0.0028101939242333174,
          "train_metric": 0.0003498122387100011,
          "validation_metric": 0.00028303239378146827,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2448445280006126
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.004246993030392332,
          "train_loss": 0.0042596845887601376,
          "validation_loss": 0.003958098590373993,
          "train_metric": 0.00042901982669718564,
          "validation_metric": 0.00039864517748355865,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2566016300006595
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.00489171885419637,
          "train_loss": 0.00455753318965435,
          "validation_loss": 0.0042374636977910995,
          "train_metric": 0.0004590180760715157,
          "validation_metric": 0.00042678180034272373,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.268157575999794
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.005277245662000496,
          "train_loss": 0.004305545706301928,
          "validation_loss": 0.004021473228931427,
          "train_metric": 0.00043363880831748247,
          "validation_metric": 0.0004050280258525163,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2796373370001675
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.004978880861017387,
          "train_loss": 0.004296343307942152,
          "validation_loss": 0.004120469558984041,
          "train_metric": 0.00043271196773275733,
          "validation_metric": 0.00041499859071336687,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2911614920012653
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.00468597786675673,
          "train_loss": 0.005980344954878092,
          "validation_loss": 0.005888110026717186,
          "train_metric": 0.0006023184978403151,
          "validation_metric": 0.0005930289044044912,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.302629194000474
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.004660628270357847,
          "train_loss": 0.0037582367658615112,
          "validation_loss": 0.0033440175466239452,
          "train_metric": 0.00037851586239412427,
          "validation_metric": 0.00033679718035273254,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.3142116330000135
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.003654991965959198,
          "train_loss": 0.003443751484155655,
          "validation_loss": 0.002804565941914916,
          "train_metric": 0.00034684204729273915,
          "validation_metric": 0.0002824655966833234,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.3263652500008902
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0035429956114967354,
          "train_loss": 0.0034250919707119465,
          "validation_loss": 0.0027722795493900776,
          "train_metric": 0.00034496275475248694,
          "validation_metric": 0.00027921388391405344,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.338506705000327
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0033964433387154713,
          "train_loss": 0.0033373748883605003,
          "validation_loss": 0.0027484381571412086,
          "train_metric": 0.00033612814149819314,
          "validation_metric": 0.00027681264327839017,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.3503884080000716
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0034313725991523825,
          "train_loss": 0.003315108362585306,
          "validation_loss": 0.0027342080138623714,
          "train_metric": 0.00033388557494618,
          "validation_metric": 0.0002753794251475483,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.3633469930009596
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.003363756331964396,
          "train_loss": 0.003332807682454586,
          "validation_loss": 0.0026358547620475292,
          "train_metric": 0.0003356681845616549,
          "validation_metric": 0.000265473616309464,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.375597154001298
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.003394447368918918,
          "train_loss": 0.0033093756064772606,
          "validation_loss": 0.0026755346916615963,
          "train_metric": 0.00033330818405374885,
          "validation_metric": 0.0002694700378924608,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.388641004999954
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0034272678967681713,
          "train_loss": 0.003337598405778408,
          "validation_loss": 0.0028037067968398333,
          "train_metric": 0.0003361506969667971,
          "validation_metric": 0.00028237910009920597,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.4003874090012687
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0033999441366177052,
          "train_loss": 0.003299613483250141,
          "validation_loss": 0.002679666271433234,
          "train_metric": 0.000332324969349429,
          "validation_metric": 0.00026988619356416166,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.4124195460008195
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0033233701724384446,
          "train_loss": 0.0032929445151239634,
          "validation_loss": 0.0027115249540656805,
          "train_metric": 0.00033165334025397897,
          "validation_metric": 0.000273094919975847,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.425606967000931
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0034257134320796467,
          "train_loss": 0.0033302553929388523,
          "validation_loss": 0.0028183499816805124,
          "train_metric": 0.0003354111104272306,
          "validation_metric": 0.0002838539076037705,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.438122558000032
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0033908312434505206,
          "train_loss": 0.0033183980267494917,
          "validation_loss": 0.0026445682160556316,
          "train_metric": 0.0003342169220559299,
          "validation_metric": 0.0002663512132130563,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.4497610650014394
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.003420515506149968,
          "train_loss": 0.003305212827399373,
          "validation_loss": 0.0027373998891562223,
          "train_metric": 0.00033288897247985005,
          "validation_metric": 0.00027570087695494294,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.462042035000195
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0034177226771134883,
          "train_loss": 0.003311052918434143,
          "validation_loss": 0.002621709369122982,
          "train_metric": 0.00033347710268571973,
          "validation_metric": 0.00026404898380860686,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.474418321000485
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0033546494451002218,
          "train_loss": 0.0032826897222548723,
          "validation_loss": 0.0027248242404311895,
          "train_metric": 0.00033062047441489995,
          "validation_metric": 0.0002744343364611268,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.4865733470014675
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0034208131837658584,
          "train_loss": 0.003283974016085267,
          "validation_loss": 0.002654506592079997,
          "train_metric": 0.00033074981183744967,
          "validation_metric": 0.00026735218125395477,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.498038562000147
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0034606509434524924,
          "train_loss": 0.003281248267740011,
          "validation_loss": 0.002706217812374234,
          "train_metric": 0.0003304753336124122,
          "validation_metric": 0.0002725603408180177,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.510006197000621
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0033839900061138906,
          "train_loss": 0.003271109890192747,
          "validation_loss": 0.002624517073854804,
          "train_metric": 0.0003294541675131768,
          "validation_metric": 0.00026433172752149403,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.522061379000661
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0033740096987457946,
          "train_loss": 0.003294023685157299,
          "validation_loss": 0.002606873167678714,
          "train_metric": 0.0003317619557492435,
          "validation_metric": 0.00026255473494529724,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.5341824930001167
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.003501383867842378,
          "train_loss": 0.003392380429431796,
          "validation_loss": 0.002975517651066184,
          "train_metric": 0.0003416682011447847,
          "validation_metric": 0.0002996832481585443,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.545911551000245
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0033806956562330015,
          "train_loss": 0.0032600394915789366,
          "validation_loss": 0.0026311965193599463,
          "train_metric": 0.00032833919976837933,
          "validation_metric": 0.00026500446256250143,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.5610291400007554
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.003466205540462397,
          "train_loss": 0.0032790200784802437,
          "validation_loss": 0.0027799999807029963,
          "train_metric": 0.00033025085576809943,
          "validation_metric": 0.00027999142184853554,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.572697527000855
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0032813841535244137,
          "train_loss": 0.0033235230948776007,
          "validation_loss": 0.00254557142034173,
          "train_metric": 0.00033473310759291053,
          "validation_metric": 0.0002563806192483753,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.5848226039997826
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0032888137284317054,
          "train_loss": 0.0032291223760694265,
          "validation_loss": 0.0026304367929697037,
          "train_metric": 0.0003252253809478134,
          "validation_metric": 0.0002649279485922307,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.5979692840010102
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0032690023581380956,
          "train_loss": 0.003236300777643919,
          "validation_loss": 0.0026543366257101297,
          "train_metric": 0.0003259483492001891,
          "validation_metric": 0.0002673350682016462,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.6097561290007434
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.003310160172986798,
          "train_loss": 0.0032251335214823484,
          "validation_loss": 0.0026602488942444324,
          "train_metric": 0.0003248236607760191,
          "validation_metric": 0.000267930532572791,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.62193914800082
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0032751587641541846,
          "train_loss": 0.0032287430949509144,
          "validation_loss": 0.0026411309372633696,
          "train_metric": 0.0003251871676184237,
          "validation_metric": 0.00026600505225360394,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.6337768730008975
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0032767695593065582,
          "train_loss": 0.0032596541568636894,
          "validation_loss": 0.0025702870916575193,
          "train_metric": 0.0003283004043623805,
          "validation_metric": 0.0002588698989711702,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.64503342900025
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0033360056186211295,
          "train_loss": 0.003217906691133976,
          "validation_loss": 0.00267424457706511,
          "train_metric": 0.00032409574487246573,
          "validation_metric": 0.0002693401474971324,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.656600445001459
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.003360993716341909,
          "train_loss": 0.0032395748421549797,
          "validation_loss": 0.0027502395678311586,
          "train_metric": 0.0003262780955992639,
          "validation_metric": 0.000276994047453627,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.6681487610003387
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0032944865197350737,
          "train_loss": 0.0032459385693073273,
          "validation_loss": 0.0025588716380298138,
          "train_metric": 0.000326919078361243,
          "validation_metric": 0.0002577201812528074,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.679862550001417
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0033557125061633997,
          "train_loss": 0.0032903742976486683,
          "validation_loss": 0.0028760770801454782,
          "train_metric": 0.0003313944907858968,
          "validation_metric": 0.0002896679216064513,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.6916029439998965
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0034638987672224175,
          "train_loss": 0.0032037878409028053,
          "validation_loss": 0.0026772862765938044,
          "train_metric": 0.0003226738190278411,
          "validation_metric": 0.0002696464362088591,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.703224623001006
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0032625313324388117,
          "train_loss": 0.0031887178774923086,
          "validation_loss": 0.002649936592206359,
          "train_metric": 0.0003211559378542006,
          "validation_metric": 0.0002668919041752815,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.714788672001305
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.0032212128426181152,
          "train_loss": 0.003185426350682974,
          "validation_loss": 0.002615133533254266,
          "train_metric": 0.0003208244452252984,
          "validation_metric": 0.00026338669704273343,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.726598555000237
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.003236273958464153,
          "train_loss": 0.0031818924471735954,
          "validation_loss": 0.002637655008584261,
          "train_metric": 0.0003204685926903039,
          "validation_metric": 0.00026565493317320943,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.7381437510011892
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0034170788421761245,
          "train_loss": 0.003176460275426507,
          "validation_loss": 0.0026054810732603073,
          "train_metric": 0.0003199214697815478,
          "validation_metric": 0.00026241448358632624,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.749799641000209
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.003371571674506413,
          "train_loss": 0.0032044786494225264,
          "validation_loss": 0.0026433393359184265,
          "train_metric": 0.0003227434353902936,
          "validation_metric": 0.0002662274637259543,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.7647884400012117
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.003808269524597563,
          "train_loss": 0.003583140904083848,
          "validation_loss": 0.003304982092231512,
          "train_metric": 0.0003608808619901538,
          "validation_metric": 0.0003328657185193151,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.7767183870000736
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.00379628679365851,
          "train_loss": 0.003284762380644679,
          "validation_loss": 0.0025702770799398422,
          "train_metric": 0.00033082926529459655,
          "validation_metric": 0.0002588688803371042,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.789589085999978
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.00343310424068477,
          "train_loss": 0.003190014511346817,
          "validation_loss": 0.0025181062519550323,
          "train_metric": 0.00032128661405295134,
          "validation_metric": 0.0002536144747864455,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.8024521300012566
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.003265115705289645,
          "train_loss": 0.0032136538065969944,
          "validation_loss": 0.002760939300060272,
          "train_metric": 0.000323667423799634,
          "validation_metric": 0.0002780717040877789,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.814476765001018
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0032789917713671457,
          "train_loss": 0.0031451918184757233,
          "validation_loss": 0.0025916248559951782,
          "train_metric": 0.000316772231599316,
          "validation_metric": 0.00026101895491592586,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.8273438310006895
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0032622651488054544,
          "train_loss": 0.0031454917043447495,
          "validation_loss": 0.002622887957841158,
          "train_metric": 0.0003168024413753301,
          "validation_metric": 0.0002641676983330399,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.8393582999997307
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0033295289176749066,
          "train_loss": 0.003271393943578005,
          "validation_loss": 0.0025309042539447546,
          "train_metric": 0.0003294828056823462,
          "validation_metric": 0.0002549033670220524,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.851510780001263
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0034171511069871485,
          "train_loss": 0.0031975845340639353,
          "validation_loss": 0.002690460067242384,
          "train_metric": 0.0003220489888917655,
          "validation_metric": 0.00027097330894321203,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.8634550370006764
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.003198142112523783,
          "train_loss": 0.0031379039864987135,
          "validation_loss": 0.0025767937768250704,
          "train_metric": 0.00031603817478753626,
          "validation_metric": 0.00025952522992156446,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.8750890999999683
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0031873438856564462,
          "train_loss": 0.003133350983262062,
          "validation_loss": 0.0025940779596567154,
          "train_metric": 0.0003155796439386904,
          "validation_metric": 0.00026126604643650353,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.8869881780010473
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.003193895499862265,
          "train_loss": 0.00316820852458477,
          "validation_loss": 0.0027068082708865404,
          "train_metric": 0.00031909035169519484,
          "validation_metric": 0.00027261979994364083,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.898889351999969
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.003249718291044701,
          "train_loss": 0.0031685743015259504,
          "validation_loss": 0.0024775576312094927,
          "train_metric": 0.0003191272262483835,
          "validation_metric": 0.00024953053798526525,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.9114441750007245
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0031820887234061956,
          "train_loss": 0.0031559327617287636,
          "validation_loss": 0.0026929208543151617,
          "train_metric": 0.0003178539627697319,
          "validation_metric": 0.0002712211571633816,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.9239702820013918
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0032849433046067134,
          "train_loss": 0.0031279786489903927,
          "validation_loss": 0.002529376884922385,
          "train_metric": 0.00031503854552283883,
          "validation_metric": 0.0002547496114857495,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.9356811360012216
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0031643665715819225,
          "train_loss": 0.0031211848836392164,
          "validation_loss": 0.0025920832995325327,
          "train_metric": 0.0003143542562611401,
          "validation_metric": 0.0002610651426948607,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.9473581470010686
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.003185617570125032,
          "train_loss": 0.0031179580837488174,
          "validation_loss": 0.0024945305194705725,
          "train_metric": 0.00031402934109792113,
          "validation_metric": 0.00025123998057097197,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.9625634270014416
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.003202807594789192,
          "train_loss": 0.003098254557698965,
          "validation_loss": 0.002537110587581992,
          "train_metric": 0.00031204483821056783,
          "validation_metric": 0.00025552845909260213,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.9740511600011814
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.003168398434354458,
          "train_loss": 0.00310004991479218,
          "validation_loss": 0.002602499211207032,
          "train_metric": 0.000312225689413026,
          "validation_metric": 0.0002621142193675041,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 2.9871428340011335
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0032333881681552157,
          "train_loss": 0.0031497436575591564,
          "validation_loss": 0.0027282540686428547,
          "train_metric": 0.00031723061692900956,
          "validation_metric": 0.0002747797698248178,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.0828527530011343
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.003287018567789346,
          "train_loss": 0.0032387461978942156,
          "validation_loss": 0.002475345740094781,
          "train_metric": 0.0003261946840211749,
          "validation_metric": 0.0002493077772669494,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.0954122690000077
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0032694537367206067,
          "train_loss": 0.0031459047459065914,
          "validation_loss": 0.0027048385236412287,
          "train_metric": 0.0003168440016452223,
          "validation_metric": 0.0002724214573390782,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.107357019000119
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0032005063840188086,
          "train_loss": 0.0031651898752897978,
          "validation_loss": 0.002466502133756876,
          "train_metric": 0.00031878636218607426,
          "validation_metric": 0.0002484170545358211,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.1199622780004574
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.003237259814341087,
          "train_loss": 0.0030692946165800095,
          "validation_loss": 0.0025095450691878796,
          "train_metric": 0.0003091281687375158,
          "validation_metric": 0.0002527521864976734,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.1323586569997133
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0032446420373162255,
          "train_loss": 0.0031107838731259108,
          "validation_loss": 0.0024399454705417156,
          "train_metric": 0.00031330675119534135,
          "validation_metric": 0.00024574241251684725,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.14623144899997
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0033037856919690967,
          "train_loss": 0.003122161142528057,
          "validation_loss": 0.002464447170495987,
          "train_metric": 0.0003144526563119143,
          "validation_metric": 0.00024821009719744325,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.1626045190005243
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0031705431174486876,
          "train_loss": 0.003077561967074871,
          "validation_loss": 0.0025155600160360336,
          "train_metric": 0.00030996071291156113,
          "validation_metric": 0.00025335801183246076,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.1741981680006575
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.003296022907306906,
          "train_loss": 0.0032985000871121883,
          "validation_loss": 0.0024662360083311796,
          "train_metric": 0.0003322128322906792,
          "validation_metric": 0.0002483902790118009,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.188286018999861
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.003344375792948995,
          "train_loss": 0.0031045624054968357,
          "validation_loss": 0.002671007066965103,
          "train_metric": 0.0003126801166217774,
          "validation_metric": 0.00026901400997303426,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2002205530006904
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0031062286652741022,
          "train_loss": 0.0030456765089184046,
          "validation_loss": 0.0024706211406737566,
          "train_metric": 0.00030674939625896513,
          "validation_metric": 0.0002488319296389818,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.211881209001149
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0031421139210578986,
          "train_loss": 0.003040972398594022,
          "validation_loss": 0.002500698668882251,
          "train_metric": 0.00030627561500295997,
          "validation_metric": 0.00025186120183207095,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2241906720009865
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.003213242380297743,
          "train_loss": 0.0030759014189243317,
          "validation_loss": 0.0024529160000383854,
          "train_metric": 0.0003097935696132481,
          "validation_metric": 0.0002470487088430673,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2363248159999785
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0031773574082762934,
          "train_loss": 0.003044589888304472,
          "validation_loss": 0.0024653819855302572,
          "train_metric": 0.00030663993675261736,
          "validation_metric": 0.0002483042189851403,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2478831610005727
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0033336397173115984,
          "train_loss": 0.003172302385792136,
          "validation_loss": 0.0028079014737159014,
          "train_metric": 0.0003195027238689363,
          "validation_metric": 0.00028280154219828546,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2608994300007907
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.003164140034641605,
          "train_loss": 0.003134344471618533,
          "validation_loss": 0.002412977861240506,
          "train_metric": 0.00031567973201163113,
          "validation_metric": 0.00024302626843564212,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2753801500002737
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0031818489223951474,
          "train_loss": 0.003022608580067754,
          "validation_loss": 0.0024668031837791204,
          "train_metric": 0.00030442606657743454,
          "validation_metric": 0.00024844740983098745,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2874933470011456
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.003229128022212535,
          "train_loss": 0.00306120328605175,
          "validation_loss": 0.0026365017984062433,
          "train_metric": 0.00030831314506940544,
          "validation_metric": 0.0002655388379935175,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.2989542140003323
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0031447452056454495,
          "train_loss": 0.0030399400275200605,
          "validation_loss": 0.0026055159978568554,
          "train_metric": 0.00030617162701673806,
          "validation_metric": 0.00026241803425364196,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.3107680949997302
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.003158588799124118,
          "train_loss": 0.0030133032705634832,
          "validation_loss": 0.0025319145061075687,
          "train_metric": 0.00030348889413289726,
          "validation_metric": 0.0002550051431171596,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.3226422980005736
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.0031350676727015525,
          "train_loss": 0.0030068857595324516,
          "validation_loss": 0.002464682562276721,
          "train_metric": 0.00030284246895462275,
          "validation_metric": 0.0002482338168192655,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.334023020999666
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.0030222645509638824,
          "train_loss": 0.002992362482473254,
          "validation_loss": 0.0024855530355125666,
          "train_metric": 0.0003013797686435282,
          "validation_metric": 0.0002503358409740031,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.346699081001134
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.003086579679802526,
          "train_loss": 0.003034101566299796,
          "validation_loss": 0.0023783822543919086,
          "train_metric": 0.0003055836132261902,
          "validation_metric": 0.00023954191419761628,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.3624220630008494
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.0030856430021231063,
          "train_loss": 0.003048783401027322,
          "validation_loss": 0.0026010903529822826,
          "train_metric": 0.0003070622915402055,
          "validation_metric": 0.00026197227998636663,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.3755676440014213
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.003007641469594091,
          "train_loss": 0.0030399784445762634,
          "validation_loss": 0.002386159962043166,
          "train_metric": 0.00030617552693001926,
          "validation_metric": 0.00024032527289818972,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.387598956000147
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.003076895125559531,
          "train_loss": 0.0030128268990665674,
          "validation_loss": 0.0025553342420607805,
          "train_metric": 0.0003034409019164741,
          "validation_metric": 0.0002573639212641865,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.3995311580001726
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.003123821079498157,
          "train_loss": 0.0030177466105669737,
          "validation_loss": 0.0025965976528823376,
          "train_metric": 0.00030393636552616954,
          "validation_metric": 0.00026151983183808625,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.4113527389999945
        }
      ],
      "best": {
        "epoch": 236,
        "updates": 3776
      },
      "best_validation_metric": 0.00023954191419761628,
      "config": {
        "label": "Step decay",
        "batch": 64,
        "lr": 0.003,
        "schedule": "step",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.015477141667556586
    },
    "449a997d9d64_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.9943018145859241,
          "train_loss": 0.9896844029426575,
          "validation_loss": 0.7139147520065308,
          "train_metric": 0.09967739135026932,
          "validation_metric": 0.07190287858247757,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.018347197999901255
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.9853110238909721,
          "train_loss": 0.9803768396377563,
          "validation_loss": 0.7076583504676819,
          "train_metric": 0.09873996675014496,
          "validation_metric": 0.07127276062965393,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.03662215400072455
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.9755979813635349,
          "train_loss": 0.9702013731002808,
          "validation_loss": 0.7006769776344299,
          "train_metric": 0.09771512448787689,
          "validation_metric": 0.07056961953639984,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.05500254899925494
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.9649721086025238,
          "train_loss": 0.9587562084197998,
          "validation_loss": 0.6929641366004944,
          "train_metric": 0.09656240791082382,
          "validation_metric": 0.06979280710220337,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.07222344799993152
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.952792976051569,
          "train_loss": 0.9458925127983093,
          "validation_loss": 0.684228241443634,
          "train_metric": 0.09526683390140533,
          "validation_metric": 0.06891295313835144,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.08903942500000994
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.939091183245182,
          "train_loss": 0.9309921264648438,
          "validation_loss": 0.6739834547042847,
          "train_metric": 0.09376612305641174,
          "validation_metric": 0.06788115203380585,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.1064245289999235
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.9232434555888176,
          "train_loss": 0.9139925837516785,
          "validation_loss": 0.6623551845550537,
          "train_metric": 0.09205397963523865,
          "validation_metric": 0.0667099803686142,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.1235118410004361
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.9051029123365879,
          "train_loss": 0.8944838643074036,
          "validation_loss": 0.6490920782089233,
          "train_metric": 0.09008914232254028,
          "validation_metric": 0.06537417322397232,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.1409406239999953
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.8839896023273468,
          "train_loss": 0.8725901246070862,
          "validation_loss": 0.6340537667274475,
          "train_metric": 0.0878840833902359,
          "validation_metric": 0.06385958194732666,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.15813874299965391
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.8607098124921322,
          "train_loss": 0.8467447757720947,
          "validation_loss": 0.6165114045143127,
          "train_metric": 0.08528102934360504,
          "validation_metric": 0.062092773616313934,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.1756427039999835
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.833630483597517,
          "train_loss": 0.8178194165229797,
          "validation_loss": 0.5966500043869019,
          "train_metric": 0.08236777782440186,
          "validation_metric": 0.06009240821003914,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.19317009000042162
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.8029162660241127,
          "train_loss": 0.785693883895874,
          "validation_loss": 0.5744593739509583,
          "train_metric": 0.07913221418857574,
          "validation_metric": 0.0578574538230896,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.21037517900003877
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.768752682954073,
          "train_loss": 0.749731719493866,
          "validation_loss": 0.549737811088562,
          "train_metric": 0.07551023364067078,
          "validation_metric": 0.05536757409572601,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.22779545500088716
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.7312517017126083,
          "train_loss": 0.7095880508422852,
          "validation_loss": 0.5222169160842896,
          "train_metric": 0.07146710902452469,
          "validation_metric": 0.05259577929973602,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.3419582940005057
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.6895790696144104,
          "train_loss": 0.6662587523460388,
          "validation_loss": 0.492220014333725,
          "train_metric": 0.06710314005613327,
          "validation_metric": 0.049574606120586395,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.3608661189991835
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.645062904804945,
          "train_loss": 0.6196703314781189,
          "validation_loss": 0.4599333703517914,
          "train_metric": 0.06241092085838318,
          "validation_metric": 0.04632280394434929,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.37832689599963487
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.5969773307442665,
          "train_loss": 0.5710557699203491,
          "validation_loss": 0.42596492171287537,
          "train_metric": 0.057514648884534836,
          "validation_metric": 0.042901623994112015,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.39610580199951073
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.5477220509201288,
          "train_loss": 0.5197492837905884,
          "validation_loss": 0.3900874853134155,
          "train_metric": 0.05234724283218384,
          "validation_metric": 0.03928818181157112,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.4133682569990924
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.49581429176032543,
          "train_loss": 0.4684082865715027,
          "validation_loss": 0.35427379608154297,
          "train_metric": 0.047176361083984375,
          "validation_metric": 0.03568115830421448,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.43122048300028837
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.44455941393971443,
          "train_loss": 0.4168574810028076,
          "validation_loss": 0.3178141415119171,
          "train_metric": 0.04198436066508293,
          "validation_metric": 0.032009076327085495,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.44866068400006043
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.39362641610205173,
          "train_loss": 0.36742857098579407,
          "validation_loss": 0.28288158774375916,
          "train_metric": 0.037006065249443054,
          "validation_metric": 0.02849079668521881,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.4664586260005308
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.34596330113708973,
          "train_loss": 0.3201301693916321,
          "validation_loss": 0.24913088977336884,
          "train_metric": 0.032242339104413986,
          "validation_metric": 0.02509154938161373,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.48383420199934335
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.3001038869842887,
          "train_loss": 0.278373658657074,
          "validation_loss": 0.21898077428340912,
          "train_metric": 0.028036778792738914,
          "validation_metric": 0.022054942324757576,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.5196634339990851
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.26044175308197737,
          "train_loss": 0.2409510463476181,
          "validation_loss": 0.19200821220874786,
          "train_metric": 0.024267705157399178,
          "validation_metric": 0.019338365644216537,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.536730271000124
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.2258833944797516,
          "train_loss": 0.20934198796749115,
          "validation_loss": 0.1689254641532898,
          "train_metric": 0.02108415588736534,
          "validation_metric": 0.0170135535299778,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.5545894789993326
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.19696312211453915,
          "train_loss": 0.1842518299818039,
          "validation_loss": 0.15037813782691956,
          "train_metric": 0.018557168543338776,
          "validation_metric": 0.01514553651213646,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.5722772539993457
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.17439176933839917,
          "train_loss": 0.1648680865764618,
          "validation_loss": 0.1359533816576004,
          "train_metric": 0.01660490781068802,
          "validation_metric": 0.013692726381123066,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.5896299649994035
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.15752068534493446,
          "train_loss": 0.1503371000289917,
          "validation_loss": 0.12504246830940247,
          "train_metric": 0.015141400508582592,
          "validation_metric": 0.012593818828463554,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6076897040002223
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.14525999780744314,
          "train_loss": 0.1401074081659317,
          "validation_loss": 0.1172628328204155,
          "train_metric": 0.014111104421317577,
          "validation_metric": 0.011810284107923508,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6251089060006052
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.13670429680496454,
          "train_loss": 0.13324607908725739,
          "validation_loss": 0.11196985840797424,
          "train_metric": 0.01342005655169487,
          "validation_metric": 0.011277193203568459,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6425224630002049
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.13094539754092693,
          "train_loss": 0.12860429286956787,
          "validation_loss": 0.10829285532236099,
          "train_metric": 0.012952551245689392,
          "validation_metric": 0.01090686023235321,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6603515089991561
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.12710986100137234,
          "train_loss": 0.1253206580877304,
          "validation_loss": 0.10571585595607758,
          "train_metric": 0.012621837668120861,
          "validation_metric": 0.010647313669323921,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6782986880007229
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.12421329878270626,
          "train_loss": 0.12306449562311172,
          "validation_loss": 0.10386376082897186,
          "train_metric": 0.012394605204463005,
          "validation_metric": 0.010460779070854187,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6957225749993086
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.12229849258437753,
          "train_loss": 0.12113011628389359,
          "validation_loss": 0.10223116725683212,
          "train_metric": 0.012199783697724342,
          "validation_metric": 0.010296348482370377,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7132936020007037
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.12042548810131848,
          "train_loss": 0.1195594072341919,
          "validation_loss": 0.10088501125574112,
          "train_metric": 0.012041585519909859,
          "validation_metric": 0.010160769335925579,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7309316659993783
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.11891111196018755,
          "train_loss": 0.11813880503177643,
          "validation_loss": 0.09966189414262772,
          "train_metric": 0.011898508295416832,
          "validation_metric": 0.01003758143633604,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7485184640008811
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.11754736071452498,
          "train_loss": 0.11678655445575714,
          "validation_loss": 0.09847135096788406,
          "train_metric": 0.011762314476072788,
          "validation_metric": 0.009917673654854298,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7658785050007282
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.11619517067447305,
          "train_loss": 0.1154809445142746,
          "validation_loss": 0.09730514138936996,
          "train_metric": 0.011630818247795105,
          "validation_metric": 0.009800217114388943,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7830807940008526
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.1150668771006167,
          "train_loss": 0.11414261907339096,
          "validation_loss": 0.09612337499856949,
          "train_metric": 0.011496026068925858,
          "validation_metric": 0.00968119315803051,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8000981250006589
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.11369487992487848,
          "train_loss": 0.11297142505645752,
          "validation_loss": 0.09505239129066467,
          "train_metric": 0.01137806847691536,
          "validation_metric": 0.00957332830876112,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8174350600002072
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.11244331672787666,
          "train_loss": 0.11173751205205917,
          "validation_loss": 0.09394705295562744,
          "train_metric": 0.011253793723881245,
          "validation_metric": 0.009462003596127033,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8348492570003145
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.11132497410289943,
          "train_loss": 0.11051861941814423,
          "validation_loss": 0.09283256530761719,
          "train_metric": 0.011131029576063156,
          "validation_metric": 0.009349755011498928,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8523048320002999
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.11005886225029826,
          "train_loss": 0.10934886336326599,
          "validation_loss": 0.09177695959806442,
          "train_metric": 0.011013217270374298,
          "validation_metric": 0.009243438951671124,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8693022850002308
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.10881931520998478,
          "train_loss": 0.10816773772239685,
          "validation_loss": 0.09070353955030441,
          "train_metric": 0.010894258506596088,
          "validation_metric": 0.009135328233242035,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8866219909996289
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.107677893480286,
          "train_loss": 0.10704832524061203,
          "validation_loss": 0.08967214822769165,
          "train_metric": 0.010781515389680862,
          "validation_metric": 0.009031449444591999,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9041509950002364
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.10655733663588762,
          "train_loss": 0.10591863095760345,
          "validation_loss": 0.08864213526248932,
          "train_metric": 0.010667736642062664,
          "validation_metric": 0.00892771128565073,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9218851559999166
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.10541942459531128,
          "train_loss": 0.10479678213596344,
          "validation_loss": 0.08761441707611084,
          "train_metric": 0.010554749518632889,
          "validation_metric": 0.008824203163385391,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9535111719997076
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.10434031998738647,
          "train_loss": 0.10374023765325546,
          "validation_loss": 0.08663442730903625,
          "train_metric": 0.010448336601257324,
          "validation_metric": 0.008725503459572792,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9991353780005738
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.10332725034095347,
          "train_loss": 0.10266448557376862,
          "validation_loss": 0.085646852850914,
          "train_metric": 0.010339991189539433,
          "validation_metric": 0.00862603634595871,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0173250080006255
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.10214825300499797,
          "train_loss": 0.10155133903026581,
          "validation_loss": 0.08464055508375168,
          "train_metric": 0.010227877646684647,
          "validation_metric": 0.00852468702942133,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.034782377000738
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.10108291613869369,
          "train_loss": 0.1005076915025711,
          "validation_loss": 0.08367742598056793,
          "train_metric": 0.010122767649590969,
          "validation_metric": 0.008427683264017105,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0518852509994758
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.10004622023552656,
          "train_loss": 0.09944866597652435,
          "validation_loss": 0.08270899206399918,
          "train_metric": 0.010016105137765408,
          "validation_metric": 0.008330147713422775,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0689074130004883
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.09902641410008073,
          "train_loss": 0.09839694201946259,
          "validation_loss": 0.08173978328704834,
          "train_metric": 0.009910179302096367,
          "validation_metric": 0.008232531137764454,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0868303070001275
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.09796786750666797,
          "train_loss": 0.09742075204849243,
          "validation_loss": 0.08082297444343567,
          "train_metric": 0.009811860509216785,
          "validation_metric": 0.008140193298459053,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1048751329999504
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.09700845275074244,
          "train_loss": 0.0963837057352066,
          "validation_loss": 0.07988496869802475,
          "train_metric": 0.009707414545118809,
          "validation_metric": 0.008045720867812634,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1225891220001358
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.09596755914390087,
          "train_loss": 0.0953906998038292,
          "validation_loss": 0.0789724811911583,
          "train_metric": 0.009607401676476002,
          "validation_metric": 0.007953818887472153,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1399894019996282
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.09498536726459861,
          "train_loss": 0.09443250298500061,
          "validation_loss": 0.07808508723974228,
          "train_metric": 0.009510897099971771,
          "validation_metric": 0.007864444516599178,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1620959869997023
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.09414595481939614,
          "train_loss": 0.09344843029975891,
          "validation_loss": 0.0771893635392189,
          "train_metric": 0.009411783888936043,
          "validation_metric": 0.007774230092763901,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.209498103000442
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.09306749817915261,
          "train_loss": 0.09249792993068695,
          "validation_loss": 0.07630752772092819,
          "train_metric": 0.009316053241491318,
          "validation_metric": 0.007685413584113121,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2520016090002173
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.09212745982222259,
          "train_loss": 0.09156220406293869,
          "validation_loss": 0.07544578611850739,
          "train_metric": 0.009221809916198254,
          "validation_metric": 0.00759862270206213,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2954219559997
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.09126117033883929,
          "train_loss": 0.09069523215293884,
          "validation_loss": 0.07463369518518448,
          "train_metric": 0.009134490974247456,
          "validation_metric": 0.007516831625252962,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3439200590000837
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.09029121813364327,
          "train_loss": 0.08973676711320877,
          "validation_loss": 0.07378701865673065,
          "train_metric": 0.009037958458065987,
          "validation_metric": 0.007431558333337307,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3873741200004588
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.08947839005850255,
          "train_loss": 0.08884567767381668,
          "validation_loss": 0.0729583278298378,
          "train_metric": 0.008948211558163166,
          "validation_metric": 0.007348095532506704,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4286872659995424
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0884581666905433,
          "train_loss": 0.08797237277030945,
          "validation_loss": 0.07217346131801605,
          "train_metric": 0.008860254660248756,
          "validation_metric": 0.007269047200679779,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4768147279992263
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.08760111359879375,
          "train_loss": 0.08711306750774384,
          "validation_loss": 0.07137975841760635,
          "train_metric": 0.008773707784712315,
          "validation_metric": 0.007189107593148947,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.537542829000813
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0867364399600774,
          "train_loss": 0.08623617142438889,
          "validation_loss": 0.0705832690000534,
          "train_metric": 0.008685391396284103,
          "validation_metric": 0.0071088881231844425,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.5861589979995188
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.08586725371424109,
          "train_loss": 0.08537875860929489,
          "validation_loss": 0.06980658322572708,
          "train_metric": 0.008599034510552883,
          "validation_metric": 0.007030664011836052,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.607064932000867
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.08515517483465374,
          "train_loss": 0.08454187214374542,
          "validation_loss": 0.06904791295528412,
          "train_metric": 0.008514747954905033,
          "validation_metric": 0.006954252254217863,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6253940419992432
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.08429324068129063,
          "train_loss": 0.08375386148691177,
          "validation_loss": 0.06833390146493912,
          "train_metric": 0.008435382507741451,
          "validation_metric": 0.006882340181618929,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6439259750004567
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.08345717843621969,
          "train_loss": 0.08297674357891083,
          "validation_loss": 0.06761575490236282,
          "train_metric": 0.008357114158570766,
          "validation_metric": 0.006810010876506567,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.662696851999499
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.08269387320615351,
          "train_loss": 0.08218131214380264,
          "validation_loss": 0.06690255552530289,
          "train_metric": 0.008277000859379768,
          "validation_metric": 0.00673817889764905,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.680952590000743
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.08182024280540645,
          "train_loss": 0.08141115307807922,
          "validation_loss": 0.06622083485126495,
          "train_metric": 0.008199433796107769,
          "validation_metric": 0.0066695199348032475,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.698464604000037
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.08108759624883533,
          "train_loss": 0.08066307008266449,
          "validation_loss": 0.06555499136447906,
          "train_metric": 0.008124088868498802,
          "validation_metric": 0.006602459121495485,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.716131466000661
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.08050170028582215,
          "train_loss": 0.07992199808359146,
          "validation_loss": 0.06489024311304092,
          "train_metric": 0.008049449883401394,
          "validation_metric": 0.006535508204251528,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7333776720006426
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.07958854944445193,
          "train_loss": 0.07920457422733307,
          "validation_loss": 0.0642402321100235,
          "train_metric": 0.007977194152772427,
          "validation_metric": 0.00647004134953022,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7511647060000541
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.07897405209951103,
          "train_loss": 0.07850033044815063,
          "validation_loss": 0.06362123042345047,
          "train_metric": 0.007906265556812286,
          "validation_metric": 0.006407698150724173,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7693748580004467
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.07829839317128062,
          "train_loss": 0.07784141600131989,
          "validation_loss": 0.06302309781312943,
          "train_metric": 0.007839903235435486,
          "validation_metric": 0.006347456015646458,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7870221510002011
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.07762519526295364,
          "train_loss": 0.07717825472354889,
          "validation_loss": 0.062436532229185104,
          "train_metric": 0.007773112505674362,
          "validation_metric": 0.006288378033787012,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.804068683000878
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0769558111205697,
          "train_loss": 0.07650423049926758,
          "validation_loss": 0.06185693293809891,
          "train_metric": 0.007705227006226778,
          "validation_metric": 0.0062300036661326885,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.821285469999566
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.07629563694354147,
          "train_loss": 0.07592993229627609,
          "validation_loss": 0.06134214997291565,
          "train_metric": 0.00764738442376256,
          "validation_metric": 0.006178156938403845,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8383161180008756
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.07565130898728967,
          "train_loss": 0.07528962194919586,
          "validation_loss": 0.06078570336103439,
          "train_metric": 0.007582894992083311,
          "validation_metric": 0.006122113671153784,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8561745410006552
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.07505030801985413,
          "train_loss": 0.07468125969171524,
          "validation_loss": 0.060257986187934875,
          "train_metric": 0.007521623279899359,
          "validation_metric": 0.0060689630918204784,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.87375370799964
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.07447201630566269,
          "train_loss": 0.07409517467021942,
          "validation_loss": 0.05974483862519264,
          "train_metric": 0.007462593726813793,
          "validation_metric": 0.006017281673848629,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8917067149996
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.07396993669681251,
          "train_loss": 0.07352398335933685,
          "validation_loss": 0.05925152450799942,
          "train_metric": 0.007405066397041082,
          "validation_metric": 0.005967597011476755,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9090832029996818
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.07326678861863911,
          "train_loss": 0.07298067212104797,
          "validation_loss": 0.05877823755145073,
          "train_metric": 0.007350345607846975,
          "validation_metric": 0.0059199281968176365,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9266487349996169
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.07284632208757102,
          "train_loss": 0.07249709963798523,
          "validation_loss": 0.058365415781736374,
          "train_metric": 0.007301642559468746,
          "validation_metric": 0.00587835069745779,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.94427239200013
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.07232128956820816,
          "train_loss": 0.07194451242685318,
          "validation_loss": 0.05790150538086891,
          "train_metric": 0.0072459876537323,
          "validation_metric": 0.005831628106534481,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9629719389995444
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.07172190048731863,
          "train_loss": 0.07144371420145035,
          "validation_loss": 0.05747826024889946,
          "train_metric": 0.007195549085736275,
          "validation_metric": 0.005789000075310469,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9808248040008039
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.07126747816801071,
          "train_loss": 0.070955790579319,
          "validation_loss": 0.05705316364765167,
          "train_metric": 0.007146407384425402,
          "validation_metric": 0.005746185779571533,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9985904469995148
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.07079767412506044,
          "train_loss": 0.07047431170940399,
          "validation_loss": 0.05665476992726326,
          "train_metric": 0.007097914349287748,
          "validation_metric": 0.005706061143428087,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.015700134999861
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.07033652998507023,
          "train_loss": 0.0700506716966629,
          "validation_loss": 0.05629298463463783,
          "train_metric": 0.007055248133838177,
          "validation_metric": 0.005669623613357544,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0325970579997374
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0698581961914897,
          "train_loss": 0.06960249692201614,
          "validation_loss": 0.05592990666627884,
          "train_metric": 0.007010108791291714,
          "validation_metric": 0.005633055232465267,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0716574660000333
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.06959370081312954,
          "train_loss": 0.0691768005490303,
          "validation_loss": 0.0555720180273056,
          "train_metric": 0.006967233493924141,
          "validation_metric": 0.005597009789198637,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0924203659997147
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.06903745140880346,
          "train_loss": 0.0687522292137146,
          "validation_loss": 0.055232059210538864,
          "train_metric": 0.006924472749233246,
          "validation_metric": 0.005562771577388048,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.109620351000558
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.06872752972412854,
          "train_loss": 0.06840752065181732,
          "validation_loss": 0.054942887276411057,
          "train_metric": 0.006889754440635443,
          "validation_metric": 0.005533645860850811,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.128124394999759
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.06831621308811009,
          "train_loss": 0.06798596680164337,
          "validation_loss": 0.05460682883858681,
          "train_metric": 0.00684729777276516,
          "validation_metric": 0.005499800201505423,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.1460585389995686
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.06788001488894224,
          "train_loss": 0.0676264613866806,
          "validation_loss": 0.054305870085954666,
          "train_metric": 0.006811089348047972,
          "validation_metric": 0.005469487980008125,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.163649111000268
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.06763400242198259,
          "train_loss": 0.06726036220788956,
          "validation_loss": 0.05401407554745674,
          "train_metric": 0.006774216890335083,
          "validation_metric": 0.005440100096166134,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.180863247000161
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.06715502147562802,
          "train_loss": 0.0669270008802414,
          "validation_loss": 0.05375130847096443,
          "train_metric": 0.00674064178019762,
          "validation_metric": 0.0054136342369019985,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.19833600200036
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.06689434126019478,
          "train_loss": 0.06662080436944962,
          "validation_loss": 0.053504422307014465,
          "train_metric": 0.006709802895784378,
          "validation_metric": 0.0053887697868049145,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.21595944100045
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.06649691483471543,
          "train_loss": 0.06628832221031189,
          "validation_loss": 0.05323585867881775,
          "train_metric": 0.006676316726952791,
          "validation_metric": 0.00536172091960907,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.250847564999276
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.06619086884893477,
          "train_loss": 0.06597143411636353,
          "validation_loss": 0.0529804565012455,
          "train_metric": 0.006644401699304581,
          "validation_metric": 0.00533599779009819,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.290158686999348
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.06591652811039239,
          "train_loss": 0.0656813457608223,
          "validation_loss": 0.05276951193809509,
          "train_metric": 0.006615184247493744,
          "validation_metric": 0.0053147510625422,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.3375108119998913
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.06558107049204409,
          "train_loss": 0.0653940737247467,
          "validation_loss": 0.05255172774195671,
          "train_metric": 0.0065862517803907394,
          "validation_metric": 0.005292817484587431,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.3788141250006447
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0653142910450697,
          "train_loss": 0.06511838734149933,
          "validation_loss": 0.052338942885398865,
          "train_metric": 0.0065584853291511536,
          "validation_metric": 0.005271386355161667,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.396321797999917
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.06511762458831072,
          "train_loss": 0.06482289731502533,
          "validation_loss": 0.05209053307771683,
          "train_metric": 0.0065287258476018906,
          "validation_metric": 0.005246368236839771,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4138181540001824
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.06475476745981723,
          "train_loss": 0.06455384939908981,
          "validation_loss": 0.051893200725317,
          "train_metric": 0.006501627620309591,
          "validation_metric": 0.005226492881774902,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4318619540008513
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0644943678053096,
          "train_loss": 0.06429731845855713,
          "validation_loss": 0.05170604959130287,
          "train_metric": 0.006475790869444609,
          "validation_metric": 0.005207644775509834,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4503953189996537
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.06423791847191751,
          "train_loss": 0.0640510842204094,
          "validation_loss": 0.051516301929950714,
          "train_metric": 0.006450990680605173,
          "validation_metric": 0.00518853310495615,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.467967389000478
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.06397462182212621,
          "train_loss": 0.0638231635093689,
          "validation_loss": 0.051348861306905746,
          "train_metric": 0.006428035907447338,
          "validation_metric": 0.0051716691814363,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.484999697000603
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.06376051565166563,
          "train_loss": 0.06357251852750778,
          "validation_loss": 0.0511525496840477,
          "train_metric": 0.006402791012078524,
          "validation_metric": 0.005151897203177214,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5024536260007153
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.06347989803180099,
          "train_loss": 0.06333249807357788,
          "validation_loss": 0.05096606910228729,
          "train_metric": 0.006378617137670517,
          "validation_metric": 0.005133115686476231,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.520021804000862
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.06351470109075308,
          "train_loss": 0.06315816938877106,
          "validation_loss": 0.0508488304913044,
          "train_metric": 0.006361058913171291,
          "validation_metric": 0.005121308378875256,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5376118149997637
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.06308515975251794,
          "train_loss": 0.06290486454963684,
          "validation_loss": 0.050644148141145706,
          "train_metric": 0.006335548125207424,
          "validation_metric": 0.005100692622363567,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5549095059996034
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.06285714742261916,
          "train_loss": 0.06266985833644867,
          "validation_loss": 0.05047190934419632,
          "train_metric": 0.006311878561973572,
          "validation_metric": 0.005083345342427492,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5731959700005973
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.06267671228852123,
          "train_loss": 0.062464579939842224,
          "validation_loss": 0.05032409355044365,
          "train_metric": 0.006291204132139683,
          "validation_metric": 0.005068458151072264,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.590932631999749
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.062393844476900995,
          "train_loss": 0.06225660815834999,
          "validation_loss": 0.050157636404037476,
          "train_metric": 0.006270257290452719,
          "validation_metric": 0.005051694344729185,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.6087194369993085
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.06224898714572191,
          "train_loss": 0.062048010528087616,
          "validation_loss": 0.04999476298689842,
          "train_metric": 0.0062492480501532555,
          "validation_metric": 0.005035289563238621,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.6267486790002295
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.06205824261996895,
          "train_loss": 0.06185130774974823,
          "validation_loss": 0.049847811460494995,
          "train_metric": 0.006229437422007322,
          "validation_metric": 0.005020489450544119,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.644937499999287
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.06194425164721906,
          "train_loss": 0.06167780980467796,
          "validation_loss": 0.04974302649497986,
          "train_metric": 0.006211963482201099,
          "validation_metric": 0.005009935703128576,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.6621712260002823
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.06160260958131403,
          "train_loss": 0.061457742005586624,
          "validation_loss": 0.049586884677410126,
          "train_metric": 0.006189798936247826,
          "validation_metric": 0.004994209855794907,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.679693918000339
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.06141616380773485,
          "train_loss": 0.06126127392053604,
          "validation_loss": 0.04942944645881653,
          "train_metric": 0.0061700111255049706,
          "validation_metric": 0.0049783531576395035,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.697326000999965
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.061224304838106036,
          "train_loss": 0.06106162071228027,
          "validation_loss": 0.04926992952823639,
          "train_metric": 0.0061499024741351604,
          "validation_metric": 0.0049622864462435246,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7149540030004573
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.06107645551674068,
          "train_loss": 0.060869112610816956,
          "validation_loss": 0.049136362969875336,
          "train_metric": 0.006130514666438103,
          "validation_metric": 0.004948833957314491,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.732518479000646
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.060917692142538726,
          "train_loss": 0.060708481818437576,
          "validation_loss": 0.04904143884778023,
          "train_metric": 0.006114335265010595,
          "validation_metric": 0.004939273931086063,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7498164049993647
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.06064629927277565,
          "train_loss": 0.06051091104745865,
          "validation_loss": 0.04888243228197098,
          "train_metric": 0.006094437092542648,
          "validation_metric": 0.00492325983941555,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7673686769994674
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.06048415997065604,
          "train_loss": 0.0603451281785965,
          "validation_loss": 0.04872376471757889,
          "train_metric": 0.006077740807086229,
          "validation_metric": 0.004907278809696436,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7851908630000253
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.06030433694832027,
          "train_loss": 0.06014231592416763,
          "validation_loss": 0.04859435558319092,
          "train_metric": 0.006057313643395901,
          "validation_metric": 0.004894246347248554,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.8028734380004607
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.06014341115951538,
          "train_loss": 0.059979114681482315,
          "validation_loss": 0.048479437828063965,
          "train_metric": 0.006040876731276512,
          "validation_metric": 0.004882671870291233,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.820989190000546
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.05994406470563263,
          "train_loss": 0.0598013699054718,
          "validation_loss": 0.0483432412147522,
          "train_metric": 0.006022975314408541,
          "validation_metric": 0.004868954885751009,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.8386140700004034
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.059768461738713086,
          "train_loss": 0.05961894989013672,
          "validation_loss": 0.04819135367870331,
          "train_metric": 0.0060046021826565266,
          "validation_metric": 0.004853656515479088,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.8561570570000185
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.05962919280864298,
          "train_loss": 0.05944575369358063,
          "validation_loss": 0.04807816818356514,
          "train_metric": 0.005987158510833979,
          "validation_metric": 0.004842257592827082,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.874173427000642
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.05942036525812,
          "train_loss": 0.059264518320560455,
          "validation_loss": 0.04792984202504158,
          "train_metric": 0.005968904588371515,
          "validation_metric": 0.004827318247407675,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.891715744999601
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.059264499112032354,
          "train_loss": 0.059082772582769394,
          "validation_loss": 0.04779456928372383,
          "train_metric": 0.00595060084015131,
          "validation_metric": 0.004813693929463625,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.910615943999801
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.05906604928895831,
          "train_loss": 0.05890888348221779,
          "validation_loss": 0.0476716086268425,
          "train_metric": 0.005933087319135666,
          "validation_metric": 0.004801309667527676,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.927448098000241
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.05887986556626856,
          "train_loss": 0.058752212673425674,
          "validation_loss": 0.04757228493690491,
          "train_metric": 0.005917307920753956,
          "validation_metric": 0.0047913058660924435,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.9643716209993727
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.058916289592161775,
          "train_loss": 0.05864281952381134,
          "validation_loss": 0.047503724694252014,
          "train_metric": 0.005906290374696255,
          "validation_metric": 0.004784401040524244,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0083428969992383
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.05865616165101528,
          "train_loss": 0.05843430012464523,
          "validation_loss": 0.047267548739910126,
          "train_metric": 0.005885289050638676,
          "validation_metric": 0.00476061413064599,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0261095969999587
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.05847864272072911,
          "train_loss": 0.05823766812682152,
          "validation_loss": 0.047165628522634506,
          "train_metric": 0.005865485407412052,
          "validation_metric": 0.004750349093228579,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.04389273400011
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.05825378780718893,
          "train_loss": 0.058065660297870636,
          "validation_loss": 0.047003086656332016,
          "train_metric": 0.005848160944879055,
          "validation_metric": 0.004733979236334562,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0616076290007186
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.05808089452330023,
          "train_loss": 0.05788706988096237,
          "validation_loss": 0.04687994718551636,
          "train_metric": 0.005830173846334219,
          "validation_metric": 0.004721576347947121,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0789728590007144
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.05785480735357851,
          "train_loss": 0.05772862955927849,
          "validation_loss": 0.04678908362984657,
          "train_metric": 0.00581421609967947,
          "validation_metric": 0.004712425637990236,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0966137510004046
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.05774653400294483,
          "train_loss": 0.057552650570869446,
          "validation_loss": 0.04664923995733261,
          "train_metric": 0.0057964930310845375,
          "validation_metric": 0.004698340781033039,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.114527256000656
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.05754193931352347,
          "train_loss": 0.0573858767747879,
          "validation_loss": 0.04650094732642174,
          "train_metric": 0.005779695697128773,
          "validation_metric": 0.004683405160903931,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1325231320006424
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.057381619699299335,
          "train_loss": 0.05723278969526291,
          "validation_loss": 0.046393558382987976,
          "train_metric": 0.005764276720583439,
          "validation_metric": 0.004672589711844921,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.150634985000579
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.05721505475230515,
          "train_loss": 0.05704518035054207,
          "validation_loss": 0.046241432428359985,
          "train_metric": 0.00574538204818964,
          "validation_metric": 0.004657267592847347,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1684674989992345
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.05702238483354449,
          "train_loss": 0.05688362941145897,
          "validation_loss": 0.04611372575163841,
          "train_metric": 0.005729110911488533,
          "validation_metric": 0.00464440556243062,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1862674999993033
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.05686482565943152,
          "train_loss": 0.056712035089731216,
          "validation_loss": 0.045962609350681305,
          "train_metric": 0.005711828358471394,
          "validation_metric": 0.004629185423254967,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.204089904000284
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.05675346544012427,
          "train_loss": 0.056547749787569046,
          "validation_loss": 0.045862823724746704,
          "train_metric": 0.005695282947272062,
          "validation_metric": 0.004619135055691004,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.2216921519993775
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.056545853381976485,
          "train_loss": 0.05636799335479736,
          "validation_loss": 0.04570659622550011,
          "train_metric": 0.00567717757076025,
          "validation_metric": 0.00460340129211545,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.2388613809998787
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.05635213025379926,
          "train_loss": 0.05619708448648453,
          "validation_loss": 0.0455770418047905,
          "train_metric": 0.005659965332597494,
          "validation_metric": 0.0045903525315225124,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.259966005000024
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.05625436222180724,
          "train_loss": 0.056061916053295135,
          "validation_loss": 0.045505836606025696,
          "train_metric": 0.0056463503278791904,
          "validation_metric": 0.004583180882036686,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.305943704999663
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.05630800442304462,
          "train_loss": 0.05589764565229416,
          "validation_loss": 0.0453062579035759,
          "train_metric": 0.005629806779325008,
          "validation_metric": 0.004563080612570047,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.3342902660006075
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.055814682389609516,
          "train_loss": 0.05569475144147873,
          "validation_loss": 0.045182645320892334,
          "train_metric": 0.005609370302408934,
          "validation_metric": 0.004550631158053875,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.351580168999135
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.05572481395211071,
          "train_loss": 0.05554748699069023,
          "validation_loss": 0.04509114846587181,
          "train_metric": 0.005594539921730757,
          "validation_metric": 0.0045414152555167675,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.3693105000002106
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.055498196044936776,
          "train_loss": 0.05536197870969772,
          "validation_loss": 0.04491767659783363,
          "train_metric": 0.005575855262577534,
          "validation_metric": 0.004523943178355694,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.3917254970001522
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.05542772833723575,
          "train_loss": 0.055191539227962494,
          "validation_loss": 0.04478727653622627,
          "train_metric": 0.005558689124882221,
          "validation_metric": 0.004510810133069754,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.450678126000639
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.05515679030213505,
          "train_loss": 0.055026181042194366,
          "validation_loss": 0.044647540897130966,
          "train_metric": 0.0055420356802642345,
          "validation_metric": 0.004496736451983452,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4765563729997666
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.055021790438331664,
          "train_loss": 0.054854899644851685,
          "validation_loss": 0.04451264813542366,
          "train_metric": 0.005524784792214632,
          "validation_metric": 0.004483150318264961,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5079536409994034
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.054846093989908695,
          "train_loss": 0.05468733608722687,
          "validation_loss": 0.04439740628004074,
          "train_metric": 0.005507908295840025,
          "validation_metric": 0.004471544176340103,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5342819149991556
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.054659615037962794,
          "train_loss": 0.05451902002096176,
          "validation_loss": 0.0442500114440918,
          "train_metric": 0.005490956362336874,
          "validation_metric": 0.00445669936016202,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5560215229997993
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.05452826071996242,
          "train_loss": 0.05435533449053764,
          "validation_loss": 0.04411807283759117,
          "train_metric": 0.005474471021443605,
          "validation_metric": 0.004443409852683544,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5772386730004655
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.054288562503643334,
          "train_loss": 0.054176609963178635,
          "validation_loss": 0.04398948326706886,
          "train_metric": 0.0054564690217375755,
          "validation_metric": 0.004430459346622229,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.589882669999497
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.05428814922925085,
          "train_loss": 0.05401518940925598,
          "validation_loss": 0.04383929446339607,
          "train_metric": 0.005440211854875088,
          "validation_metric": 0.004415333736687899,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.602374444999441
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.05406310362741351,
          "train_loss": 0.05386461317539215,
          "validation_loss": 0.04376307874917984,
          "train_metric": 0.005425046663731337,
          "validation_metric": 0.0044076573103666306,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.614938696999161
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0538207096979022,
          "train_loss": 0.05367431789636612,
          "validation_loss": 0.043581314384937286,
          "train_metric": 0.005405880510807037,
          "validation_metric": 0.004389350302517414,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.62718443400081
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.053814564482308924,
          "train_loss": 0.053525667637586594,
          "validation_loss": 0.04348144680261612,
          "train_metric": 0.0053909095004200935,
          "validation_metric": 0.00437929155305028,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.6398247390006873
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.05343074654228985,
          "train_loss": 0.05334046110510826,
          "validation_loss": 0.04330768436193466,
          "train_metric": 0.00537225604057312,
          "validation_metric": 0.0043617915362119675,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.652363781000531
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.053419652045704424,
          "train_loss": 0.053184229880571365,
          "validation_loss": 0.04316252842545509,
          "train_metric": 0.005356520880013704,
          "validation_metric": 0.004347171168774366,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.6649620169991977
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.05317984824068844,
          "train_loss": 0.053001150488853455,
          "validation_loss": 0.043046556413173676,
          "train_metric": 0.005338081158697605,
          "validation_metric": 0.004335491918027401,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.6771887890008657
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.053085311548784375,
          "train_loss": 0.052854135632514954,
          "validation_loss": 0.04295792803168297,
          "train_metric": 0.0053232754580676556,
          "validation_metric": 0.004326565191149712,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.689760277999085
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.052809933898970485,
          "train_loss": 0.05266397073864937,
          "validation_loss": 0.042774394154548645,
          "train_metric": 0.005304122809320688,
          "validation_metric": 0.004308079835027456,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7021559620006883
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.052698828279972076,
          "train_loss": 0.0525045208632946,
          "validation_loss": 0.042612116783857346,
          "train_metric": 0.005288063548505306,
          "validation_metric": 0.004291736055165529,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7147265240000706
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.052481479360722005,
          "train_loss": 0.05231931060552597,
          "validation_loss": 0.04247700423002243,
          "train_metric": 0.0052694096229970455,
          "validation_metric": 0.0042781285010278225,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.726789103999181
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.052472277311608195,
          "train_loss": 0.05217496305704117,
          "validation_loss": 0.04241590201854706,
          "train_metric": 0.005254871211946011,
          "validation_metric": 0.004271973855793476,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7391703499997675
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.052119326195679605,
          "train_loss": 0.05197242647409439,
          "validation_loss": 0.04219749942421913,
          "train_metric": 0.005234472919255495,
          "validation_metric": 0.004249977879226208,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7518646089993126
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.051966557395644486,
          "train_loss": 0.051799971610307693,
          "validation_loss": 0.04206150770187378,
          "train_metric": 0.005217103287577629,
          "validation_metric": 0.0042362818494439125,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.764317633000246
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.05177964235190302,
          "train_loss": 0.05162586644291878,
          "validation_loss": 0.04194172844290733,
          "train_metric": 0.005199568811804056,
          "validation_metric": 0.004224217031151056,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7766433300002973
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.051713238237425685,
          "train_loss": 0.05144738033413887,
          "validation_loss": 0.0417877659201622,
          "train_metric": 0.005181591492146254,
          "validation_metric": 0.004208710510283709,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7890282459993614
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.05144214618485421,
          "train_loss": 0.05127592384815216,
          "validation_loss": 0.04165603592991829,
          "train_metric": 0.005164323374629021,
          "validation_metric": 0.004195443354547024,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8013674569992872
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.051249282667413354,
          "train_loss": 0.05110051482915878,
          "validation_loss": 0.04149608686566353,
          "train_metric": 0.00514665711671114,
          "validation_metric": 0.0041793342679739,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8135881449998124
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.051059840712696314,
          "train_loss": 0.050924014300107956,
          "validation_loss": 0.04135344550013542,
          "train_metric": 0.005128880497068167,
          "validation_metric": 0.004164967220276594,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8259034490001795
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.05094924953300506,
          "train_loss": 0.05075191333889961,
          "validation_loss": 0.04121342673897743,
          "train_metric": 0.0051115467213094234,
          "validation_metric": 0.004150865133851767,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8377467690006597
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.05072047468274832,
          "train_loss": 0.050569236278533936,
          "validation_loss": 0.041069772094488144,
          "train_metric": 0.005093148443847895,
          "validation_metric": 0.004136397037655115,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8498674780003057
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.05053954536560923,
          "train_loss": 0.050393346697092056,
          "validation_loss": 0.04092844948172569,
          "train_metric": 0.005075433291494846,
          "validation_metric": 0.004122163634747267,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.862098296000113
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.05035687796771526,
          "train_loss": 0.05023371800780296,
          "validation_loss": 0.040834397077560425,
          "train_metric": 0.005059355869889259,
          "validation_metric": 0.00411269161850214,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.874360544999945
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.050246915430761874,
          "train_loss": 0.05005204305052757,
          "validation_loss": 0.04066097363829613,
          "train_metric": 0.005041058640927076,
          "validation_metric": 0.004095224663615227,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.886585070000365
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.050007922924123704,
          "train_loss": 0.04987179487943649,
          "validation_loss": 0.04050586372613907,
          "train_metric": 0.005022904835641384,
          "validation_metric": 0.004079602193087339,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.898782833999576
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.049836845020763576,
          "train_loss": 0.04969437047839165,
          "validation_loss": 0.04034560173749924,
          "train_metric": 0.00500503508374095,
          "validation_metric": 0.004063460975885391,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.910926852000557
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.04965066828299314,
          "train_loss": 0.04952279105782509,
          "validation_loss": 0.04020213335752487,
          "train_metric": 0.004987753462046385,
          "validation_metric": 0.004049011040478945,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.923259657000017
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.04956131731159985,
          "train_loss": 0.04934108257293701,
          "validation_loss": 0.04005476459860802,
          "train_metric": 0.004969453439116478,
          "validation_metric": 0.004034169018268585,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9355211010006315
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.04930702783167362,
          "train_loss": 0.049158379435539246,
          "validation_loss": 0.03991033136844635,
          "train_metric": 0.004951052367687225,
          "validation_metric": 0.004019621759653091,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.948345839000467
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.04921581910457462,
          "train_loss": 0.04899795725941658,
          "validation_loss": 0.039769284427165985,
          "train_metric": 0.004934894852340221,
          "validation_metric": 0.0040054162964224815,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9607622969997465
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.048988139955326915,
          "train_loss": 0.04882999509572983,
          "validation_loss": 0.03968121483922005,
          "train_metric": 0.004917978309094906,
          "validation_metric": 0.003996546845883131,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.973352179000358
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.04886782635003328,
          "train_loss": 0.04862122982740402,
          "validation_loss": 0.03947623446583748,
          "train_metric": 0.004896952770650387,
          "validation_metric": 0.003975901752710342,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.985761989000821
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.048793428228236735,
          "train_loss": 0.04845011979341507,
          "validation_loss": 0.03933831676840782,
          "train_metric": 0.00487971818074584,
          "validation_metric": 0.00396201154217124,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.998494868999842
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.04839885514229536,
          "train_loss": 0.04827623814344406,
          "validation_loss": 0.03917921707034111,
          "train_metric": 0.004862206522375345,
          "validation_metric": 0.003945987671613693,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.011196299999938
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.04835412523243576,
          "train_loss": 0.0480988584458828,
          "validation_loss": 0.03903878480195999,
          "train_metric": 0.004844340495765209,
          "validation_metric": 0.003931843210011721,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.024081801000648
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.04812730778940022,
          "train_loss": 0.04793790727853775,
          "validation_loss": 0.03893771395087242,
          "train_metric": 0.004828130826354027,
          "validation_metric": 0.0039216638542711735,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.036506301999907
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.04792195535264909,
          "train_loss": 0.04774564877152443,
          "validation_loss": 0.038738369941711426,
          "train_metric": 0.004808766767382622,
          "validation_metric": 0.0039015866350382566,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.0489115939999465
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.047710827086120844,
          "train_loss": 0.04756542667746544,
          "validation_loss": 0.03859221562743187,
          "train_metric": 0.004790615756064653,
          "validation_metric": 0.0038868666160851717,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.0614071220006736
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.04757405037526041,
          "train_loss": 0.04738318547606468,
          "validation_loss": 0.03844774141907692,
          "train_metric": 0.004772260785102844,
          "validation_metric": 0.003872315399348736,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.073932735000199
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.047333957511000335,
          "train_loss": 0.04720160737633705,
          "validation_loss": 0.03830154612660408,
          "train_metric": 0.004753973335027695,
          "validation_metric": 0.003857591887935996,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.086580176999632
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.047192181926220655,
          "train_loss": 0.047023922204971313,
          "validation_loss": 0.038174524903297424,
          "train_metric": 0.004736077506095171,
          "validation_metric": 0.0038447983097285032,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.184783095999592
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.047022208804264665,
          "train_loss": 0.04684695973992348,
          "validation_loss": 0.03800595924258232,
          "train_metric": 0.004718254320323467,
          "validation_metric": 0.003827820997685194,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.19762772499962
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.04680420539807528,
          "train_loss": 0.046663619577884674,
          "validation_loss": 0.03785888850688934,
          "train_metric": 0.0046997894532978535,
          "validation_metric": 0.003813008312135935,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.209744186000535
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0466418422292918,
          "train_loss": 0.04649099335074425,
          "validation_loss": 0.03771970793604851,
          "train_metric": 0.004682402592152357,
          "validation_metric": 0.00379899051040411,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.221762851000676
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.04647693771403283,
          "train_loss": 0.04630625993013382,
          "validation_loss": 0.03755088523030281,
          "train_metric": 0.004663797095417976,
          "validation_metric": 0.0037819878198206425,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.233487007999429
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.046407196088694036,
          "train_loss": 0.0461648628115654,
          "validation_loss": 0.03741995245218277,
          "train_metric": 0.0046495553106069565,
          "validation_metric": 0.0037688002921640873,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.245542541000759
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.046212150366045535,
          "train_loss": 0.04599359631538391,
          "validation_loss": 0.037325721234083176,
          "train_metric": 0.004632307216525078,
          "validation_metric": 0.0037593096494674683,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.257270648000485
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.04615531221497804,
          "train_loss": 0.0457879900932312,
          "validation_loss": 0.03711565211415291,
          "train_metric": 0.004611598793417215,
          "validation_metric": 0.0037381527945399284,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.2691946150007425
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.04579040827229619,
          "train_loss": 0.04561654105782509,
          "validation_loss": 0.03701232746243477,
          "train_metric": 0.00459433114156127,
          "validation_metric": 0.003727746196091175,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.281240391999745
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.04553134483285248,
          "train_loss": 0.045402321964502335,
          "validation_loss": 0.0368051640689373,
          "train_metric": 0.00457275565713644,
          "validation_metric": 0.003706881310790777,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.293562308999753
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.04538600076921284,
          "train_loss": 0.04522489756345749,
          "validation_loss": 0.036651115864515305,
          "train_metric": 0.004554886370897293,
          "validation_metric": 0.003691366408020258,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.3061336430000665
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.04534842341672629,
          "train_loss": 0.04505122825503349,
          "validation_loss": 0.036512020975351334,
          "train_metric": 0.0045373947359621525,
          "validation_metric": 0.0036773569881916046,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.31856963700011
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.045054107904434204,
          "train_loss": 0.04487717151641846,
          "validation_loss": 0.03637956082820892,
          "train_metric": 0.004519864451140165,
          "validation_metric": 0.0036640155594795942,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.330865497000559
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.04489806713536382,
          "train_loss": 0.04470595717430115,
          "validation_loss": 0.03620670363306999,
          "train_metric": 0.004502620082348585,
          "validation_metric": 0.0036466065794229507,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.343280606999542
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.04484301304910332,
          "train_loss": 0.04457611218094826,
          "validation_loss": 0.036151085048913956,
          "train_metric": 0.004489542450755835,
          "validation_metric": 0.0036410046741366386,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.355615447000673
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.04458111815620214,
          "train_loss": 0.044339682906866074,
          "validation_loss": 0.03590480238199234,
          "train_metric": 0.004465730395168066,
          "validation_metric": 0.003616200527176261,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.367665568999655
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.04439289728179574,
          "train_loss": 0.04415823146700859,
          "validation_loss": 0.0357668474316597,
          "train_metric": 0.004447455517947674,
          "validation_metric": 0.003602305892854929,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.379960350999681
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.04423615534324199,
          "train_loss": 0.044001054018735886,
          "validation_loss": 0.035649001598358154,
          "train_metric": 0.0044316258281469345,
          "validation_metric": 0.003590436652302742,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.392300506000538
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0439056612085551,
          "train_loss": 0.04379836469888687,
          "validation_loss": 0.0354619137942791,
          "train_metric": 0.004411210771650076,
          "validation_metric": 0.0035715943668037653,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.404550212000686
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.04378662328235805,
          "train_loss": 0.04362908750772476,
          "validation_loss": 0.035315267741680145,
          "train_metric": 0.004394161514937878,
          "validation_metric": 0.003556824754923582,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.416861777999657
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.04366158548509702,
          "train_loss": 0.04344441369175911,
          "validation_loss": 0.035161424428224564,
          "train_metric": 0.004375562537461519,
          "validation_metric": 0.003541329875588417,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.429274099000395
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.04342546430416405,
          "train_loss": 0.04327751323580742,
          "validation_loss": 0.03503123298287392,
          "train_metric": 0.00435875216498971,
          "validation_metric": 0.0035282177850604057,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.441353832999084
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0432821981376037,
          "train_loss": 0.043082598596811295,
          "validation_loss": 0.03488022834062576,
          "train_metric": 0.004339121747761965,
          "validation_metric": 0.0035130088217556477,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.453687797000384
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.04317728616297245,
          "train_loss": 0.04293176531791687,
          "validation_loss": 0.03472498804330826,
          "train_metric": 0.004323930013924837,
          "validation_metric": 0.0034973740112036467,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.465972949999923
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.042883855872787535,
          "train_loss": 0.042748115956783295,
          "validation_loss": 0.03458872064948082,
          "train_metric": 0.004305433016270399,
          "validation_metric": 0.003483649343252182,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.47813467800006
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.043159221415407956,
          "train_loss": 0.0426119901239872,
          "validation_loss": 0.03445277363061905,
          "train_metric": 0.004291723947972059,
          "validation_metric": 0.003469957271590829,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.490655843999775
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.04297937254887074,
          "train_loss": 0.042543694376945496,
          "validation_loss": 0.03441200405359268,
          "train_metric": 0.004284844733774662,
          "validation_metric": 0.0034658508375287056,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.5029337180003495
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.042465966078452766,
          "train_loss": 0.042209576815366745,
          "validation_loss": 0.0341569185256958,
          "train_metric": 0.004251193720847368,
          "validation_metric": 0.003440160071477294,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.515484626999751
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.04217631393112242,
          "train_loss": 0.04203122481703758,
          "validation_loss": 0.03399286046624184,
          "train_metric": 0.0042332313023507595,
          "validation_metric": 0.003423636546358466,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.5275554950003425
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.04202009714208543,
          "train_loss": 0.04187146946787834,
          "validation_loss": 0.03386116400361061,
          "train_metric": 0.004217141307890415,
          "validation_metric": 0.003410372417420149,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.540487006999683
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.04185493034310639,
          "train_loss": 0.041700512170791626,
          "validation_loss": 0.03370814770460129,
          "train_metric": 0.00419992208480835,
          "validation_metric": 0.0033949613571166992,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.55244005699933
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.04167568206321448,
          "train_loss": 0.04151003435254097,
          "validation_loss": 0.033553216606378555,
          "train_metric": 0.00418073870241642,
          "validation_metric": 0.003379357047379017,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.564707449000707
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.04151977290166542,
          "train_loss": 0.041348762810230255,
          "validation_loss": 0.03341120108962059,
          "train_metric": 0.004164496436715126,
          "validation_metric": 0.0033650542609393597,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.576923679000174
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.041482306784018874,
          "train_loss": 0.041166264563798904,
          "validation_loss": 0.03325209766626358,
          "train_metric": 0.004146115854382515,
          "validation_metric": 0.0033490294590592384,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.589623936999487
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.04118693678174168,
          "train_loss": 0.041018251329660416,
          "validation_loss": 0.03312823548913002,
          "train_metric": 0.004131208173930645,
          "validation_metric": 0.003336555091664195,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.601972403999753
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.0410125651396811,
          "train_loss": 0.04081521928310394,
          "validation_loss": 0.03297288715839386,
          "train_metric": 0.0041107595898211,
          "validation_metric": 0.003320908872410655,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.614779223000369
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.04077491711359471,
          "train_loss": 0.040637340396642685,
          "validation_loss": 0.03281473368406296,
          "train_metric": 0.004092843737453222,
          "validation_metric": 0.003304979531094432,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.763501142000678
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 3840
      },
      "best_validation_metric": 0.003304979531094432,
      "config": {
        "label": "LR 0.0001",
        "batch": 64,
        "lr": 0.0001,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.05748895138280426
    },
    "1298f0867b9f_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7750630155205727,
          "train_loss": 0.3853791058063507,
          "validation_loss": 0.30187591910362244,
          "train_metric": 0.03881397098302841,
          "validation_metric": 0.03040383756160736,
          "lr_first": 0.003,
          "lr_last": 0.002999887429249521,
          "seconds": 0.01802909899924998
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.2060780874453485,
          "train_loss": 0.1268892139196396,
          "validation_loss": 0.10953299701213837,
          "train_metric": 0.01277981884777546,
          "validation_metric": 0.011031762696802616,
          "lr_first": 0.0029998719197231404,
          "lr_last": 0.0029995192175544397,
          "seconds": 0.036053073999937624
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12402724288403988,
          "train_loss": 0.09904995560646057,
          "validation_loss": 0.08297355473041534,
          "train_metric": 0.009975949302315712,
          "validation_metric": 0.008356792852282524,
          "lr_first": 0.0029994877008384588,
          "lr_last": 0.002998894927685237,
          "seconds": 0.054469717000756646
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09277393901720643,
          "train_loss": 0.08497162163257599,
          "validation_loss": 0.06828556954860687,
          "train_metric": 0.008558030240237713,
          "validation_metric": 0.006877472624182701,
          "lr_first": 0.002998847409179882,
          "lr_last": 0.002998014666610766,
          "seconds": 0.07274423200033198
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07505651679821312,
          "train_loss": 0.06761742383241653,
          "validation_loss": 0.05565846338868141,
          "train_metric": 0.00681017991155386,
          "validation_metric": 0.00560571625828743,
          "lr_first": 0.0029979511544580887,
          "lr_last": 0.0029968785851592315,
          "seconds": 0.09055452500069805
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06592379149515182,
          "train_loss": 0.06523515284061432,
          "validation_loss": 0.05283120647072792,
          "train_metric": 0.0065702456049621105,
          "validation_metric": 0.005320965312421322,
          "lr_first": 0.002996799090241712,
          "lr_last": 0.002995486877992343,
          "seconds": 0.10872963200017693
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06306629558093846,
          "train_loss": 0.0651402622461319,
          "validation_loss": 0.05299976468086243,
          "train_metric": 0.006560687907040119,
          "validation_metric": 0.0053379419259727,
          "lr_first": 0.0029953914139310266,
          "lr_last": 0.0029938397835719568,
          "seconds": 0.12584397299906414
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06272630509920418,
          "train_loss": 0.05920081213116646,
          "validation_loss": 0.04778161644935608,
          "train_metric": 0.0059624891728162766,
          "validation_metric": 0.004812389612197876,
          "lr_first": 0.002993728366724119,
          "lr_last": 0.002991937584119225,
          "seconds": 0.14385212600063824
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05564913444686681,
          "train_loss": 0.05258726328611374,
          "validation_loss": 0.04344264790415764,
          "train_metric": 0.005296397488564253,
          "validation_metric": 0.004375385120511055,
          "lr_first": 0.0029918102335755687,
          "lr_last": 0.0029897806055662304,
          "seconds": 0.16172502799963695
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05134535860270262,
          "train_loss": 0.05154154449701309,
          "validation_loss": 0.04278421029448509,
          "train_metric": 0.005191075149923563,
          "validation_metric": 0.004309068899601698,
          "lr_first": 0.0029896373431476165,
          "lr_last": 0.0029873692175001417,
          "seconds": 0.1801565309997386
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04969988949596882,
          "train_loss": 0.048038214445114136,
          "validation_loss": 0.03884723037481308,
          "train_metric": 0.004838232882320881,
          "validation_metric": 0.003912550397217274,
          "lr_first": 0.002987210067753847,
          "lr_last": 0.0029847038330998916,
          "seconds": 0.1986162470002455
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04531278368085623,
          "train_loss": 0.041745539754629135,
          "validation_loss": 0.034716952592134476,
          "train_metric": 0.004204457625746727,
          "validation_metric": 0.003496564691886306,
          "lr_first": 0.0029845288232954064,
          "lr_last": 0.0029817849090653776,
          "seconds": 0.21645641699979024
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.040732288267463446,
          "train_loss": 0.03914289548993111,
          "validation_loss": 0.032656002789735794,
          "train_metric": 0.003942329436540604,
          "validation_metric": 0.0032889926806092262,
          "lr_first": 0.0029815940691897305,
          "lr_last": 0.002978612945539204,
          "seconds": 0.23474143400017056
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.037410909309983253,
          "train_loss": 0.03535976633429527,
          "validation_loss": 0.029605358839035034,
          "train_metric": 0.003561306046321988,
          "validation_metric": 0.0029817435424774885,
          "lr_first": 0.002978406308291826,
          "lr_last": 0.0029751884860209923,
          "seconds": 0.2529194929993537
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03616417699959129,
          "train_loss": 0.039494212716817856,
          "validation_loss": 0.033457815647125244,
          "train_metric": 0.003977712709456682,
          "validation_metric": 0.0033697490580379963,
          "lr_first": 0.0029749660868081114,
          "lr_last": 0.0029715121172742523,
          "seconds": 0.2699869709995255
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.03397140366723761,
          "train_loss": 0.03585522621870041,
          "validation_loss": 0.0303086768835783,
          "train_metric": 0.0036112070083618164,
          "validation_metric": 0.003052579006180167,
          "lr_first": 0.0029712739942028285,
          "lr_last": 0.0029675844692258427,
          "seconds": 0.289833892000388
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.030409214639803395,
          "train_loss": 0.027377478778362274,
          "validation_loss": 0.023430563509464264,
          "train_metric": 0.0027573592960834503,
          "validation_metric": 0.002359840553253889,
          "lr_first": 0.002967330663097039,
          "lr_last": 0.002963406214858036,
          "seconds": 0.33348831399962364
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.03021729231113568,
          "train_loss": 0.036261532455682755,
          "validation_loss": 0.02839723415672779,
          "train_metric": 0.0036521286237984896,
          "validation_metric": 0.0028600660152733326,
          "lr_first": 0.0029631367691602254,
          "lr_last": 0.00295897807009321,
          "seconds": 0.350426184000753
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.03147682186681777,
          "train_loss": 0.026124197989702225,
          "validation_loss": 0.01964973285794258,
          "train_metric": 0.0026311336550861597,
          "validation_metric": 0.001979049062356353,
          "lr_first": 0.002958693030994526,
          "lr_last": 0.002954300793671172,
          "seconds": 0.367896132000169
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02388609980698675,
          "train_loss": 0.022394994273781776,
          "validation_loss": 0.018845243379473686,
          "train_metric": 0.0022555417381227016,
          "validation_metric": 0.0018980237655341625,
          "lr_first": 0.002954000210011597,
          "lr_last": 0.00294937518701916,
          "seconds": 0.38584404299945163
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023435425013303757,
          "train_loss": 0.020204074680805206,
          "validation_loss": 0.01724330335855484,
          "train_metric": 0.002034880220890045,
          "validation_metric": 0.0017366823740303516,
          "lr_first": 0.002949059110302156,
          "lr_last": 0.0029442020941145156,
          "seconds": 0.4034677429990552
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02317172527546063,
          "train_loss": 0.021212272346019745,
          "validation_loss": 0.016477253288030624,
          "train_metric": 0.002136422321200371,
          "validation_metric": 0.001659528468735516,
          "lr_first": 0.0029438705784982015,
          "lr_last": 0.002938782401340078,
          "seconds": 0.42151186000046437
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.022194949007825926,
          "train_loss": 0.01986221969127655,
          "validation_loss": 0.017614779993891716,
          "train_metric": 0.0020004501566290855,
          "validation_metric": 0.0017740960465744138,
          "lr_first": 0.0029384355036279485,
          "lr_last": 0.0029331170373323025,
          "seconds": 0.43856198499997845
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.01853561168536544,
          "train_loss": 0.017261959612369537,
          "validation_loss": 0.014878902584314346,
          "train_metric": 0.0017385614337399602,
          "validation_metric": 0.0014985486632212996,
          "lr_first": 0.002932754816963494,
          "lr_last": 0.002927206972822146,
          "seconds": 0.45627757699912763
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019558302767109126,
          "train_loss": 0.02019488997757435,
          "validation_loss": 0.0151126179844141,
          "train_metric": 0.0020339551847428083,
          "validation_metric": 0.0015220876084640622,
          "lr_first": 0.002926829491861254,
          "lr_last": 0.002921053220468738,
          "seconds": 0.472982356999637
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.018771823204588145,
          "train_loss": 0.020884867757558823,
          "validation_loss": 0.019569210708141327,
          "train_metric": 0.002103447448462248,
          "validation_metric": 0.0019709391053766012,
          "lr_first": 0.0029206605435951826,
          "lr_last": 0.0029146568346858654,
          "seconds": 0.49696215000039956
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.017258061561733484,
          "train_loss": 0.015214218758046627,
          "validation_loss": 0.01310745533555746,
          "train_metric": 0.0015323202824220061,
          "validation_metric": 0.0013201349647715688,
          "lr_first": 0.002914249029182806,
          "lr_last": 0.002908018911461302,
          "seconds": 0.5143835599992599
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015380903234472498,
          "train_loss": 0.014855005778372288,
          "validation_loss": 0.012994576245546341,
          "train_metric": 0.0014961418928578496,
          "validation_metric": 0.0013087659608572721,
          "lr_first": 0.002907596047204113,
          "lr_last": 0.0029011405881690204,
          "seconds": 0.532247147999442
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.017229885619599372,
          "train_loss": 0.01835150457918644,
          "validation_loss": 0.014845347963273525,
          "train_metric": 0.0018482963787391782,
          "validation_metric": 0.0014951692428439856,
          "lr_first": 0.002900702737613315,
          "lr_last": 0.0028940230433743067,
          "seconds": 0.5490580919995409
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.017065402818843722,
          "train_loss": 0.014565709047019482,
          "validation_loss": 0.01229576114565134,
          "train_metric": 0.0014670047676190734,
          "validation_metric": 0.0012383840512484312,
          "lr_first": 0.002893570281543524,
          "lr_last": 0.002886667496631822,
          "seconds": 0.5667393139992782
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.02183365830569528,
          "train_loss": 0.013556724414229393,
          "validation_loss": 0.011162146925926208,
          "train_metric": 0.0013653835048899055,
          "validation_metric": 0.0011242106556892395,
          "lr_first": 0.0028861999011043707,
          "lr_last": 0.002879075208276634,
          "seconds": 0.5845709309996892
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.018014416506048292,
          "train_loss": 0.015458480454981327,
          "validation_loss": 0.011149626225233078,
          "train_metric": 0.0015569216338917613,
          "validation_metric": 0.0011229496449232101,
          "lr_first": 0.0028785928591726004,
          "lr_last": 0.0028712474792082655,
          "seconds": 0.6024804249991575
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.014809048967435956,
          "train_loss": 0.01211323868483305,
          "validation_loss": 0.009718004614114761,
          "train_metric": 0.0012200011406093836,
          "validation_metric": 0.0009787617018446326,
          "lr_first": 0.0028707504591756854,
          "lr_last": 0.0028631856506677973,
          "seconds": 0.7483562340003118
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01317998644663021,
          "train_loss": 0.015515336766839027,
          "validation_loss": 0.010780859738588333,
          "train_metric": 0.0015626479871571064,
          "validation_metric": 0.0010858086170628667,
          "lr_first": 0.0028626740448684956,
          "lr_last": 0.0028548911040080457,
          "seconds": 0.7625123790003272
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.01437787211034447,
          "train_loss": 0.015200702473521233,
          "validation_loss": 0.01364054623991251,
          "train_metric": 0.0015309591544792056,
          "validation_metric": 0.0013738260604441166,
          "lr_first": 0.002854365000103041,
          "lr_last": 0.002846365260456879,
          "seconds": 0.7764711969994096
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.01640879752812907,
          "train_loss": 0.021001795306801796,
          "validation_loss": 0.015360614284873009,
          "train_metric": 0.002115223789587617,
          "validation_metric": 0.0015470648650079966,
          "lr_first": 0.0028458247485913675,
          "lr_last": 0.002837609580873701,
          "seconds": 0.7904072000001179
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.012311289872741327,
          "train_loss": 0.011123312637209892,
          "validation_loss": 0.009485164657235146,
          "train_metric": 0.0011202992172911763,
          "validation_metric": 0.0009553111158311367,
          "lr_first": 0.0028370547536616087,
          "lr_last": 0.002828625565499134,
          "seconds": 0.8047305969994341
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011617553944233805,
          "train_loss": 0.009937791153788567,
          "validation_loss": 0.008104953914880753,
          "train_metric": 0.0010008979588747025,
          "validation_metric": 0.0008163012680597603,
          "lr_first": 0.002828056518007247,
          "lr_last": 0.00281941475369796,
          "seconds": 0.8192905870000686
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.010852123727090657,
          "train_loss": 0.009928123094141483,
          "validation_loss": 0.00780867226421833,
          "train_metric": 0.0009999241447076201,
          "validation_metric": 0.0007864609360694885,
          "lr_first": 0.0028188315834296366,
          "lr_last": 0.0028099787236953665,
          "seconds": 0.834335468000063
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012826824604417197,
          "train_loss": 0.013163914903998375,
          "validation_loss": 0.0094850342720747,
          "train_metric": 0.001325821503996849,
          "validation_metric": 0.0009552979026921093,
          "lr_first": 0.00280938153057383,
          "lr_last": 0.0028003190923065156,
          "seconds": 0.8489121149996208
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010832579660927877,
          "train_loss": 0.010110517032444477,
          "validation_loss": 0.007634029723703861,
          "train_metric": 0.0010182942496612668,
          "validation_metric": 0.0007688715122640133,
          "lr_first": 0.0027997079786577336,
          "lr_last": 0.00279043751465952,
          "seconds": 0.8641419640007371
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010855459579033777,
          "train_loss": 0.009975188411772251,
          "validation_loss": 0.007316493894904852,
          "train_metric": 0.001004664460197091,
          "validation_metric": 0.0007368904771283269,
          "lr_first": 0.00278981258519467,
          "lr_last": 0.0027803356839118383,
          "seconds": 0.8793279740002617
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011440545640653,
          "train_loss": 0.008938828483223915,
          "validation_loss": 0.007149587385356426,
          "train_metric": 0.000900286016985774,
          "validation_metric": 0.0007200803374871612,
          "lr_first": 0.0027796970457093662,
          "lr_last": 0.002770015330960163,
          "seconds": 0.8941955880000023
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.009919340285705402,
          "train_loss": 0.009017202071845531,
          "validation_loss": 0.007138718385249376,
          "train_metric": 0.0009081795578822494,
          "validation_metric": 0.0007189855095930398,
          "lr_first": 0.0027693630934474377,
          "lr_last": 0.0027594782241438446,
          "seconds": 0.9094071920007991
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013510818214854226,
          "train_loss": 0.009734764695167542,
          "validation_loss": 0.008079209364950657,
          "train_metric": 0.0009804497240111232,
          "validation_metric": 0.0008137084660120308,
          "lr_first": 0.0027588124990784056,
          "lr_last": 0.0027487261689418912,
          "seconds": 0.9223967990001256
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.012186518855742179,
          "train_loss": 0.01038524229079485,
          "validation_loss": 0.0076666297391057014,
          "train_metric": 0.0010459637269377708,
          "validation_metric": 0.0007721548317931592,
          "lr_first": 0.0027480470703923022,
          "lr_last": 0.002737761007663611,
          "seconds": 0.9349212369997986
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011637854971922934,
          "train_loss": 0.01331253256648779,
          "validation_loss": 0.012220054864883423,
          "train_metric": 0.0013407894875854254,
          "validation_metric": 0.0012307590804994106,
          "lr_first": 0.0027370686519899118,
          "lr_last": 0.0027265846191329374,
          "seconds": 0.9471121650003624
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010312451006029733,
          "train_loss": 0.014557366259396076,
          "validation_loss": 0.01398459728807211,
          "train_metric": 0.0014661647146567702,
          "validation_metric": 0.0014084774302318692,
          "lr_first": 0.0027258791249667085,
          "lr_last": 0.0027151989183665085,
          "seconds": 0.9590835519993561
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.012566177872940898,
          "train_loss": 0.010833770968019962,
          "validation_loss": 0.008831595070660114,
          "train_metric": 0.0010911377612501383,
          "validation_metric": 0.0008894859347492456,
          "lr_first": 0.002714480406590544,
          "lr_last": 0.0027036058562455397,
          "seconds": 0.9712057619999541
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.011326618216116913,
          "train_loss": 0.008397739380598068,
          "validation_loss": 0.006359912920743227,
          "train_metric": 0.0008457896765321493,
          "validation_metric": 0.0006405471358448267,
          "lr_first": 0.0027028744499731326,
          "lr_last": 0.002691807419181544,
          "seconds": 0.9835364390000905
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009222109016263857,
          "train_loss": 0.013731792569160461,
          "validation_loss": 0.012639855034649372,
          "train_metric": 0.0013830158859491348,
          "validation_metric": 0.0012730398448184133,
          "lr_first": 0.002691063243735394,
          "lr_last": 0.0026798056287759726,
          "seconds": 0.9960750640002516
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.00912014368805103,
          "train_loss": 0.008336761966347694,
          "validation_loss": 0.00598534382879734,
          "train_metric": 0.0008396482444368303,
          "validation_metric": 0.0006028218776918948,
          "lr_first": 0.0026790488116667116,
          "lr_last": 0.002667602541473826,
          "seconds": 1.0087111070006358
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008446889885817654,
          "train_loss": 0.010209133848547935,
          "validation_loss": 0.007553229108452797,
          "train_metric": 0.001028226688504219,
          "validation_metric": 0.0007607336156070232,
          "lr_first": 0.0026668332123781697,
          "lr_last": 0.0026552002482112917,
          "seconds": 1.0207394779990864
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009160654866718687,
          "train_loss": 0.009508301503956318,
          "validation_loss": 0.008701737970113754,
          "train_metric": 0.0009576412849128246,
          "validation_metric": 0.0008764071972109377,
          "lr_first": 0.00265441853894982,
          "lr_last": 0.0026426008740574737,
          "seconds": 1.0329267649995018
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.012372840006719343,
          "train_loss": 0.0077375504188239574,
          "validation_loss": 0.006023839116096497,
          "train_metric": 0.0007792976684868336,
          "validation_metric": 0.000606698973570019,
          "lr_first": 0.0026418069185720444,
          "lr_last": 0.0026298065778502713,
          "seconds": 1.0445165139990422
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.009446370342629962,
          "train_loss": 0.008718976750969887,
          "validation_loss": 0.006388162728399038,
          "train_metric": 0.0008781432989053428,
          "validation_metric": 0.0006433923845179379,
          "lr_first": 0.002629000512181069,
          "lr_last": 0.0026168195518264732,
          "seconds": 1.0563280820006185
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.009182927693473175,
          "train_loss": 0.007619915064424276,
          "validation_loss": 0.006061785854399204,
          "train_metric": 0.0007674499647691846,
          "validation_metric": 0.0006105208303779364,
          "lr_first": 0.002616001514088701,
          "lr_last": 0.0026036420212461322,
          "seconds": 1.0684839490004379
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007525247550802305,
          "train_loss": 0.007059510797262192,
          "validation_loss": 0.005410072859376669,
          "train_metric": 0.0007110079750418663,
          "validation_metric": 0.0005448826705105603,
          "lr_first": 0.0026028121516063455,
          "lr_last": 0.0025902762440112768,
          "seconds": 1.0809749599993665
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007746773786493577,
          "train_loss": 0.007113334257155657,
          "validation_loss": 0.006275789346545935,
          "train_metric": 0.0007164290291257203,
          "validation_metric": 0.0006320745451375842,
          "lr_first": 0.0025894346846633653,
          "lr_last": 0.002576724510279026,
          "seconds": 1.0929607079997368
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0071924759686226025,
          "train_loss": 0.006564946845173836,
          "validation_loss": 0.005219299346208572,
          "train_metric": 0.0006611973512917757,
          "validation_metric": 0.0005256687290966511,
          "lr_first": 0.0025758714054198526,
          "lr_last": 0.0025629891420691935,
          "seconds": 1.1055105539999204
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.008716863827430643,
          "train_loss": 0.007181818131357431,
          "validation_loss": 0.00621588621288538,
          "train_metric": 0.0007233264623209834,
          "validation_metric": 0.0006260413792915642,
          "lr_first": 0.0025621246378738837,
          "lr_last": 0.0025490724928664103,
          "seconds": 1.1172558130001562
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007154869308578782,
          "train_loss": 0.006395393051207066,
          "validation_loss": 0.005038523580878973,
          "train_metric": 0.0006441204459406435,
          "validation_metric": 0.000507461663801223,
          "lr_first": 0.0025481967374633083,
          "lr_last": 0.002534976947216873,
          "seconds": 1.1295314590006456
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.0069932489859638736,
          "train_loss": 0.006367839872837067,
          "validation_loss": 0.004876235965639353,
          "train_metric": 0.000641345395706594,
          "validation_metric": 0.0004911166033707559,
          "lr_first": 0.002534090090662157,
          "lr_last": 0.002520704920319766,
          "seconds": 1.1419428369990783
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007627464379766025,
          "train_loss": 0.009519575163722038,
          "validation_loss": 0.008810346946120262,
          "train_metric": 0.0009587766835466027,
          "validation_metric": 0.000887345930095762,
          "lr_first": 0.0025198071145717425,
          "lr_last": 0.0025062588576134272,
          "seconds": 1.1541679530000692
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.009238842641934752,
          "train_loss": 0.009743427857756615,
          "validation_loss": 0.008758672513067722,
          "train_metric": 0.0009813224896788597,
          "validation_metric": 0.0008821414667181671,
          "lr_first": 0.0025053502565064887,
          "lr_last": 0.00249164123435633,
          "seconds": 1.166014043999894
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.007725637391558848,
          "train_loss": 0.007653144653886557,
          "validation_loss": 0.00591690419241786,
          "train_metric": 0.0007707965560257435,
          "validation_metric": 0.000595928868278861,
          "lr_first": 0.0024907219935745985,
          "lr_last": 0.0024768545552029684,
          "seconds": 1.1780016270004126
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008741326426388696,
          "train_loss": 0.009318678639829159,
          "validation_loss": 0.006230821367353201,
          "train_metric": 0.0009385432349517941,
          "validation_metric": 0.0006275454652495682,
          "lr_first": 0.0024759248322536198,
          "lr_last": 0.002461901353774695,
          "seconds": 1.1896402799993666
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.007818438207323197,
          "train_loss": 0.0074435025453567505,
          "validation_loss": 0.005035430658608675,
          "train_metric": 0.0007496823091059923,
          "validation_metric": 0.0005071500199846923,
          "lr_first": 0.0024609613079609708,
          "lr_last": 0.0024467841922255927,
          "seconds": 1.2013049769993813
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.006660373794147745,
          "train_loss": 0.0063566043972969055,
          "validation_loss": 0.0055084992200136185,
          "train_metric": 0.000640213955193758,
          "validation_metric": 0.0005547958426177502,
          "lr_first": 0.0024458339846195024,
          "lr_last": 0.0024315056608034706,
          "seconds": 1.2968860080000013
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006788841186789796,
          "train_loss": 0.005891900043934584,
          "validation_loss": 0.0045007457956671715,
          "train_metric": 0.0005934105720371008,
          "validation_metric": 0.00045329862041398883,
          "lr_first": 0.002430545454218197,
          "lr_last": 0.0024160683774060343,
          "seconds": 1.3097311300007277
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006943465006770566,
          "train_loss": 0.006236498709768057,
          "validation_loss": 0.004605951718986034,
          "train_metric": 0.0006281172973103821,
          "validation_metric": 0.0004638946265913546,
          "lr_first": 0.0024150983363680325,
          "lr_last": 0.0024004749871323217,
          "seconds": 1.3216573200006678
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006563952469150536,
          "train_loss": 0.005726234056055546,
          "validation_loss": 0.004905210342258215,
          "train_metric": 0.0005767252878285944,
          "validation_metric": 0.0004940348444506526,
          "lr_first": 0.0023994952778531296,
          "lr_last": 0.0023847281618294826,
          "seconds": 1.3335632369999075
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.009953707529348321,
          "train_loss": 0.015849187970161438,
          "validation_loss": 0.012503782287240028,
          "train_metric": 0.0015962723409757018,
          "validation_metric": 0.001259335083886981,
          "lr_first": 0.002383738952177242,
          "lr_last": 0.002368830599634967,
          "seconds": 1.3453266029991937
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.00975933215522673,
          "train_loss": 0.007461745757609606,
          "validation_loss": 0.005131215322762728,
          "train_metric": 0.0007515196921303868,
          "validation_metric": 0.0005167972412891686,
          "lr_first": 0.002367832059105659,
          "lr_last": 0.002352785024514216,
          "seconds": 1.3569939509998221
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006723686252371408,
          "train_loss": 0.005659264512360096,
          "validation_loss": 0.004406708758324385,
          "train_metric": 0.0005699803587049246,
          "validation_metric": 0.0004438275354914367,
          "lr_first": 0.00235177732420262,
          "lr_last": 0.0023365941857939245,
          "seconds": 1.3695990500000335
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.007989417383214459,
          "train_loss": 0.006408713757991791,
          "validation_loss": 0.005779212806373835,
          "train_metric": 0.0006454621907323599,
          "validation_metric": 0.000582061184104532,
          "lr_first": 0.002335577498364301,
          "lr_last": 0.0023202608576909545,
          "seconds": 1.3841026699992653
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007546025037299842,
          "train_loss": 0.008282897993922234,
          "validation_loss": 0.0065544117242097855,
          "train_metric": 0.0008342231158167124,
          "validation_metric": 0.0006601362256333232,
          "lr_first": 0.002319235357347461,
          "lr_last": 0.002303787838836996,
          "seconds": 1.3991826549990947
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.007588236956507899,
          "train_loss": 0.007411908358335495,
          "validation_loss": 0.007084624841809273,
          "train_metric": 0.0007465002709068358,
          "validation_metric": 0.0007135374471545219,
          "lr_first": 0.002302753701293836,
          "lr_last": 0.002287177951799022,
          "seconds": 1.4131227130001207
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.007221795625810046,
          "train_loss": 0.006492907181382179,
          "validation_loss": 0.006022248882800341,
          "train_metric": 0.0006539418827742338,
          "validation_metric": 0.0006065388442948461,
          "lr_first": 0.002286135354250341,
          "lr_last": 0.002270434042595671,
          "seconds": 1.4264191319998645
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.006292606303759385,
          "train_loss": 0.006153725553303957,
          "validation_loss": 0.004680159501731396,
          "train_metric": 0.0006197806796990335,
          "validation_metric": 0.00047136854846030474,
          "lr_first": 0.0022693831636851907,
          "lr_last": 0.002253558980209585,
          "seconds": 1.4390063089995238
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0056350331724388525,
          "train_loss": 0.005052188877016306,
          "validation_loss": 0.004083178471773863,
          "train_metric": 0.000508837983943522,
          "validation_metric": 0.0004112427996005863,
          "lr_first": 0.0022524999999999976,
          "lr_last": 0.0022365556560958325,
          "seconds": 1.452722587999233
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0058930846498697065,
          "train_loss": 0.005011626053601503,
          "validation_loss": 0.004076776560395956,
          "train_metric": 0.0005047526792623103,
          "validation_metric": 0.0004105980333406478,
          "lr_first": 0.0022354887560379452,
          "lr_last": 0.002219426983686468,
          "seconds": 1.4674981690004643
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006013314145093318,
          "train_loss": 0.005227058194577694,
          "validation_loss": 0.004124601371586323,
          "train_metric": 0.0005264501669444144,
          "validation_metric": 0.0004154147463850677,
          "lr_first": 0.002218352346588113,
          "lr_last": 0.0022021758978913284,
          "seconds": 1.4816132830001152
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.006405152089428157,
          "train_loss": 0.005471632815897465,
          "validation_loss": 0.003907328471541405,
          "train_metric": 0.000551082834135741,
          "validation_metric": 0.00039353183819912374,
          "lr_first": 0.0022010937078860405,
          "lr_last": 0.002184805354595156,
          "seconds": 1.4997532990000764
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005659401052980684,
          "train_loss": 0.005079155322164297,
          "validation_loss": 0.0038762520998716354,
          "train_metric": 0.0005115538951940835,
          "validation_metric": 0.0003904019540641457,
          "lr_first": 0.0021837157971106207,
          "lr_last": 0.0021673183301511214,
          "seconds": 1.5132745810005872
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.00734308484243229,
          "train_loss": 0.005679483525454998,
          "validation_loss": 0.005012178793549538,
          "train_metric": 0.0005720168701373041,
          "validation_metric": 0.0005048083257861435,
          "lr_first": 0.002166221591877406,
          "lr_last": 0.0021497178208708407,
          "seconds": 1.5282889310001337
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0054532499925699085,
          "train_loss": 0.004921920131891966,
          "validation_loss": 0.003712284378707409,
          "train_metric": 0.0004957177443429828,
          "validation_metric": 0.00037388771306723356,
          "lr_first": 0.0021486140897284006,
          "lr_last": 0.002132006842510969,
          "seconds": 1.5418456429997605
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.005274805138469674,
          "train_loss": 0.005001896061003208,
          "validation_loss": 0.0036283039953559637,
          "train_metric": 0.0005037725786678493,
          "validation_metric": 0.0003654294996522367,
          "lr_first": 0.002130896307618455,
          "lr_last": 0.0021141884297564742,
          "seconds": 1.5549205980005354
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005232566763879731,
          "train_loss": 0.004613750148564577,
          "validation_loss": 0.0038201620336622,
          "train_metric": 0.00046467999345622957,
          "validation_metric": 0.00038475272594951093,
          "lr_first": 0.0021130712813983224,
          "lr_last": 0.0020962656357006496,
          "seconds": 1.57045257900063
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004977997792593669,
          "train_loss": 0.004707279149442911,
          "validation_loss": 0.0036760992370545864,
          "train_metric": 0.0004740999429486692,
          "validation_metric": 0.00037024327320978045,
          "lr_first": 0.00209514206529448,
          "lr_last": 0.002078241531321991,
          "seconds": 1.5820875099998375
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004968907051079441,
          "train_loss": 0.004667274188250303,
          "validation_loss": 0.003630084916949272,
          "train_metric": 0.0004700707795564085,
          "validation_metric": 0.00036560892476700246,
          "lr_first": 0.0020771117313858087,
          "lr_last": 0.0020601192049579968,
          "seconds": 1.5940452590002678
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0048934913575067185,
          "train_loss": 0.0044909873977303505,
          "validation_loss": 0.0037394831888377666,
          "train_metric": 0.00045231578405946493,
          "validation_metric": 0.0003766270529013127,
          "lr_first": 0.0020589833690772045,
          "lr_last": 0.0020419017617759943,
          "seconds": 1.6060315190006804
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0051152177911717445,
          "train_loss": 0.004460630938410759,
          "validation_loss": 0.0036101939622312784,
          "train_metric": 0.00044925842666998506,
          "validation_metric": 0.0003636055043898523,
          "lr_first": 0.0020407600845702236,
          "lr_last": 0.0020235923232410895,
          "seconds": 1.6184462319997692
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0048191046371357515,
          "train_loss": 0.00451378570869565,
          "validation_loss": 0.003492160700261593,
          "train_metric": 0.00045461198897100985,
          "validation_metric": 0.000351717637386173,
          "lr_first": 0.0020224450003308516,
          "lr_last": 0.002005194026581318,
          "seconds": 1.6309314190002624
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.006026495226251427,
          "train_loss": 0.007725683506578207,
          "validation_loss": 0.00752123910933733,
          "train_metric": 0.0007781025487929583,
          "validation_metric": 0.000757511705160141,
          "lr_first": 0.0020040412545544867,
          "lr_last": 0.001986710024250099,
          "seconds": 1.6428804110000783
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.005038541574322153,
          "train_loss": 0.0044026123359799385,
          "validation_loss": 0.003355843247845769,
          "train_metric": 0.000443415017798543,
          "validation_metric": 0.0003379882255103439,
          "lr_first": 0.0019855520006282273,
          "lr_last": 0.0019681434833860745,
          "seconds": 1.65556831899994
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005976551641651895,
          "train_loss": 0.004706194624304771,
          "validation_loss": 0.003628641599789262,
          "train_metric": 0.0004739907162729651,
          "validation_metric": 0.00036546349292621017,
          "lr_first": 0.0019669804065905482,
          "lr_last": 0.0019494975852704384,
          "seconds": 1.6677334279993374
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004855300685449038,
          "train_loss": 0.004746371880173683,
          "validation_loss": 0.003445365699008107,
          "train_metric": 0.00047803716734051704,
          "validation_metric": 0.0003470046503935009,
          "lr_first": 0.0019483296545884795,
          "lr_last": 0.0019307755247818476,
          "seconds": 1.6803263689998857
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.00469840122241294,
          "train_loss": 0.004415492992848158,
          "validation_loss": 0.0034672627225518227,
          "train_metric": 0.0004447123210411519,
          "validation_metric": 0.00034920996404252946,
          "lr_first": 0.0019296029403323655,
          "lr_last": 0.0019119805098489838,
          "seconds": 1.6923125379998964
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.00489775332971476,
          "train_loss": 0.004656199831515551,
          "validation_loss": 0.003385203657671809,
          "train_metric": 0.00046895534615032375,
          "validation_metric": 0.0003409452619962394,
          "lr_first": 0.001910803472548287,
          "lr_last": 0.0018931157609008988,
          "seconds": 1.7040862199992262
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005341660747944843,
          "train_loss": 0.004638013429939747,
          "validation_loss": 0.0035332085099071264,
          "train_metric": 0.00046712372568435967,
          "validation_metric": 0.0003558518073987216,
          "lr_first": 0.0018919344724282693,
          "lr_last": 0.0018741845103152086,
          "seconds": 1.7161309910006821
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.00609418070234824,
          "train_loss": 0.008235132321715355,
          "validation_loss": 0.008064815774559975,
          "train_metric": 0.0008294123108498752,
          "validation_metric": 0.0008122588624246418,
          "lr_first": 0.0018729991730783457,
          "lr_last": 0.0018551900018642434,
          "seconds": 1.728159484000571
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0057049144234042615,
          "train_loss": 0.004182225558906794,
          "validation_loss": 0.0032588173635303974,
          "train_metric": 0.00042121842852793634,
          "validation_metric": 0.00032821614877320826,
          "lr_first": 0.00185400081896458,
          "lr_last": 0.0018361354901592413,
          "seconds": 1.7404490640001313
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004341783373092767,
          "train_loss": 0.004106246866285801,
          "validation_loss": 0.003416610648855567,
          "train_metric": 0.0004135661292821169,
          "validation_metric": 0.00034410852822475135,
          "lr_first": 0.0018349426653571461,
          "lr_last": 0.001817024240092691,
          "seconds": 1.752506340999389
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004418363161676098,
          "train_loss": 0.0060939304530620575,
          "validation_loss": 0.006017167121171951,
          "train_metric": 0.0006137583404779434,
          "validation_metric": 0.0006060271407477558,
          "lr_first": 0.0018158279777725514,
          "lr_last": 0.0017978595262789054,
          "seconds": 1.846459901000344
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.005630478546663653,
          "train_loss": 0.005900408606976271,
          "validation_loss": 0.005451497621834278,
          "train_metric": 0.0005942676216363907,
          "validation_metric": 0.000549054762814194,
          "lr_first": 0.0017966600314141131,
          "lr_last": 0.001778644632492937,
          "seconds": 1.8588080459994671
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0047086736594792455,
          "train_loss": 0.004396495409309864,
          "validation_loss": 0.003192623145878315,
          "train_metric": 0.0004427989770192653,
          "validation_metric": 0.0003215492470189929,
          "lr_first": 0.0017774421106107614,
          "lr_last": 0.0017593828511079182,
          "seconds": 1.8715132220004307
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.00442559746443294,
          "train_loss": 0.0038928138092160225,
          "validation_loss": 0.003182182554155588,
          "train_metric": 0.000392069952795282,
          "validation_metric": 0.00032049775472842157,
          "lr_first": 0.0017581775082542933,
          "lr_last": 0.0017400774825309276,
          "seconds": 1.884487015000559
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.005110793077619746,
          "train_loss": 0.006064290180802345,
          "validation_loss": 0.005817520897835493,
          "train_metric": 0.0006107731605879962,
          "validation_metric": 0.0005859193624928594,
          "lr_first": 0.0017388695252351471,
          "lr_last": 0.001720731834637488,
          "seconds": 1.8965726490005181
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.005748119379859418,
          "train_loss": 0.004180890042334795,
          "validation_loss": 0.0031902515329420567,
          "train_metric": 0.0004210839397273958,
          "validation_metric": 0.0003213104500900954,
          "lr_first": 0.001719521469876815,
          "lr_last": 0.0017013492222047732,
          "seconds": 1.9085654970003816
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.005522920560906641,
          "train_loss": 0.008886034600436687,
          "validation_loss": 0.00843548309057951,
          "train_metric": 0.0008949688635766506,
          "validation_metric": 0.0008495909860357642,
          "lr_first": 0.0017001366573689788,
          "lr_last": 0.0016819329663436428,
          "seconds": 1.920864891999372
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.005439661872514989,
          "train_loss": 0.004078830126672983,
          "validation_loss": 0.0032799397595226765,
          "train_metric": 0.0004108048160560429,
          "validation_metric": 0.0003303434932604432,
          "lr_first": 0.0016807184091994691,
          "lr_last": 0.0016624863939295827,
          "seconds": 1.9329448209991824
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004483416290895548,
          "train_loss": 0.0041998098604381084,
          "validation_loss": 0.0035440120846033096,
          "train_metric": 0.0004229895130265504,
          "validation_metric": 0.0003569399123080075,
          "lr_first": 0.001661270052585144,
          "lr_last": 0.0016430128370326644,
          "seconds": 1.945026269999289
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004919212842651177,
          "train_loss": 0.004099125973880291,
          "validation_loss": 0.0032409345731139183,
          "train_metric": 0.00041284895269200206,
          "validation_metric": 0.0003264150582253933,
          "lr_first": 0.0016417949199017884,
          "lr_last": 0.0016235156323466118,
          "seconds": 1.9574088459994528
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004840203964704415,
          "train_loss": 0.003972734324634075,
          "validation_loss": 0.0029226206243038177,
          "train_metric": 0.0004001192864961922,
          "validation_metric": 0.00029435561737045646,
          "lr_first": 0.001622296348113129,
          "lr_last": 0.0016039981206170785,
          "seconds": 1.969733260000794
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.005314390393323265,
          "train_loss": 0.00874849408864975,
          "validation_loss": 0.008736027404665947,
          "train_metric": 0.00088111637160182,
          "validation_metric": 0.0008798606577329338,
          "lr_first": 0.0016027776781990647,
          "lr_last": 0.0015844636460692202,
          "seconds": 1.982327826999608
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.005132029982632957,
          "train_loss": 0.0038933013565838337,
          "validation_loss": 0.0030778697691857815,
          "train_metric": 0.0003921190509572625,
          "validation_metric": 0.0003099917958024889,
          "lr_first": 0.0015832422545832014,
          "lr_last": 0.0015649155558346884,
          "seconds": 1.994187410999075
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.004579242682666518,
          "train_loss": 0.004101874306797981,
          "validation_loss": 0.0031458763405680656,
          "train_metric": 0.0004131257883273065,
          "validation_metric": 0.0003168411203660071,
          "lr_first": 0.0015636934245598084,
          "lr_last": 0.0015453571993781102,
          "seconds": 2.006408585999452
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.005377793844672851,
          "train_loss": 0.006626904476433992,
          "validation_loss": 0.004154553636908531,
          "train_metric": 0.0006674375617876649,
          "validation_metric": 0.00041843147482723,
          "lr_first": 0.0015441345377202708,
          "lr_last": 0.0015257919279231778,
          "seconds": 2.01858213199921
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.004391125388792716,
          "train_loss": 0.003837058087810874,
          "validation_loss": 0.003112463280558586,
          "train_metric": 0.0003864544560201466,
          "validation_metric": 0.0003134759026579559,
          "lr_first": 0.0015245689453791606,
          "lr_last": 0.0015062230938784333,
          "seconds": 2.0303002830005425
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004255606669175904,
          "train_loss": 0.0037995155435055494,
          "validation_loss": 0.003206335473805666,
          "train_metric": 0.00038267328636720777,
          "validation_metric": 0.0003229303692933172,
          "lr_first": 0.001505000000000001,
          "lr_last": 0.0014866540502628492,
          "seconds": 2.0422788790001505
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.004818862886168063,
          "train_loss": 0.003718474879860878,
          "validation_loss": 0.0032674747053533792,
          "train_metric": 0.00037451114621944726,
          "validation_metric": 0.000329088099533692,
          "lr_first": 0.0014854310546208404,
          "lr_last": 0.0014670881501313076,
          "seconds": 2.0546667180005898
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.004107354645384476,
          "train_loss": 0.0044875298626720905,
          "validation_loss": 0.003058638656511903,
          "train_metric": 0.00045196758583188057,
          "validation_metric": 0.0003080548485741019,
          "lr_first": 0.0014658654622797308,
          "lr_last": 0.0014475287460000708,
          "seconds": 2.066278604999752
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.004213029802485835,
          "train_loss": 0.0037369458004832268,
          "validation_loss": 0.0028395650442689657,
          "train_metric": 0.00037637149216607213,
          "validation_metric": 0.00028599059442058206,
          "lr_first": 0.0014463065754401938,
          "lr_last": 0.001427979189272343,
          "seconds": 2.078822265000781
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.004090418326086365,
          "train_loss": 0.005348946433514357,
          "validation_loss": 0.005152300000190735,
          "train_metric": 0.0005387262790463865,
          "validation_metric": 0.0005189207731746137,
          "lr_first": 0.0014267577454168008,
          "lr_last": 0.0014084428296640297,
          "seconds": 2.091050630000609
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0055987594605539925,
          "train_loss": 0.0039548128843307495,
          "validation_loss": 0.0033931871876120567,
          "train_metric": 0.00039831429603509605,
          "validation_metric": 0.000341749400831759,
          "lr_first": 0.001407222321800938,
          "lr_last": 0.0013889230146297755,
          "seconds": 2.103357328000129
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0038649923735647462,
          "train_loss": 0.0035602774005383253,
          "validation_loss": 0.0028570997528731823,
          "train_metric": 0.0003585781087167561,
          "validation_metric": 0.00028775661485269666,
          "lr_first": 0.0013877036518868736,
          "lr_last": 0.0013694230887893984,
          "seconds": 2.1156940959990607
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.00370860689872643,
          "train_loss": 0.003594902576878667,
          "validation_loss": 0.0027753515169024467,
          "train_metric": 0.0003620654169935733,
          "validation_metric": 0.0002795231994241476,
          "lr_first": 0.001368205080098215,
          "lr_last": 0.0013499463933548063,
          "seconds": 2.1282525599999644
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0036786648124689236,
          "train_loss": 0.003496788674965501,
          "validation_loss": 0.002707767765969038,
          "train_metric": 0.00035218376433476806,
          "validation_metric": 0.00027271645376458764,
          "lr_first": 0.00134872994741486,
          "lr_last": 0.0013304962655574931,
          "seconds": 2.1407721499999752
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.004056655619933736,
          "train_loss": 0.0037605417892336845,
          "validation_loss": 0.002706445986405015,
          "train_metric": 0.00037874802364967763,
          "validation_metric": 0.00027258333284407854,
          "lr_first": 0.0013292815908005342,
          "lr_last": 0.0013110760380767297,
          "seconds": 2.1531192649999866
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.003693759430461796,
          "train_loss": 0.004808132071048021,
          "validation_loss": 0.003184846369549632,
          "train_metric": 0.00048425744171254337,
          "validation_metric": 0.00032076609204523265,
          "lr_first": 0.0013098633426310243,
          "lr_last": 0.0012916890384685182,
          "seconds": 2.165047598999081
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.004447208091733046,
          "train_loss": 0.004090976435691118,
          "validation_loss": 0.0028589782305061817,
          "train_metric": 0.00041202816646546125,
          "validation_metric": 0.0002879458188544959,
          "lr_first": 0.0012904785301231876,
          "lr_last": 0.0012723385885954383,
          "seconds": 2.177288535000116
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0038563035777769983,
          "train_loss": 0.0034314915537834167,
          "validation_loss": 0.0029673450626432896,
          "train_metric": 0.00034560728818178177,
          "validation_metric": 0.00029886013362556696,
          "lr_first": 0.001271130474764856,
          "lr_last": 0.0012530280040574575,
          "seconds": 2.1889949519991205
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.003497969839372672,
          "train_loss": 0.0038376261945813894,
          "validation_loss": 0.00277746026404202,
          "train_metric": 0.00038651161594316363,
          "validation_metric": 0.0002797356282826513,
          "lr_first": 0.0012518224917457093,
          "lr_last": 0.001233760593623827,
          "seconds": 2.2008623170004284
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.004121909500099719,
          "train_loss": 0.004509721416980028,
          "validation_loss": 0.003170785028487444,
          "train_metric": 0.0004542025562841445,
          "validation_metric": 0.0003193498414475471,
          "lr_first": 0.001232557889389241,
          "lr_last": 0.0012145396586661337,
          "seconds": 2.2124904609991063
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004021929838927463,
          "train_loss": 0.003691594349220395,
          "validation_loss": 0.003199473023414612,
          "train_metric": 0.00037180384970270097,
          "validation_metric": 0.00032223924063146114,
          "lr_first": 0.0012133399685858898,
          "lr_last": 0.0011953684925926351,
          "seconds": 2.2240198389990837
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003933946274628397,
          "train_loss": 0.004135750699788332,
          "validation_loss": 0.004023158457130194,
          "train_metric": 0.0004165376303717494,
          "validation_metric": 0.0004051978175994009,
          "lr_first": 0.0011941720222274513,
          "lr_last": 0.0011762503802839421,
          "seconds": 2.2357829320008022
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0038918339378142264,
          "train_loss": 0.004687186796218157,
          "validation_loss": 0.004611196927726269,
          "train_metric": 0.0004720762372016907,
          "validation_metric": 0.0004644228611141443,
          "lr_first": 0.001175057334642857,
          "lr_last": 0.001157188597530181,
          "seconds": 2.248195015999954
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.004510648119321559,
          "train_loss": 0.007251355331391096,
          "validation_loss": 0.0071368771605193615,
          "train_metric": 0.00073032989166677,
          "validation_metric": 0.0007188001181930304,
          "lr_first": 0.0011559991810354228,
          "lr_last": 0.0011381864104696943,
          "seconds": 2.26004608599942
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.004119929471926298,
          "train_loss": 0.0037245582789182663,
          "validation_loss": 0.002622209955006838,
          "train_metric": 0.0003751238400582224,
          "validation_metric": 0.0002640993916429579,
          "lr_first": 0.0011370008269216568,
          "lr_last": 0.0011192470750294122,
          "seconds": 2.2725963329994556
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.003967557160649449,
          "train_loss": 0.004840286448597908,
          "validation_loss": 0.003246085252612829,
          "train_metric": 0.00048749594134278595,
          "validation_metric": 0.0003269337466917932,
          "lr_first": 0.0011180655275717327,
          "lr_last": 0.0011003738363669622,
          "seconds": 2.3649168350002583
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0036212112463545054,
          "train_loss": 0.00330432690680027,
          "validation_loss": 0.002786109456792474,
          "train_metric": 0.00033279971103183925,
          "validation_metric": 0.0002806067350320518,
          "lr_first": 0.001099196527451715,
          "lr_last": 0.0010815699283146308,
          "seconds": 2.377866320999601
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.003582226905564312,
          "train_loss": 0.003424267517402768,
          "validation_loss": 0.0029956665821373463,
          "train_metric": 0.0003448797215241939,
          "validation_metric": 0.00030171260004863143,
          "lr_first": 0.0010803970596676372,
          "lr_last": 0.0010628385728252573,
          "seconds": 2.389851518000796
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.003739642423170153,
          "train_loss": 0.003392196027562022,
          "validation_loss": 0.003022994613274932,
          "train_metric": 0.0003416495455894619,
          "validation_metric": 0.0003044649783987552,
          "lr_first": 0.0010616703454115225,
          "lr_last": 0.0010441829794201776,
          "seconds": 2.4026278849996743
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.003464009383606026,
          "train_loss": 0.0032913293689489365,
          "validation_loss": 0.0028332225047051907,
          "train_metric": 0.00033149062073789537,
          "validation_metric": 0.0002853517944458872,
          "lr_first": 0.001043019593409455,
          "lr_last": 0.0010256063446392812,
          "seconds": 2.4152933360001043
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.004533227787760552,
          "train_loss": 0.0035237418487668037,
          "validation_loss": 0.003192873438820243,
          "train_metric": 0.000354898365912959,
          "validation_metric": 0.0003215745382476598,
          "lr_first": 0.0010244479993717752,
          "lr_last": 0.001007111851493306,
          "seconds": 2.4272204779990716
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.004044993103889283,
          "train_loss": 0.0037454741541296244,
          "validation_loss": 0.003456474980339408,
          "train_metric": 0.00037723046261817217,
          "validation_metric": 0.0003481235180515796,
          "lr_first": 0.0010059587454455136,
          "lr_last": 0.0009887026689184423,
          "seconds": 2.43961993800076
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.004052969787153415,
          "train_loss": 0.004482593387365341,
          "validation_loss": 0.0030920966528356075,
          "train_metric": 0.00045147037599235773,
          "validation_metric": 0.0003114246646873653,
          "lr_first": 0.0009875549996691494,
          "lr_last": 0.0009703819512333538,
          "seconds": 2.452328650999334
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.004502570132899564,
          "train_loss": 0.0032964753918349743,
          "validation_loss": 0.0026145754382014275,
          "train_metric": 0.0003320089017506689,
          "validation_metric": 0.000263330468442291,
          "lr_first": 0.0009692399154297775,
          "lr_last": 0.0009521528375986995,
          "seconds": 2.4650873680002405
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.005096391541883349,
          "train_loss": 0.003604253986850381,
          "validation_loss": 0.0029142284765839577,
          "train_metric": 0.00036300724605098367,
          "validation_metric": 0.00029351041303016245,
          "lr_first": 0.0009510166309227964,
          "lr_last": 0.0009340184514792582,
          "seconds": 2.4775852860002487
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.004565108669339679,
          "train_loss": 0.0031817839480936527,
          "validation_loss": 0.0025805006735026836,
          "train_metric": 0.00032045767875388265,
          "validation_metric": 0.00025989857385866344,
          "lr_first": 0.0009328882686141923,
          "lr_last": 0.0009159819001087345,
          "seconds": 2.489971938999588
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.003920238545106258,
          "train_loss": 0.003164769383147359,
          "validation_loss": 0.002610706025734544,
          "train_metric": 0.00031874398700892925,
          "validation_metric": 0.00026294076815247536,
          "lr_first": 0.0009148579347055221,
          "lr_last": 0.0008980462739573539,
          "seconds": 2.502392329999566
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0039296135364566,
          "train_loss": 0.004066374618560076,
          "validation_loss": 0.003813035786151886,
          "train_metric": 0.0004095503827556968,
          "validation_metric": 0.0003840350254904479,
          "lr_first": 0.0008969287186016802,
          "lr_last": 0.000880214646202324,
          "seconds": 2.514076161000048
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.004408461012644693,
          "train_loss": 0.004311859142035246,
          "validation_loss": 0.004093051888048649,
          "train_metric": 0.0004342746688053012,
          "validation_metric": 0.00041223716107197106,
          "lr_first": 0.0008791036923815459,
          "lr_last": 0.0008624900722012644,
          "seconds": 2.5255523460000404
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0046408419002546,
          "train_loss": 0.004120219498872757,
          "validation_loss": 0.003873463487252593,
          "train_metric": 0.0004149734159000218,
          "validation_metric": 0.0003901210438925773,
          "lr_first": 0.0008613859102716003,
          "lr_last": 0.0008448755889686828,
          "seconds": 2.537638440999217
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.004189433740975801,
          "train_loss": 0.0037795468233525753,
          "validation_loss": 0.003678865497931838,
          "train_metric": 0.00038066215347498655,
          "validation_metric": 0.0003705218550749123,
          "lr_first": 0.0008437784081225947,
          "lr_last": 0.000827374214655601,
          "seconds": 2.5491863469997043
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.004040297368192114,
          "train_loss": 0.0049334242939949036,
          "validation_loss": 0.00490171043202281,
          "train_metric": 0.0004968764260411263,
          "validation_metric": 0.0004936823388561606,
          "lr_first": 0.0008262842028893788,
          "lr_last": 0.0008099889480324116,
          "seconds": 2.5605384379996394
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.003929976584913675,
          "train_loss": 0.0032478400971740484,
          "validation_loss": 0.002856483682990074,
          "train_metric": 0.0003271105815656483,
          "validation_metric": 0.0002876945654861629,
          "lr_first": 0.0008089062921139598,
          "lr_last": 0.0007927227679750494,
          "seconds": 2.5722041270000773
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0032048387147369795,
          "train_loss": 0.003159562824293971,
          "validation_loss": 0.002577858744189143,
          "train_metric": 0.0003182195941917598,
          "validation_metric": 0.0002596324775367975,
          "lr_first": 0.0007916476534118866,
          "lr_last": 0.0007755786329545818,
          "seconds": 2.5849530180003057
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0032147084584721597,
          "train_loss": 0.0030661607161164284,
          "validation_loss": 0.002475951099768281,
          "train_metric": 0.00030881245038472116,
          "validation_metric": 0.0002493687206879258,
          "lr_first": 0.0007745112439620533,
          "lr_last": 0.0007585594805302867,
          "seconds": 2.59825830400041
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0031313519066316076,
          "train_loss": 0.0030546493362635374,
          "validation_loss": 0.002583725843578577,
          "train_metric": 0.00030765309929847717,
          "validation_metric": 0.00026022340171039104,
          "lr_first": 0.0007575000000000017,
          "lr_last": 0.0007416682268463151,
          "seconds": 2.610891124000773
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.003307227889308706,
          "train_loss": 0.0031381601002067327,
          "validation_loss": 0.0024257302284240723,
          "train_metric": 0.0003160640480928123,
          "validation_metric": 0.0002443106786813587,
          "lr_first": 0.0007406168363148086,
          "lr_last": 0.000724907766132032,
          "seconds": 2.6234625639990554
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0033125431727967225,
          "train_loss": 0.003124601673334837,
          "validation_loss": 0.0023901560343801975,
          "train_metric": 0.00031469843816012144,
          "validation_metric": 0.00024072773521766067,
          "lr_first": 0.0007238646457496579,
          "lr_last": 0.0007082809702060954,
          "seconds": 2.6365726919993904
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.003229122215998359,
          "train_loss": 0.0031805001199245453,
          "validation_loss": 0.002417817013338208,
          "train_metric": 0.00032032831222750247,
          "validation_metric": 0.00024351367028430104,
          "lr_first": 0.000707246298706162,
          "lr_last": 0.0006917906879843936,
          "seconds": 2.6490183450005134
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0032195117746596225,
          "train_loss": 0.0032845395617187023,
          "validation_loss": 0.002978098811581731,
          "train_metric": 0.0003308068262413144,
          "validation_metric": 0.00029994320357218385,
          "lr_first": 0.0006907646426525363,
          "lr_last": 0.0006754397449918927,
          "seconds": 2.6610860010005126
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0032228408235823736,
          "train_loss": 0.003155919723212719,
          "validation_loss": 0.00283241318538785,
          "train_metric": 0.0003178526530973613,
          "validation_metric": 0.0002852702746167779,
          "lr_first": 0.0006744225016356963,
          "lr_last": 0.0006592309428784994,
          "seconds": 2.6727976629990735
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0031694747049186844,
          "train_loss": 0.003044720506295562,
          "validation_loss": 0.0024011675268411636,
          "train_metric": 0.0003066530916839838,
          "validation_metric": 0.00024183682398870587,
          "lr_first": 0.0006582226757973757,
          "lr_last": 0.0006431670589390128,
          "seconds": 2.6850589539990324
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0031651046738261357,
          "train_loss": 0.003379028057679534,
          "validation_loss": 0.003111174562945962,
          "train_metric": 0.0003403233422432095,
          "validation_metric": 0.00031334609957411885,
          "lr_first": 0.0006421679408943365,
          "lr_last": 0.0006272508456372499,
          "seconds": 2.697697349000009
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0030342688696691766,
          "train_loss": 0.003279397264122963,
          "validation_loss": 0.0024236785247921944,
          "train_metric": 0.00033028892357833683,
          "validation_metric": 0.0002441039978293702,
          "lr_first": 0.0006262610478227541,
          "lr_last": 0.0006114850301344212,
          "seconds": 2.709592932000305
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.003238496763515286,
          "train_loss": 0.002974162343889475,
          "validation_loss": 0.002384082181379199,
          "train_metric": 0.0002995467511937022,
          "validation_metric": 0.0002401160600129515,
          "lr_first": 0.0006105047221468661,
          "lr_last": 0.0005958723138218511,
          "seconds": 2.722053995999886
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0030744769974262454,
          "train_loss": 0.0029732189141213894,
          "validation_loss": 0.002393994014710188,
          "train_metric": 0.00029945169808343053,
          "validation_metric": 0.00024111432139761746,
          "lr_first": 0.0005949016636319636,
          "lr_last": 0.0005804153718581059,
          "seconds": 2.7344889609994425
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.003065528617298696,
          "train_loss": 0.0029935531783849,
          "validation_loss": 0.002558706095442176,
          "train_metric": 0.0003014997055288404,
          "validation_metric": 0.00025770350475795567,
          "lr_first": 0.0005794545457817997,
          "lr_last": 0.0005651168527106182,
          "seconds": 2.7465653399995063
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.003151561279082671,
          "train_loss": 0.003064841264858842,
          "validation_loss": 0.0023917360231280327,
          "train_metric": 0.0003086795622948557,
          "validation_metric": 0.00024088691861834377,
          "lr_first": 0.0005641660153804942,
          "lr_last": 0.0005499793777018868,
          "seconds": 2.758583210999859
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0030686569825775223,
          "train_loss": 0.0032233840320259333,
          "validation_loss": 0.0029705066699534655,
          "train_metric": 0.00032464746618643403,
          "validation_metric": 0.0002991785586345941,
          "lr_first": 0.000549038692039027,
          "lr_last": 0.0005350055405603265,
          "seconds": 2.771051059999081
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0032028430359787308,
          "train_loss": 0.0029527463484555483,
          "validation_loss": 0.002394797746092081,
          "train_metric": 0.0002973897790070623,
          "validation_metric": 0.0002411952445982024,
          "lr_first": 0.0005340751677463768,
          "lr_last": 0.0005201979069758457,
          "seconds": 2.7831194570007938
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.003187599271768704,
          "train_loss": 0.003438633168116212,
          "validation_loss": 0.0025085569359362125,
          "train_metric": 0.00034632650204002857,
          "validation_metric": 0.00025265265139751136,
          "lr_first": 0.0005192780064253987,
          "lr_last": 0.0005055590141602302,
          "seconds": 2.795117219000531
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.003315675705380272,
          "train_loss": 0.003455646801739931,
          "validation_loss": 0.003189926501363516,
          "train_metric": 0.00034804007736966014,
          "validation_metric": 0.0003212777082808316,
          "lr_first": 0.0005046497434935084,
          "lr_last": 0.0004910913704124016,
          "seconds": 2.88783082000009
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0031897933076834306,
          "train_loss": 0.002928799716755748,
          "validation_loss": 0.0024893053341656923,
          "train_metric": 0.0002949779445771128,
          "validation_metric": 0.0002507137251086533,
          "lr_first": 0.0004901928854282528,
          "lr_last": 0.0004767974546886371,
          "seconds": 2.9007356830006756
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.003024768695468083,
          "train_loss": 0.002931142458692193,
          "validation_loss": 0.0023212991654872894,
          "train_metric": 0.0002952139766421169,
          "validation_metric": 0.0002337927435291931,
          "lr_first": 0.0004759099093378383,
          "lr_last": 0.00046267971617781114,
          "seconds": 2.913924810000026
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.002995410955918487,
          "train_loss": 0.0029093832708895206,
          "validation_loss": 0.0023193289525806904,
          "train_metric": 0.0002930224291048944,
          "validation_metric": 0.00023359432816505432,
          "lr_first": 0.0004618032625366892,
          "lr_last": 0.00044874057388173796,
          "seconds": 2.9267150680007035
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0030675327725475654,
          "train_loss": 0.002911194460466504,
          "validation_loss": 0.0023199720308184624,
          "train_metric": 0.0002932048519141972,
          "validation_metric": 0.00023365909873973578,
          "lr_first": 0.0004478753621261122,
          "lr_last": 0.00043498241620069174,
          "seconds": 2.939620200000718
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.002955024901893921,
          "train_loss": 0.0029195817187428474,
          "validation_loss": 0.002358601661399007,
          "train_metric": 0.00029404956148937345,
          "validation_metric": 0.00023754974245093763,
          "lr_first": 0.00043412859458014243,
          "lr_last": 0.0004214076005241641,
          "seconds": 2.9519208109995816
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.002980225093779154,
          "train_loss": 0.0029070775490254164,
          "validation_loss": 0.0024748998694121838,
          "train_metric": 0.0002927902387455106,
          "validation_metric": 0.0002492628700565547,
          "lr_first": 0.0004205653153366304,
          "lr_last": 0.00040801845282693807,
          "seconds": 2.9686342410004727
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0029665832698810846,
          "train_loss": 0.002870069583877921,
          "validation_loss": 0.0023697579745203257,
          "train_metric": 0.0002890629111789167,
          "validation_metric": 0.00023867335403338075,
          "lr_first": 0.0004071878483936508,
          "lr_last": 0.0003948172672705435,
          "seconds": 2.987980576999689
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.002971899306430714,
          "train_loss": 0.00299074430949986,
          "validation_loss": 0.002307172166183591,
          "train_metric": 0.00030121684540063143,
          "validation_metric": 0.00023236995912157,
          "lr_first": 0.00039399848591129634,
          "lr_last": 0.0003818063058101653,
          "seconds": 3.000789309000538
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0030476244537567254,
          "train_loss": 0.002863295143470168,
          "validation_loss": 0.002385093364864588,
          "train_metric": 0.00028838060097768903,
          "validation_metric": 0.0002402178943157196,
          "lr_first": 0.0003809994878189294,
          "lr_last": 0.00036898779780706985,
          "seconds": 3.0127339659993595
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.003059105274587637,
          "train_loss": 0.0029881123919039965,
          "validation_loss": 0.0026623117737472057,
          "train_metric": 0.00030095173860900104,
          "validation_metric": 0.0002681383048184216,
          "lr_first": 0.00036819308142795417,
          "lr_last": 0.000356363939646614,
          "seconds": 3.0245094990004873
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0029568600475613493,
          "train_loss": 0.002946657594293356,
          "validation_loss": 0.0022956246975809336,
          "train_metric": 0.0002967765904031694,
          "validation_metric": 0.00023120689729694277,
          "lr_first": 0.0003555814610501785,
          "lr_last": 0.0003439368943619046,
          "seconds": 3.0369259399994917
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0029645019967574626,
          "train_loss": 0.0030668452382087708,
          "validation_loss": 0.002781570190563798,
          "train_metric": 0.0003088814555667341,
          "validation_metric": 0.00028014954295940697,
          "lr_first": 0.0003431667876218289,
          "lr_last": 0.0003317087912631771,
          "seconds": 3.049235766999118
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0032282690845022444,
          "train_loss": 0.0028725031297653913,
          "validation_loss": 0.002435979200527072,
          "train_metric": 0.00028930799453519285,
          "validation_metric": 0.0002453429333399981,
          "lr_first": 0.00033095118833328664,
          "lr_last": 0.00031968172557294704,
          "seconds": 3.061209362000227
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0029285873752087355,
          "train_loss": 0.0028447408694773912,
          "validation_loss": 0.002355503384023905,
          "train_metric": 0.00028651190223172307,
          "validation_metric": 0.0002372376766288653,
          "lr_first": 0.00031893675626460375,
          "lr_last": 0.00030785775806700536,
          "seconds": 3.073636910999994
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002875678808777593,
          "train_loss": 0.0028395922854542732,
          "validation_loss": 0.002371476497501135,
          "train_metric": 0.00028599335928447545,
          "validation_metric": 0.00023884643451310694,
          "lr_first": 0.000307125550026865,
          "lr_last": 0.00029623891472131333,
          "seconds": 3.0861024820005696
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0028810848598368466,
          "train_loss": 0.002830717246979475,
          "validation_loss": 0.002346825785934925,
          "train_metric": 0.0002850994933396578,
          "validation_metric": 0.00023636370315216482,
          "lr_first": 0.0002955195934094543,
          "lr_last": 0.00028482718636486487,
          "seconds": 3.0979241620007087
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.003046343816095032,
          "train_loss": 0.002829494886100292,
          "validation_loss": 0.00235173967666924,
          "train_metric": 0.0002849763841368258,
          "validation_metric": 0.00023685861378908157,
          "lr_first": 0.0002841208750332906,
          "lr_last": 0.0002736245283385659,
          "seconds": 3.109923649999473
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0029951735341455787,
          "train_loss": 0.0028498207684606314,
          "validation_loss": 0.0023584456648677588,
          "train_metric": 0.000287023518467322,
          "validation_metric": 0.00023753401183057576,
          "lr_first": 0.00027293134801008756,
          "lr_last": 0.00026263286016019817,
          "seconds": 3.121662739000385
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.0033657952735666186,
          "train_loss": 0.0031220640521496534,
          "validation_loss": 0.0028614038601517677,
          "train_metric": 0.00031444287742488086,
          "validation_metric": 0.0002881901164073497,
          "lr_first": 0.0002619529296076959,
          "lr_last": 0.0002518540651955178,
          "seconds": 3.133760718999838
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.003317658534797374,
          "train_loss": 0.0029006963595747948,
          "validation_loss": 0.00229745265096426,
          "train_metric": 0.00029214753885753453,
          "validation_metric": 0.00023139100812841207,
          "lr_first": 0.00025118750092159196,
          "lr_last": 0.0002412899903355538,
          "seconds": 3.1460920259996783
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.00301753226085566,
          "train_loss": 0.0028293917421251535,
          "validation_loss": 0.0022763079032301903,
          "train_metric": 0.00028496593586169183,
          "validation_metric": 0.00022926137899048626,
          "lr_first": 0.00024063690655256104,
          "lr_last": 0.00023094244568014875,
          "seconds": 3.158576051999262
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0028809879113396164,
          "train_loss": 0.0028296576347202063,
          "validation_loss": 0.00240425206720829,
          "train_metric": 0.0002849927404895425,
          "validation_metric": 0.00024214749282691628,
          "lr_first": 0.0002303029542906331,
          "lr_last": 0.00022081320422781038,
          "seconds": 3.170495940999899
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0028862695617135614,
          "train_loss": 0.0028058448806405067,
          "validation_loss": 0.0023175315000116825,
          "train_metric": 0.00028259438113309443,
          "validation_metric": 0.0002334132877876982,
          "lr_first": 0.00022018741480532905,
          "lr_last": 0.00021090400157191562,
          "seconds": 3.182710199000212
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.002889569914259482,
          "train_loss": 0.0028034127317368984,
          "validation_loss": 0.0023252975661307573,
          "train_metric": 0.00028234950150363147,
          "validation_metric": 0.00023419548233505338,
          "lr_first": 0.00021029202134226492,
          "lr_last": 0.00020121653560332717,
          "seconds": 3.1949450579995755
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.003011568715010071,
          "train_loss": 0.002819349989295006,
          "validation_loss": 0.0022992962040007114,
          "train_metric": 0.00028395463596098125,
          "validation_metric": 0.00023157669056672603,
          "lr_first": 0.00020061846942616898,
          "lr_last": 0.00019175246621946768,
          "seconds": 3.206679852999514
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0029234046523924917,
          "train_loss": 0.002821162575855851,
          "validation_loss": 0.002296356251463294,
          "train_metric": 0.0002841371460817754,
          "validation_metric": 0.00023128061729948968,
          "lr_first": 0.0001911684165703626,
          "lr_last": 0.00018251341503990455,
          "seconds": 3.2187840299993695
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0028236772923264652,
          "train_loss": 0.0028087743557989597,
          "validation_loss": 0.0023533711209893227,
          "train_metric": 0.0002828894939739257,
          "validation_metric": 0.0002370229340158403,
          "lr_first": 0.00018194348199275317,
          "lr_last": 0.00017350096512849466,
          "seconds": 3.231181398999979
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0028303528742981143,
          "train_loss": 0.0027953297831118107,
          "validation_loss": 0.002319583436474204,
          "train_metric": 0.0002815354382619262,
          "validation_metric": 0.00023361992498394102,
          "lr_first": 0.00017294524633839077,
          "lr_last": 0.00016471666072213429,
          "seconds": 3.2431829619999917
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0028352485824143514,
          "train_loss": 0.0027956790290772915,
          "validation_loss": 0.002330092014744878,
          "train_metric": 0.0002815705956891179,
          "validation_metric": 0.00023467834398616105,
          "lr_first": 0.00016417525140863167,
          "lr_last": 0.00015616200696616197,
          "seconds": 3.2550238519997947
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.002844275935785845,
          "train_loss": 0.002800989430397749,
          "validation_loss": 0.002270342782139778,
          "train_metric": 0.00028210534946992993,
          "validation_metric": 0.00022866061772219837,
          "lr_first": 0.00015563499989695927,
          "lr_last": 0.00014783846965645865,
          "seconds": 3.26762952900026
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0028206057904753834,
          "train_loss": 0.002808370627462864,
          "validation_loss": 0.002392958616837859,
          "train_metric": 0.00028284877771511674,
          "validation_metric": 0.00024101004237309098,
          "lr_first": 0.00014732595513150429,
          "lr_last": 0.0001397474749882931,
          "seconds": 3.2801046929998847
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0028644638950936496,
          "train_loss": 0.002789258724078536,
          "validation_loss": 0.0023236500564962626,
          "train_metric": 0.00028092393768019974,
          "validation_metric": 0.00023402951774187386,
          "lr_first": 0.0001392495408243121,
          "lr_last": 0.00013189040931194934,
          "seconds": 3.2920813679993444
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0028159845678601414,
          "train_loss": 0.002805313328281045,
          "validation_loss": 0.002263243542984128,
          "train_metric": 0.0002825408591888845,
          "validation_metric": 0.00022794559481553733,
          "lr_first": 0.00013140714082739743,
          "lr_last": 0.00012426861889518266,
          "seconds": 3.3051473289997375
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.002847984476829879,
          "train_loss": 0.0027984222397208214,
          "validation_loss": 0.0023639767896384,
          "train_metric": 0.0002818467910401523,
          "validation_metric": 0.00023809107369743288,
          "lr_first": 0.00012380009889562699,
          "lr_last": 0.00011688340969254351,
          "seconds": 3.3171016720007174
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0028738930450344924,
          "train_loss": 0.002824904629960656,
          "validation_loss": 0.002239382127299905,
          "train_metric": 0.00028451404068619013,
          "validation_metric": 0.000225542375119403,
          "lr_first": 0.0001164297184564739,
          "lr_last": 0.00010973604712160967,
          "seconds": 3.3296205919996282
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.002803988587402273,
          "train_loss": 0.00282206479460001,
          "validation_loss": 0.002431299537420273,
          "train_metric": 0.0002842280373442918,
          "validation_metric": 0.00024487156770192087,
          "lr_first": 0.00010929726238668399,
          "lr_last": 0.00010282775584616308,
          "seconds": 3.424289606000457
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0028306910462561063,
          "train_loss": 0.002779656555503607,
          "validation_loss": 0.002291738986968994,
          "train_metric": 0.0002799568173941225,
          "validation_metric": 0.00023081556719262153,
          "lr_first": 0.00010240395279588673,
          "lr_last": 9.615971956635015e-05,
          "seconds": 3.436765483000272
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.00289959105066373,
          "train_loss": 0.00279002427123487,
          "validation_loss": 0.0023589017800986767,
          "train_metric": 0.00028100103372707963,
          "validation_metric": 0.00023757995222695172,
          "lr_first": 9.575097081719392e-05,
          "lr_last": 8.973308081585969e-05,
          "seconds": 3.4490715869997075
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0028213033474457916,
          "train_loss": 0.0028015202842652798,
          "validation_loss": 0.0022431863471865654,
          "train_metric": 0.0002821589005179703,
          "validation_metric": 0.00022592549794353545,
          "lr_first": 8.933945640481766e-05,
          "lr_last": 8.354894076615733e-05,
          "seconds": 3.46124788599991
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.002853176680218894,
          "train_loss": 0.002785155549645424,
          "validation_loss": 0.002345575951039791,
          "train_metric": 0.0002805106632877141,
          "validation_metric": 0.00023623781453352422,
          "lr_first": 8.317050813874592e-05,
          "lr_last": 7.760835903780457e-05,
          "seconds": 3.4728253840003163
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0028200562846905086,
          "train_loss": 0.0027899155393242836,
          "validation_loss": 0.0022468212991952896,
          "train_metric": 0.000280990032479167,
          "validation_metric": 0.00022629162413068116,
          "lr_first": 7.724518303650655e-05,
          "lr_last": 7.191235351889875e-05,
          "seconds": 3.484698648999256
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.002789554353512358,
          "train_loss": 0.0027743384707719088,
          "validation_loss": 0.002298033330589533,
          "train_metric": 0.00027942124870605767,
          "validation_metric": 0.00023144949227571487,
          "lr_first": 7.156449637205192e-05,
          "lr_last": 6.646190019066193e-05,
          "seconds": 3.496924893999676
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0028346268663881347,
          "train_loss": 0.002778013236820698,
          "validation_loss": 0.002332567237317562,
          "train_metric": 0.0002797913330141455,
          "validation_metric": 0.0002349276328459382,
          "lr_first": 6.61294215017977e-05,
          "lr_last": 6.125793296021323e-05,
          "seconds": 3.5086795030001667
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0027687265755957924,
          "train_loss": 0.0027732402086257935,
          "validation_loss": 0.0022783384192734957,
          "train_metric": 0.0002793105668388307,
          "validation_metric": 0.00022946586250327528,
          "lr_first": 6.094088969784328e-05,
          "lr_last": 5.630134350054723e-05,
          "seconds": 3.5204172159992595
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0028203403708175756,
          "train_loss": 0.002771972445771098,
          "validation_loss": 0.002283102134242654,
          "train_metric": 0.0002791829174384475,
          "validation_metric": 0.00022994571190793067,
          "lr_first": 5.599978998840269e-05,
          "lr_last": 5.1592981097751505e-05,
          "seconds": 3.532525816999623
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.002791501443425659,
          "train_loss": 0.0027729703579097986,
          "validation_loss": 0.0022729020565748215,
          "train_metric": 0.0002792834129650146,
          "validation_metric": 0.00022891839034855366,
          "lr_first": 5.130696900547388e-05,
          "lr_last": 4.7133652505484666e-05,
          "seconds": 3.544768538999051
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.002780102084216196,
          "train_loss": 0.0027708997949957848,
          "validation_loss": 0.002295117126777768,
          "train_metric": 0.0002790748549159616,
          "validation_metric": 0.00023115580552257597,
          "lr_first": 4.686323083977387e-05,
          "lr_last": 4.292412180674455e-05,
          "seconds": 3.5566289689995756
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0027809544699266553,
          "train_loss": 0.002769840881228447,
          "validation_loss": 0.0022919897455722094,
          "train_metric": 0.00027896821848116815,
          "validation_metric": 0.00023084081476554275,
          "lr_first": 4.266933690296076e-05,
          "lr_last": 3.896511028294589e-05,
          "seconds": 3.5689373320001323
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0027921399159822613,
          "train_loss": 0.0027716525364667177,
          "validation_loss": 0.0022676661610603333,
          "train_metric": 0.00027915069949813187,
          "validation_metric": 0.00022839104349259287,
          "lr_first": 3.872600579717075e-05,
          "lr_last": 3.52572962903319e-05,
          "seconds": 3.581299306000801
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.002783896132314112,
          "train_loss": 0.0027690776623785496,
          "validation_loss": 0.0022945101372897625,
          "train_metric": 0.0002788913552649319,
          "validation_metric": 0.00023109464382287115,
          "lr_first": 3.503391319188826e-05,
          "lr_last": 3.18013151437425e-05,
          "seconds": 3.594107499000529
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0027837993256980553,
          "train_loss": 0.002768389182165265,
          "validation_loss": 0.0022935180459171534,
          "train_metric": 0.0002788220008369535,
          "validation_metric": 0.00023099475947674364,
          "lr_first": 3.159369170817401e-05,
          "lr_last": 2.8597759007755454e-05,
          "seconds": 3.6060191489996214
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.0027781518947449513,
          "train_loss": 0.0027680096682161093,
          "validation_loss": 0.002288673771545291,
          "train_metric": 0.0002787838166113943,
          "validation_metric": 0.0002305068337591365,
          "lr_first": 2.8405930810269434e-05,
          "lr_last": 2.5647176795223176e-05,
          "seconds": 3.6181918250003946
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0027760677148762625,
          "train_loss": 0.0027681482024490833,
          "validation_loss": 0.0023019874934107065,
          "train_metric": 0.00027879775734618306,
          "validation_metric": 0.0002318477927474305,
          "lr_first": 2.5471176704593375e-05,
          "lr_last": 2.295007407321785e-05,
          "seconds": 3.630137853000633
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.0027739885772462003,
          "train_loss": 0.002767356112599373,
          "validation_loss": 0.0022858199663460255,
          "train_metric": 0.0002787179546430707,
          "validation_metric": 0.00023021944798529148,
          "lr_first": 2.2789932246153667e-05,
          "lr_last": 2.0506912976406118e-05,
          "seconds": 3.642318255999271
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0027743585378630087,
          "train_loss": 0.0027670746203511953,
          "validation_loss": 0.002289380179718137,
          "train_metric": 0.00027868966571986675,
          "validation_metric": 0.0002305779926246032,
          "lr_first": 2.036265685238534e-05,
          "lr_last": 1.831811212786458e-05,
          "seconds": 3.654524875999414
        }
      ],
      "best": {
        "epoch": 212,
        "updates": 3392
      },
      "best_validation_metric": 0.000225542375119403,
      "config": {
        "label": "Cosine decay",
        "batch": 64,
        "lr": 0.003,
        "schedule": "cosine",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.015018068288545069
    },
    "ed8dd0384733_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.18389458605088294,
          "train_loss": 0.06916050612926483,
          "validation_loss": 0.06209268793463707,
          "train_metric": 0.006965593434870243,
          "validation_metric": 0.0062537482008337975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.015022739999039914
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.045738674933090806,
          "train_loss": 0.03489449992775917,
          "validation_loss": 0.032735906541347504,
          "train_metric": 0.0035144463181495667,
          "validation_metric": 0.003297040704637766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.029512199998862343
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.02863602398429066,
          "train_loss": 0.02373126894235611,
          "validation_loss": 0.017781876027584076,
          "train_metric": 0.0023901264648884535,
          "validation_metric": 0.0017909255111590028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04414163399997051
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.03047496487852186,
          "train_loss": 0.018814587965607643,
          "validation_loss": 0.01837657205760479,
          "train_metric": 0.001894936547614634,
          "validation_metric": 0.001850820961408317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05794547499863256
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.01888181286631152,
          "train_loss": 0.01885739341378212,
          "validation_loss": 0.016206376254558563,
          "train_metric": 0.001899247639812529,
          "validation_metric": 0.0016322467708960176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08596314299938967
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.01619607725297101,
          "train_loss": 0.011927196756005287,
          "validation_loss": 0.011007381603121758,
          "train_metric": 0.0012012633960694075,
          "validation_metric": 0.001108623226173222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09996576199955598
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.012518655508756638,
          "train_loss": 0.010872814804315567,
          "validation_loss": 0.01140688918530941,
          "train_metric": 0.0010950700379908085,
          "validation_metric": 0.0011488600866869092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11335899799996696
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.009969741659006104,
          "train_loss": 0.009192890487611294,
          "validation_loss": 0.009333253838121891,
          "train_metric": 0.000925874279346317,
          "validation_metric": 0.00094001111574471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12724858499859693
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.010474069189513102,
          "train_loss": 0.00656439084559679,
          "validation_loss": 0.00658778939396143,
          "train_metric": 0.0006611414137296379,
          "validation_metric": 0.0006634980672970414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1410995139995066
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.013046462088823318,
          "train_loss": 0.023848779499530792,
          "validation_loss": 0.01683362014591694,
          "train_metric": 0.0024019619449973106,
          "validation_metric": 0.0016954205930233002,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15414160099862784
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0150880494620651,
          "train_loss": 0.01769361086189747,
          "validation_loss": 0.013639772310853004,
          "train_metric": 0.0017820355715230107,
          "validation_metric": 0.0013737480621784925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16711052099890367
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.013406007667072117,
          "train_loss": 0.010042965412139893,
          "validation_loss": 0.01069402415305376,
          "train_metric": 0.0010114905890077353,
          "validation_metric": 0.0010770629160106182,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18045739799890725
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.007156405947171152,
          "train_loss": 0.0063883112743496895,
          "validation_loss": 0.0051514930091798306,
          "train_metric": 0.0006434072856791317,
          "validation_metric": 0.0005188394570723176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1941643539994402
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.006110915433964692,
          "train_loss": 0.005998062901198864,
          "validation_loss": 0.008657868020236492,
          "train_metric": 0.0006041027954779565,
          "validation_metric": 0.0008719888282939792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20757954699911352
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.0062475081358570606,
          "train_loss": 0.01197153702378273,
          "validation_loss": 0.007577677723020315,
          "train_metric": 0.0012057292042300105,
          "validation_metric": 0.0007631958578713238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2205952949989296
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.0113911721855402,
          "train_loss": 0.008418227545917034,
          "validation_loss": 0.005429287441074848,
          "train_metric": 0.0008478531381115317,
          "validation_metric": 0.0005468179006129503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23392090400011512
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.006947965404833667,
          "train_loss": 0.006220518611371517,
          "validation_loss": 0.004411776550114155,
          "train_metric": 0.0006265078554861248,
          "validation_metric": 0.00044433801667764783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24767749899910996
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.015920912410365418,
          "train_loss": 0.005532862152904272,
          "validation_loss": 0.0060539236292243,
          "train_metric": 0.0005572495865635574,
          "validation_metric": 0.0006097290315665305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2610718949999864
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.0066804879024857655,
          "train_loss": 0.005352771375328302,
          "validation_loss": 0.006394522730261087,
          "train_metric": 0.0005391114973463118,
          "validation_metric": 0.0006440328434109688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2742253950000304
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.006032283257809468,
          "train_loss": 0.007059706840664148,
          "validation_loss": 0.006349417380988598,
          "train_metric": 0.0007110278238542378,
          "validation_metric": 0.000639490142930299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2874597709997033
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.011215290942345746,
          "train_loss": 0.010435744188725948,
          "validation_loss": 0.007207291200757027,
          "train_metric": 0.0010510500287637115,
          "validation_metric": 0.0007258919067680836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30072923300031107
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.006948341120732948,
          "train_loss": 0.004147992469370365,
          "validation_loss": 0.004910673946142197,
          "train_metric": 0.0004177706432528794,
          "validation_metric": 0.0004945851396769285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31467731799966714
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.005095057058497332,
          "train_loss": 0.005464919842779636,
          "validation_loss": 0.00746883312240243,
          "train_metric": 0.000550406810361892,
          "validation_metric": 0.0007522335508838296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3288703259986505
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.005976300351903774,
          "train_loss": 0.006015682592988014,
          "validation_loss": 0.004650956951081753,
          "train_metric": 0.0006058774888515472,
          "validation_metric": 0.00046842731535434723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34254120900004636
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.008279251887870487,
          "train_loss": 0.004652548581361771,
          "validation_loss": 0.004610558971762657,
          "train_metric": 0.0004685877065639943,
          "validation_metric": 0.00046435859985649586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3564520370000537
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.005110172278364189,
          "train_loss": 0.004339386243373156,
          "validation_loss": 0.003442443208768964,
          "train_metric": 0.0004370470705907792,
          "validation_metric": 0.0003467102360446006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37093613099932554
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.004839025321416557,
          "train_loss": 0.004047211725264788,
          "validation_loss": 0.0022406503558158875,
          "train_metric": 0.000407620333135128,
          "validation_metric": 0.0002256701118312776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3857458549991861
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.006638483784627169,
          "train_loss": 0.0031345903407782316,
          "validation_loss": 0.0033052018843591213,
          "train_metric": 0.0003157044993713498,
          "validation_metric": 0.00033288789563812315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39954999899964605
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.004400948935654014,
          "train_loss": 0.007335529197007418,
          "validation_loss": 0.007428395561873913,
          "train_metric": 0.0007388076628558338,
          "validation_metric": 0.0007481608190573752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4129146889990807
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.007314472968573682,
          "train_loss": 0.01119297370314598,
          "validation_loss": 0.009809521026909351,
          "train_metric": 0.001127315335907042,
          "validation_metric": 0.000987979001365602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4262829919989599
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.011946768732741475,
          "train_loss": 0.015263335779309273,
          "validation_loss": 0.009868407621979713,
          "train_metric": 0.00153726723510772,
          "validation_metric": 0.0009939097799360752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4401130619989999
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01918425314943306,
          "train_loss": 0.008226007223129272,
          "validation_loss": 0.006636414676904678,
          "train_metric": 0.0008284933865070343,
          "validation_metric": 0.0006683953106403351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4536601250001695
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.007922696531750262,
          "train_loss": 0.00950158853083849,
          "validation_loss": 0.00873914361000061,
          "train_metric": 0.0009569652029313147,
          "validation_metric": 0.0008801745716482401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4678145849993598
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.0060403976967791095,
          "train_loss": 0.004429431166499853,
          "validation_loss": 0.0031370145734399557,
          "train_metric": 0.00044611605699174106,
          "validation_metric": 0.0003159486223012209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48139092499877734
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.007136315063689835,
          "train_loss": 0.00483589805662632,
          "validation_loss": 0.004416401963680983,
          "train_metric": 0.00048705397057347,
          "validation_metric": 0.00044480382348410785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49499998799910827
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.006166164930618834,
          "train_loss": 0.008886346593499184,
          "validation_loss": 0.006666672416031361,
          "train_metric": 0.0008950002957135439,
          "validation_metric": 0.0006714428309351206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5082901129990205
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013328372235264396,
          "train_loss": 0.006340267136693001,
          "validation_loss": 0.005754882004112005,
          "train_metric": 0.0006385684828273952,
          "validation_metric": 0.0005796106997877359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5218518559995573
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.012484577498980798,
          "train_loss": 0.010533424094319344,
          "validation_loss": 0.008602084591984749,
          "train_metric": 0.0010608879383653402,
          "validation_metric": 0.0008663704502396286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5381561679987499
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.008715812757145613,
          "train_loss": 0.010210011154413223,
          "validation_loss": 0.010367657989263535,
          "train_metric": 0.0010283149313181639,
          "validation_metric": 0.0010441925842314959,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5521661049988325
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.006645105022471398,
          "train_loss": 0.003892696462571621,
          "validation_loss": 0.004213661886751652,
          "train_metric": 0.00039205807843245566,
          "validation_metric": 0.00042438460513949394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5658646779993433
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.0034325837914366275,
          "train_loss": 0.005157135892659426,
          "validation_loss": 0.004753299057483673,
          "train_metric": 0.0005194079130887985,
          "validation_metric": 0.0004787349025718868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5790487639987987
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.0053725406469311565,
          "train_loss": 0.005116073414683342,
          "validation_loss": 0.00443333899602294,
          "train_metric": 0.0005152720841579139,
          "validation_metric": 0.0004465096862986684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5941187359985634
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.004826908516406547,
          "train_loss": 0.0030361060053110123,
          "validation_loss": 0.002795793814584613,
          "train_metric": 0.0003057854773942381,
          "validation_metric": 0.0002815821208059788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6078358109989495
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.005764299785369076,
          "train_loss": 0.002777745481580496,
          "validation_loss": 0.002067248336970806,
          "train_metric": 0.00027976438286714256,
          "validation_metric": 0.00020820568897761405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.625724874998923
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.00684713403461501,
          "train_loss": 0.016613522544503212,
          "validation_loss": 0.0151804955676198,
          "train_metric": 0.001673253020271659,
          "validation_metric": 0.0015289239818230271,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6402422159990238
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.013230672280769795,
          "train_loss": 0.011494782753288746,
          "validation_loss": 0.008084407076239586,
          "train_metric": 0.0011577124241739511,
          "validation_metric": 0.0008142318110913038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6611805969987472
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.009761531080584973,
          "train_loss": 0.01654219627380371,
          "validation_loss": 0.01096273586153984,
          "train_metric": 0.001666069496423006,
          "validation_metric": 0.0011041266843676567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6745940519995202
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.007402211602311581,
          "train_loss": 0.004573901183903217,
          "validation_loss": 0.0048045748844742775,
          "train_metric": 0.00046066660434007645,
          "validation_metric": 0.00048389926087111235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6878367779991095
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.006362319247273263,
          "train_loss": 0.01003568060696125,
          "validation_loss": 0.010160641744732857,
          "train_metric": 0.001010756939649582,
          "validation_metric": 0.0010233427165076137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.701137418000144
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009506951872026548,
          "train_loss": 0.003585672238841653,
          "validation_loss": 0.0027831955812871456,
          "train_metric": 0.0003611357242334634,
          "validation_metric": 0.00028031328110955656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7143411170000036
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.0032224172391579486,
          "train_loss": 0.0028891030233353376,
          "validation_loss": 0.0022915666922926903,
          "train_metric": 0.00029097992228344083,
          "validation_metric": 0.00023079822130966932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7272994989998551
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0029776674891763832,
          "train_loss": 0.0018656619358807802,
          "validation_loss": 0.0016751085640862584,
          "train_metric": 0.0001879026385722682,
          "validation_metric": 0.00016871080151759088,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7421917779993237
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.0025117844124906696,
          "train_loss": 0.0015789213357493281,
          "validation_loss": 0.001564625184983015,
          "train_metric": 0.00015902318409644067,
          "validation_metric": 0.00015758331574033946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7576592589994107
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.004702272948634345,
          "train_loss": 0.0031877593137323856,
          "validation_loss": 0.00207777158357203,
          "train_metric": 0.0003210594004485756,
          "validation_metric": 0.00020926553406752646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7791523639989464
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.0043695231070159934,
          "train_loss": 0.002947391476482153,
          "validation_loss": 0.0018236770993098617,
          "train_metric": 0.0002968504559248686,
          "validation_metric": 0.0001836740702856332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8011166819997015
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0030838565071462654,
          "train_loss": 0.0019234869396314025,
          "validation_loss": 0.0018800768302753568,
          "train_metric": 0.0001937265769811347,
          "validation_metric": 0.0001893544540507719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8281827880000492
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.00831347810162697,
          "train_loss": 0.010747790336608887,
          "validation_loss": 0.011046103201806545,
          "train_metric": 0.00108247809112072,
          "validation_metric": 0.0011125231394544244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8457500929998787
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.013217429805081338,
          "train_loss": 0.014079894870519638,
          "validation_loss": 0.011847943998873234,
          "train_metric": 0.0014180755242705345,
          "validation_metric": 0.0011932814959436655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8596655820001615
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.00968716957140714,
          "train_loss": 0.003559175645932555,
          "validation_loss": 0.002408534288406372,
          "train_metric": 0.0003584671358112246,
          "validation_metric": 0.00024257875338662416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.873152410998955
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.004268310447514523,
          "train_loss": 0.003076690249145031,
          "validation_loss": 0.0026662994641810656,
          "train_metric": 0.0003098729357589036,
          "validation_metric": 0.00026853993767872453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8865011489997414
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.003192644871887751,
          "train_loss": 0.002256835578009486,
          "validation_loss": 0.001771004986949265,
          "train_metric": 0.00022730021737515926,
          "validation_metric": 0.00017836914048530161,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9004199299997708
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.010353232479246799,
          "train_loss": 0.006857823580503464,
          "validation_loss": 0.007976052351295948,
          "train_metric": 0.0006906948983669281,
          "validation_metric": 0.0008033186313696206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9139798579999479
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.00696221791440621,
          "train_loss": 0.004263373091816902,
          "validation_loss": 0.003222549334168434,
          "train_metric": 0.00042939133709296584,
          "validation_metric": 0.000324563414324075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9275757879986486
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.004308240255340934,
          "train_loss": 0.0050133042968809605,
          "validation_loss": 0.003281049197539687,
          "train_metric": 0.0005049216561019421,
          "validation_metric": 0.00033045525196939707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9411536469997372
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.005824062769534066,
          "train_loss": 0.0041847084648907185,
          "validation_loss": 0.004596705548465252,
          "train_metric": 0.00042146845953539014,
          "validation_metric": 0.0004629633331205696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.954542624998794
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0044172290654387325,
          "train_loss": 0.0028241947293281555,
          "validation_loss": 0.0023591758217662573,
          "train_metric": 0.00028444253257475793,
          "validation_metric": 0.00023760757176205516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9684722290003265
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007902841818577144,
          "train_loss": 0.008420509286224842,
          "validation_loss": 0.00681733526289463,
          "train_metric": 0.0008480828255414963,
          "validation_metric": 0.0006866169860586524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9817232879995572
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.006470535372500308,
          "train_loss": 0.006183000281453133,
          "validation_loss": 0.004538337234407663,
          "train_metric": 0.0006227291305549443,
          "validation_metric": 0.00045708479592576623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9948719009989873
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0047544184199068695,
          "train_loss": 0.0038780677132308483,
          "validation_loss": 0.0040322886779904366,
          "train_metric": 0.00039058475522324443,
          "validation_metric": 0.0004061173531226814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0082616019990382
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0037328602775232866,
          "train_loss": 0.004080135840922594,
          "validation_loss": 0.002278213854879141,
          "train_metric": 0.00041093630716204643,
          "validation_metric": 0.0002294533624080941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.021323618999304
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0020912929430778604,
          "train_loss": 0.0014375519240275025,
          "validation_loss": 0.0014171263901516795,
          "train_metric": 0.00014478497905656695,
          "validation_metric": 0.00014272777480073273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0357893599993986
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.001979758586458047,
          "train_loss": 0.002612967509776354,
          "validation_loss": 0.002011799719184637,
          "train_metric": 0.00026316847652196884,
          "validation_metric": 0.0002026211004704237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0564940240001306
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0026347338080086047,
          "train_loss": 0.005731229670345783,
          "validation_loss": 0.005995870567858219,
          "train_metric": 0.0005772284930571914,
          "validation_metric": 0.0006038821302354336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0879557659991406
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.003633381023973925,
          "train_loss": 0.015063708648085594,
          "validation_loss": 0.014628863893449306,
          "train_metric": 0.0015171617269515991,
          "validation_metric": 0.0014733655843883753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1149642679993121
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.009834867814788595,
          "train_loss": 0.004153165500611067,
          "validation_loss": 0.003681073896586895,
          "train_metric": 0.0004182916018180549,
          "validation_metric": 0.00037074426654726267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.131668632999208
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.003664548541564727,
          "train_loss": 0.0019357223063707352,
          "validation_loss": 0.0015469479840248823,
          "train_metric": 0.00019495884771458805,
          "validation_metric": 0.0001558029471198097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1463962079997145
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.010253546482999809,
          "train_loss": 0.010063789784908295,
          "validation_loss": 0.008222606964409351,
          "train_metric": 0.001013588160276413,
          "validation_metric": 0.0008281508926302195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1615536459994473
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.007564818821265362,
          "train_loss": 0.006584870629012585,
          "validation_loss": 0.006203601602464914,
          "train_metric": 0.0006632040021941066,
          "validation_metric": 0.0006248039426282048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1757749150001473
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.004099438592675142,
          "train_loss": 0.002144117606803775,
          "validation_loss": 0.0014944118447601795,
          "train_metric": 0.0002159476571250707,
          "validation_metric": 0.00015051168156787753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.190206015999138
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0032105334757943638,
          "train_loss": 0.0017245177878066897,
          "validation_loss": 0.00147560506593436,
          "train_metric": 0.0001736871199682355,
          "validation_metric": 0.00014861753152217716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2053838899992115
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.002015630409005098,
          "train_loss": 0.002402056474238634,
          "validation_loss": 0.0018092561513185501,
          "train_metric": 0.00024192631826736033,
          "validation_metric": 0.00018222165817860514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2205059150001034
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0018255817849421874,
          "train_loss": 0.002874446101486683,
          "validation_loss": 0.002978114178404212,
          "train_metric": 0.00028950365958735347,
          "validation_metric": 0.0002999447751790285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2354797329990106
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.00219643452874152,
          "train_loss": 0.0022699374239891768,
          "validation_loss": 0.001636217231862247,
          "train_metric": 0.00022861978504806757,
          "validation_metric": 0.0001647938188398257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2501140829990618
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0024114554944389965,
          "train_loss": 0.0011883843690156937,
          "validation_loss": 0.0010088231647387147,
          "train_metric": 0.0001196897283080034,
          "validation_metric": 0.00010160497186006978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2654655659989658
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005671732622431591,
          "train_loss": 0.0026004889514297247,
          "validation_loss": 0.0026728517841547728,
          "train_metric": 0.0002619117440190166,
          "validation_metric": 0.00026919986703433096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2802533799986122
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.002263538113766117,
          "train_loss": 0.0016506339889019728,
          "validation_loss": 0.0011818850180134177,
          "train_metric": 0.00016624582349322736,
          "validation_metric": 0.00011903511767741293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2949415690000023
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0025546831348037813,
          "train_loss": 0.0010019607143476605,
          "validation_loss": 0.000665078463498503,
          "train_metric": 0.00010091381409438327,
          "validation_metric": 6.698427023366094e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3101966209997045
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0013269855953694787,
          "train_loss": 0.0011451351456344128,
          "validation_loss": 0.0010535723995417356,
          "train_metric": 0.00011533381621120498,
          "validation_metric": 0.0001061119619407691,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.324667048000265
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.012651161105168285,
          "train_loss": 0.011599946767091751,
          "validation_loss": 0.007627279497683048,
          "train_metric": 0.0011683041229844093,
          "validation_metric": 0.000768191646784544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3387374880003335
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.009440422960324213,
          "train_loss": 0.0051046391017735004,
          "validation_loss": 0.0038668003398925066,
          "train_metric": 0.0005141206202097237,
          "validation_metric": 0.0003894499968737364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3527742059995944
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.007274967574630864,
          "train_loss": 0.007724083960056305,
          "validation_loss": 0.008014827035367489,
          "train_metric": 0.0007779414299875498,
          "validation_metric": 0.0008072240743786097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3670420799990097
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.01032038229459431,
          "train_loss": 0.00949662085622549,
          "validation_loss": 0.008266611024737358,
          "train_metric": 0.0009564649080857635,
          "validation_metric": 0.0008325828821398318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3804717709990655
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.009481584202148952,
          "train_loss": 0.005834406707435846,
          "validation_loss": 0.005998155102133751,
          "train_metric": 0.0005876200739294291,
          "validation_metric": 0.0006041122251190245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3943045569994865
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0069979504332877696,
          "train_loss": 0.004951157607138157,
          "validation_loss": 0.003516714321449399,
          "train_metric": 0.000498662528116256,
          "validation_metric": 0.00035419058986008167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4081206060000113
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004454804940905888,
          "train_loss": 0.0033845941070467234,
          "validation_loss": 0.0028575609903782606,
          "train_metric": 0.00034088391112163663,
          "validation_metric": 0.0002878030645661056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4217698869997548
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004547477743471973,
          "train_loss": 0.004957916215062141,
          "validation_loss": 0.002591877244412899,
          "train_metric": 0.0004993431502953172,
          "validation_metric": 0.0002610443625599146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4354575369998201
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.002780291160888737,
          "train_loss": 0.0016864355420693755,
          "validation_loss": 0.0012292545288801193,
          "train_metric": 0.00016985161346383393,
          "validation_metric": 0.00012380599218886346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.44935355799862
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.005378440884669544,
          "train_loss": 0.01067050639539957,
          "validation_loss": 0.009048969484865665,
          "train_metric": 0.0010746942134574056,
          "validation_metric": 0.0009113791747950017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4630152529989573
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.00727358645235654,
          "train_loss": 0.0034815578255802393,
          "validation_loss": 0.0025574821047484875,
          "train_metric": 0.00035064981784671545,
          "validation_metric": 0.0002575802500359714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4767850520001957
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0032759085988800507,
          "train_loss": 0.0029653972014784813,
          "validation_loss": 0.0021796049550175667,
          "train_metric": 0.0002986639447044581,
          "validation_metric": 0.00021952184033580124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4903520179996121
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0022715423729096074,
          "train_loss": 0.0015629044501110911,
          "validation_loss": 0.0010809411760419607,
          "train_metric": 0.00015741001698188484,
          "validation_metric": 0.00010886842937907204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5041128670000035
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0015194572515611071,
          "train_loss": 0.0013308447087183595,
          "validation_loss": 0.0009314727503806353,
          "train_metric": 0.0001340378075838089,
          "validation_metric": 9.381453128298745e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5180737359987688
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0022806995366408955,
          "train_loss": 0.0025346858892589808,
          "validation_loss": 0.0024042658042162657,
          "train_metric": 0.0002552842488512397,
          "validation_metric": 0.00024214886070694774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5313551839990396
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.01612027273949934,
          "train_loss": 0.014160738326609135,
          "validation_loss": 0.011336752213537693,
          "train_metric": 0.0014262177282944322,
          "validation_metric": 0.0011417961213737726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5446664599985525
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.012640742410439998,
          "train_loss": 0.009316591545939445,
          "validation_loss": 0.008496502414345741,
          "train_metric": 0.0009383329306729138,
          "validation_metric": 0.000855736609082669,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.558039348999955
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.006215162909938954,
          "train_loss": 0.00460429023951292,
          "validation_loss": 0.004066161345690489,
          "train_metric": 0.0004637272213585675,
          "validation_metric": 0.0004095288459211588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5718147959996713
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0031090339543879963,
          "train_loss": 0.009074498899281025,
          "validation_loss": 0.006904708221554756,
          "train_metric": 0.0009139501489698887,
          "validation_metric": 0.0006954169366508722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5854056409989425
        }
      ],
      "best": {
        "epoch": 87,
        "updates": 1392
      },
      "best_validation_metric": 6.698427023366094e-05,
      "config": {
        "label": "LayerNorm",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.008184391867063852
    },
    "32d8a0bb9dc3_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.725183792412281,
          "train_loss": 0.271520733833313,
          "validation_loss": 0.22441114485263824,
          "train_metric": 0.027346575632691383,
          "validation_metric": 0.022601868957281113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0441367720013659
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.19251352921128273,
          "train_loss": 0.13974319398403168,
          "validation_loss": 0.1179056391119957,
          "train_metric": 0.014074422419071198,
          "validation_metric": 0.011875025928020477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07219726700168394
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12273477180860937,
          "train_loss": 0.10032490640878677,
          "validation_loss": 0.08398673683404922,
          "train_metric": 0.010104356333613396,
          "validation_metric": 0.008458836935460567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10534372400070424
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09152746060863137,
          "train_loss": 0.08048801124095917,
          "validation_loss": 0.06612011045217514,
          "train_metric": 0.008106458932161331,
          "validation_metric": 0.006659375037997961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13163933900068514
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07133611291646957,
          "train_loss": 0.06516025960445404,
          "validation_loss": 0.05265787988901138,
          "train_metric": 0.0065627028234303,
          "validation_metric": 0.005303508602082729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15786437600036152
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06069809605833143,
          "train_loss": 0.05711313337087631,
          "validation_loss": 0.047087281942367554,
          "train_metric": 0.005752225872129202,
          "validation_metric": 0.004742457997053862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18846230700000888
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05503121088258922,
          "train_loss": 0.05024784803390503,
          "validation_loss": 0.041228801012039185,
          "train_metric": 0.005060779396444559,
          "validation_metric": 0.004152413923293352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2178561640012049
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04905405070167035,
          "train_loss": 0.04574151337146759,
          "validation_loss": 0.03793175891041756,
          "train_metric": 0.004606917500495911,
          "validation_metric": 0.0038203480653464794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25148212400017655
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04519221524242312,
          "train_loss": 0.05024033412337303,
          "validation_loss": 0.04192633181810379,
          "train_metric": 0.005060022231191397,
          "validation_metric": 0.004222666844725609,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27935675900153
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.049359321128576994,
          "train_loss": 0.046037573367357254,
          "validation_loss": 0.039142146706581116,
          "train_metric": 0.004636736121028662,
          "validation_metric": 0.003942254465073347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30935597400093684
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04093247506534681,
          "train_loss": 0.03829994797706604,
          "validation_loss": 0.03261260688304901,
          "train_metric": 0.0038574300706386566,
          "validation_metric": 0.003284622449427843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.332005416001266
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.036824317183345556,
          "train_loss": 0.032820843160152435,
          "validation_loss": 0.028341077268123627,
          "train_metric": 0.003305594902485609,
          "validation_metric": 0.002854409161955118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36297993100015447
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03441062942147255,
          "train_loss": 0.03700372204184532,
          "validation_loss": 0.02941253036260605,
          "train_metric": 0.0037268793676048517,
          "validation_metric": 0.0029623224399983883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4038193810010853
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.032163326046429574,
          "train_loss": 0.02908608876168728,
          "validation_loss": 0.024541635066270828,
          "train_metric": 0.0029294448904693127,
          "validation_metric": 0.0024717431515455246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.450933004000035
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03053868602728471,
          "train_loss": 0.02759496122598648,
          "validation_loss": 0.023318752646446228,
          "train_metric": 0.0027792633045464754,
          "validation_metric": 0.0023485796991735697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4946302630014543
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.026482005312573165,
          "train_loss": 0.027983058243989944,
          "validation_loss": 0.023301327601075172,
          "train_metric": 0.002818351611495018,
          "validation_metric": 0.002346824621781707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5416870870012644
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.02441761939553544,
          "train_loss": 0.023983746767044067,
          "validation_loss": 0.02047683112323284,
          "train_metric": 0.002415555063635111,
          "validation_metric": 0.0020623516757041216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5845752150016779
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.023378114216029644,
          "train_loss": 0.02086450532078743,
          "validation_loss": 0.017464669421315193,
          "train_metric": 0.002101396443322301,
          "validation_metric": 0.0017589774215593934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6309901370004809
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.02140462800161913,
          "train_loss": 0.02289021760225296,
          "validation_loss": 0.018739325925707817,
          "train_metric": 0.002305419184267521,
          "validation_metric": 0.001887356280349195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6736752290016739
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02563406975241378,
          "train_loss": 0.02189042419195175,
          "validation_loss": 0.017806872725486755,
          "train_metric": 0.0022047236561775208,
          "validation_metric": 0.001793443108908832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7152248640013568
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023129807494115084,
          "train_loss": 0.020913323387503624,
          "validation_loss": 0.017478393390774727,
          "train_metric": 0.0021063131280243397,
          "validation_metric": 0.00176035996992141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7321397210016585
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02696609584381804,
          "train_loss": 0.026205450296401978,
          "validation_loss": 0.021345825865864754,
          "train_metric": 0.0026393167208880186,
          "validation_metric": 0.002149873413145542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7487391009999556
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.019721148069947958,
          "train_loss": 0.01606700010597706,
          "validation_loss": 0.01406087540090084,
          "train_metric": 0.0016182095278054476,
          "validation_metric": 0.0014161599101498723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7663568310017581
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.017273633362492546,
          "train_loss": 0.014825965277850628,
          "validation_loss": 0.012481079436838627,
          "train_metric": 0.0014932170743122697,
          "validation_metric": 0.0012570485705509782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7843270920002396
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019757675006985664,
          "train_loss": 0.01432263944298029,
          "validation_loss": 0.012306571938097477,
          "train_metric": 0.001442523906007409,
          "validation_metric": 0.0012394727673381567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8019784510015597
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.017510230012703687,
          "train_loss": 0.014158128760755062,
          "validation_loss": 0.01238947082310915,
          "train_metric": 0.0014259549789130688,
          "validation_metric": 0.0012478221906349063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8189519220013608
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015726784709841013,
          "train_loss": 0.013463691808283329,
          "validation_loss": 0.01134172547608614,
          "train_metric": 0.0013560139341279864,
          "validation_metric": 0.001142297056503594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8361309360006999
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.013707112200791016,
          "train_loss": 0.012812243774533272,
          "validation_loss": 0.01123434491455555,
          "train_metric": 0.001290402258746326,
          "validation_metric": 0.0011314821895211935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8542370070008474
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016051386279286817,
          "train_loss": 0.02502411976456642,
          "validation_loss": 0.020168717950582504,
          "train_metric": 0.0025203379336744547,
          "validation_metric": 0.002031319309026003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8744756100004452
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01542237150715664,
          "train_loss": 0.014315691776573658,
          "validation_loss": 0.011095796711742878,
          "train_metric": 0.001441823784261942,
          "validation_metric": 0.001117528066970408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9117582290000428
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.015414124500239268,
          "train_loss": 0.021503446623682976,
          "validation_loss": 0.018202824518084526,
          "train_metric": 0.0021657487377524376,
          "validation_metric": 0.0018333217594772577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9285116270002618
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.020806275395443663,
          "train_loss": 0.017160432413220406,
          "validation_loss": 0.012870152480900288,
          "train_metric": 0.0017283360939472914,
          "validation_metric": 0.0012962345499545336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9447653310016904
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01630209616268985,
          "train_loss": 0.011724606156349182,
          "validation_loss": 0.009741163812577724,
          "train_metric": 0.0011808593990281224,
          "validation_metric": 0.0009810944320634007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9628303190002043
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013805997223244049,
          "train_loss": 0.023595713078975677,
          "validation_loss": 0.019702864810824394,
          "train_metric": 0.0023764735087752342,
          "validation_metric": 0.0019844004418700933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9798160860009375
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.015536964187049307,
          "train_loss": 0.011063741520047188,
          "validation_loss": 0.00977888610213995,
          "train_metric": 0.0011142995208501816,
          "validation_metric": 0.000984893529675901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9960877440007607
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.012316050342633389,
          "train_loss": 0.011551490053534508,
          "validation_loss": 0.0093732550740242,
          "train_metric": 0.0011634236434474587,
          "validation_metric": 0.0009440399007871747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.013489811000909
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.01322234922554344,
          "train_loss": 0.01286236196756363,
          "validation_loss": 0.010114856995642185,
          "train_metric": 0.0012954497942700982,
          "validation_metric": 0.001018731389194727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.030029697001737
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011155385203892365,
          "train_loss": 0.011489743366837502,
          "validation_loss": 0.010611145757138729,
          "train_metric": 0.0011572049697861075,
          "validation_metric": 0.0010687157046049833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0468310190008197
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.011601212609093636,
          "train_loss": 0.014726401306688786,
          "validation_loss": 0.010290934704244137,
          "train_metric": 0.0014831890584900975,
          "validation_metric": 0.0010364654008299112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0635103390013683
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012412100812071003,
          "train_loss": 0.012658714316785336,
          "validation_loss": 0.01127684861421585,
          "train_metric": 0.0012749392772093415,
          "validation_metric": 0.00113576278090477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.080123261001063
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011675719724735245,
          "train_loss": 0.009901653975248337,
          "validation_loss": 0.007692788261920214,
          "train_metric": 0.0009972583502531052,
          "validation_metric": 0.0007747894851490855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0975040620014624
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010013210485340096,
          "train_loss": 0.008797609247267246,
          "validation_loss": 0.007866949774324894,
          "train_metric": 0.0008860630914568901,
          "validation_metric": 0.0007923304219730198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1146381820017268
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009386818463099189,
          "train_loss": 0.013002975843846798,
          "validation_loss": 0.009492841549217701,
          "train_metric": 0.001309612300246954,
          "validation_metric": 0.0009560841717757285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1313603990001866
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01224605915194843,
          "train_loss": 0.009541220031678677,
          "validation_loss": 0.0073880041018128395,
          "train_metric": 0.0009609567932784557,
          "validation_metric": 0.0007440927438437939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1496499780005252
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.01069485695916228,
          "train_loss": 0.01067455392330885,
          "validation_loss": 0.009969848208129406,
          "train_metric": 0.0010751018999144435,
          "validation_metric": 0.0010041267378255725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1676477480014
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010661405161954463,
          "train_loss": 0.009650606662034988,
          "validation_loss": 0.008344125933945179,
          "train_metric": 0.0009719737572595477,
          "validation_metric": 0.0008403898100368679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.184343061000618
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011100387433543801,
          "train_loss": 0.015353727154433727,
          "validation_loss": 0.010953432880342007,
          "train_metric": 0.0015463712625205517,
          "validation_metric": 0.0011031896574422717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2010217310016742
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010565859702182934,
          "train_loss": 0.008216824382543564,
          "validation_loss": 0.007050653453916311,
          "train_metric": 0.0008275684667751193,
          "validation_metric": 0.0007101160590536892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2182181850002962
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00831423222552985,
          "train_loss": 0.007751990109682083,
          "validation_loss": 0.007047897204756737,
          "train_metric": 0.0007807521033100784,
          "validation_metric": 0.000709838408511132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2363135420000617
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.008484421006869525,
          "train_loss": 0.008210123516619205,
          "validation_loss": 0.006662710104137659,
          "train_metric": 0.0008268934907391667,
          "validation_metric": 0.0006710437010042369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2547312560000137
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009121570168645121,
          "train_loss": 0.008138127624988556,
          "validation_loss": 0.006662833970040083,
          "train_metric": 0.0008196425042115152,
          "validation_metric": 0.0006710562738589942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.271314589001122
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.012653169003897347,
          "train_loss": 0.008039749227464199,
          "validation_loss": 0.006858522072434425,
          "train_metric": 0.0008097341051325202,
          "validation_metric": 0.0006907652714289725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2881438870008424
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008644956105854362,
          "train_loss": 0.008325216360390186,
          "validation_loss": 0.00712231732904911,
          "train_metric": 0.0008384853135794401,
          "validation_metric": 0.0007173336343839765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3049930580000364
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.00779974230681546,
          "train_loss": 0.008252358995378017,
          "validation_loss": 0.007306789513677359,
          "train_metric": 0.0008311473648063838,
          "validation_metric": 0.0007359131122939289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.322019661000013
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.00925846655445639,
          "train_loss": 0.00652270158752799,
          "validation_loss": 0.0057301875203847885,
          "train_metric": 0.0006569424876943231,
          "validation_metric": 0.0005771234282292426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3400438050011871
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.012231946529936977,
          "train_loss": 0.00858000572770834,
          "validation_loss": 0.00622341874986887,
          "train_metric": 0.0008641468011774123,
          "validation_metric": 0.0006267998833209276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.357745753000927
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008446793217444792,
          "train_loss": 0.00689212791621685,
          "validation_loss": 0.006349670700728893,
          "train_metric": 0.0006941498722881079,
          "validation_metric": 0.0006395156960934401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3750606240009802
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.008316159102832898,
          "train_loss": 0.007118558511137962,
          "validation_loss": 0.006148797459900379,
          "train_metric": 0.0007169551681727171,
          "validation_metric": 0.0006192844011820853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.417699673000243
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007200873791589402,
          "train_loss": 0.006566303316503763,
          "validation_loss": 0.005335439462214708,
          "train_metric": 0.0006613340810872614,
          "validation_metric": 0.0005373659078031778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4610598870003741
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006584036775166169,
          "train_loss": 0.006239462643861771,
          "validation_loss": 0.005712360609322786,
          "train_metric": 0.0006284158444032073,
          "validation_metric": 0.0005753280129283667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.478259337000054
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.006660603481577709,
          "train_loss": 0.006314736790955067,
          "validation_loss": 0.005067717283964157,
          "train_metric": 0.0006359972176142037,
          "validation_metric": 0.0005104019655846059,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4960539100011374
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.006349309740471654,
          "train_loss": 0.005507147405296564,
          "validation_loss": 0.004698514938354492,
          "train_metric": 0.0005546596948988736,
          "validation_metric": 0.00047321722377091646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5139210150009603
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.008569663259549998,
          "train_loss": 0.015207894146442413,
          "validation_loss": 0.0099509721621871,
          "train_metric": 0.001531683374196291,
          "validation_metric": 0.0010022253263741732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5307424510010605
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.010702520652557723,
          "train_loss": 0.0073378500528633595,
          "validation_loss": 0.007015361916273832,
          "train_metric": 0.0007390414248220623,
          "validation_metric": 0.0007065616082400084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5471635640005843
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.007163544389186427,
          "train_loss": 0.0053948210552334785,
          "validation_loss": 0.0044744606129825115,
          "train_metric": 0.0005433466285467148,
          "validation_metric": 0.00045065133599564433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5644967920015915
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005669868871336803,
          "train_loss": 0.005300572142004967,
          "validation_loss": 0.0045133717358112335,
          "train_metric": 0.0005338541814126074,
          "validation_metric": 0.00045457022497430444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.58137309400081
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0069226891282596625,
          "train_loss": 0.007420714013278484,
          "validation_loss": 0.006919790990650654,
          "train_metric": 0.0007473871810361743,
          "validation_metric": 0.000696935981977731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5975503730005585
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00731872973847203,
          "train_loss": 0.006452545523643494,
          "validation_loss": 0.0051038870587944984,
          "train_metric": 0.00064987683435902,
          "validation_metric": 0.0005140448338352144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.613870933000726
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.009972294181352481,
          "train_loss": 0.0056060985662043095,
          "validation_loss": 0.004278527572751045,
          "train_metric": 0.0005646257195621729,
          "validation_metric": 0.00043091760016977787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6368519660009042
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.007715228144661523,
          "train_loss": 0.00778129231184721,
          "validation_loss": 0.007210019044578075,
          "train_metric": 0.0007837032317183912,
          "validation_metric": 0.0007261668215505779,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6619036820011388
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006288037599006202,
          "train_loss": 0.005706281866878271,
          "validation_loss": 0.00420799758285284,
          "train_metric": 0.0005747157265432179,
          "validation_metric": 0.00042381405364722013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.679613419000816
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00576221395022003,
          "train_loss": 0.005137464962899685,
          "validation_loss": 0.004070205148309469,
          "train_metric": 0.000517426582518965,
          "validation_metric": 0.0004099361540284008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6977159730013227
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.005247840948868543,
          "train_loss": 0.006295354105532169,
          "validation_loss": 0.005427327938377857,
          "train_metric": 0.0006340449908748269,
          "validation_metric": 0.0005466206348501146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7148870030014223
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005847955937497318,
          "train_loss": 0.0056092459708452225,
          "validation_loss": 0.004303757566958666,
          "train_metric": 0.0005649426602758467,
          "validation_metric": 0.0004334586556069553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.732290958001613
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00761535503261257,
          "train_loss": 0.0046490756794810295,
          "validation_loss": 0.004038749262690544,
          "train_metric": 0.00046823782031424344,
          "validation_metric": 0.00040676796925254166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7498071630016057
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0055943846455193125,
          "train_loss": 0.005045777186751366,
          "validation_loss": 0.003924061544239521,
          "train_metric": 0.000508192228153348,
          "validation_metric": 0.0003952171537093818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7674918660013645
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.00631263930699788,
          "train_loss": 0.005989368073642254,
          "validation_loss": 0.005381503142416477,
          "train_metric": 0.0006032272358424962,
          "validation_metric": 0.0005420053494162858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7848634120000497
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.005129424927872606,
          "train_loss": 0.004593792837113142,
          "validation_loss": 0.0038168870378285646,
          "train_metric": 0.00046266999561339617,
          "validation_metric": 0.0003844228631351143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8029463070015481
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00618596728600096,
          "train_loss": 0.004681388847529888,
          "validation_loss": 0.004123073071241379,
          "train_metric": 0.0004714923270512372,
          "validation_metric": 0.00041526081622578204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8211262230015564
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004919681974570267,
          "train_loss": 0.004786643199622631,
          "validation_loss": 0.0038449480198323727,
          "train_metric": 0.00048209322267211974,
          "validation_metric": 0.00038724910700693727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.863371529001597
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.004694133349403273,
          "train_loss": 0.004141520708799362,
          "validation_loss": 0.0031629661098122597,
          "train_metric": 0.0004171188338659704,
          "validation_metric": 0.00031856237910687923,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.8813086900008784
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.004251900412782561,
          "train_loss": 0.004105106461793184,
          "validation_loss": 0.003548264503479004,
          "train_metric": 0.00041345125646330416,
          "validation_metric": 0.00035736820427700877,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.898307058001592
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.004312076867790893,
          "train_loss": 0.0044862693175673485,
          "validation_loss": 0.003908704500645399,
          "train_metric": 0.00045184060581959784,
          "validation_metric": 0.00039367034332826734,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9150687220007967
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0041945908378693275,
          "train_loss": 0.0037214390467852354,
          "validation_loss": 0.0031034399289637804,
          "train_metric": 0.0003748097224161029,
          "validation_metric": 0.00031256707734428346,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.933854261000306
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.004040578634885605,
          "train_loss": 0.004069951828569174,
          "validation_loss": 0.0036799642257392406,
          "train_metric": 0.00040991068817675114,
          "validation_metric": 0.0003706325078383088,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.950740140000562
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.0040998850599862635,
          "train_loss": 0.003756114048883319,
          "validation_loss": 0.0032717068679630756,
          "train_metric": 0.00037830209475941956,
          "validation_metric": 0.00032951432513073087,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0668679059999704
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.003695391867950093,
          "train_loss": 0.003698589513078332,
          "validation_loss": 0.0030183366034179926,
          "train_metric": 0.0003725083952303976,
          "validation_metric": 0.00030399582465179265,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.085234368001693
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0038144910140545107,
          "train_loss": 0.003679414512589574,
          "validation_loss": 0.0030063677113503218,
          "train_metric": 0.00037057712324894965,
          "validation_metric": 0.00030279034399427474,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.103437447000033
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004059057060658233,
          "train_loss": 0.004055152181535959,
          "validation_loss": 0.0035193392541259527,
          "train_metric": 0.0004084200190845877,
          "validation_metric": 0.0003544549399521202,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.121198909000668
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.003807932043855544,
          "train_loss": 0.0036410149186849594,
          "validation_loss": 0.0029214532114565372,
          "train_metric": 0.00036670968984253705,
          "validation_metric": 0.00029423809610307217,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1397713240003213
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.003686351432406809,
          "train_loss": 0.0037514667492359877,
          "validation_loss": 0.0030232840217649937,
          "train_metric": 0.00037783404695801437,
          "validation_metric": 0.00030449405312538147,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1574758300012036
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.003930527913325932,
          "train_loss": 0.005040637217462063,
          "validation_loss": 0.0035659174900501966,
          "train_metric": 0.0005076745292171836,
          "validation_metric": 0.0003591461281757802,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1748313390016847
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.005824663428938948,
          "train_loss": 0.0059798420406877995,
          "validation_loss": 0.004049479495733976,
          "train_metric": 0.0006022677989676595,
          "validation_metric": 0.0004078487982042134,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1918428570006654
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0051419840747257695,
          "train_loss": 0.00473429448902607,
          "validation_loss": 0.004228889476507902,
          "train_metric": 0.0004768207436427474,
          "validation_metric": 0.0004259182314854115,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2092609670016827
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004308916140871588,
          "train_loss": 0.0038505804259330034,
          "validation_loss": 0.003104234579950571,
          "train_metric": 0.00038781636976636946,
          "validation_metric": 0.0003126471128780395,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.226720906000992
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004323200191720389,
          "train_loss": 0.004481015261262655,
          "validation_loss": 0.003321533091366291,
          "train_metric": 0.0004513114399742335,
          "validation_metric": 0.0003345326113048941,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.243986640000003
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.003677783897728659,
          "train_loss": 0.0035569113679230213,
          "validation_loss": 0.00318676489405334,
          "train_metric": 0.0003582390781957656,
          "validation_metric": 0.00032095922506414354,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.261359308000465
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.003776038960495498,
          "train_loss": 0.00384995574131608,
          "validation_loss": 0.00298296264372766,
          "train_metric": 0.00038775341818109155,
          "validation_metric": 0.0003004330792464316,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2781771710015164
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.003750573203433305,
          "train_loss": 0.005627904087305069,
          "validation_loss": 0.004870121832937002,
          "train_metric": 0.0005668219528160989,
          "validation_metric": 0.0004905008245259523,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.2953029990003415
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0043092323176097125,
          "train_loss": 0.003388982266187668,
          "validation_loss": 0.002842513844370842,
          "train_metric": 0.00034132585278712213,
          "validation_metric": 0.000286287599010393,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.3131399460016837
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.004072784082381986,
          "train_loss": 0.0034068094100803137,
          "validation_loss": 0.0029256423003971577,
          "train_metric": 0.0003431214136071503,
          "validation_metric": 0.000294659985229373,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.330466282000998
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.004581090506690089,
          "train_loss": 0.007572703994810581,
          "validation_loss": 0.005017973482608795,
          "train_metric": 0.0007626950391568244,
          "validation_metric": 0.000505391857586801,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.3478337320011633
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.004626081841706764,
          "train_loss": 0.003329619998112321,
          "validation_loss": 0.0028826307971030474,
          "train_metric": 0.0003353471402078867,
          "validation_metric": 0.00029032799648121,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.365059146000931
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0035796783849946223,
          "train_loss": 0.004181520082056522,
          "validation_loss": 0.0030523089226335287,
          "train_metric": 0.00042114738607779145,
          "validation_metric": 0.0003074173873756081,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.3817236450013297
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003572121735487599,
          "train_loss": 0.0034728446044027805,
          "validation_loss": 0.003101638052612543,
          "train_metric": 0.00034977219183929265,
          "validation_metric": 0.0003123856149613857,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.3979555670011905
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0034893589472630993,
          "train_loss": 0.0033146669156849384,
          "validation_loss": 0.0026726454962044954,
          "train_metric": 0.00033384113339707255,
          "validation_metric": 0.00026917908689938486,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.415628796999954
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0032544165005674586,
          "train_loss": 0.005361037328839302,
          "validation_loss": 0.004820838570594788,
          "train_metric": 0.0005399439833126962,
          "validation_metric": 0.0004855372535530478,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.433048480001162
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.003910472441930324,
          "train_loss": 0.0032545693684369326,
          "validation_loss": 0.002825879491865635,
          "train_metric": 0.0003277883224654943,
          "validation_metric": 0.00028461223701015115,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.449783580001167
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.003456186612311285,
          "train_loss": 0.003233571071177721,
          "validation_loss": 0.002793659921735525,
          "train_metric": 0.00032567346352152526,
          "validation_metric": 0.00028136715991422534,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.46625280400076
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0036010880357935093,
          "train_loss": 0.0034271988552063704,
          "validation_loss": 0.0027039104606956244,
          "train_metric": 0.00034517492167651653,
          "validation_metric": 0.0002723280049394816,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.483298792001733
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.003300941491033882,
          "train_loss": 0.0030907911714166403,
          "validation_loss": 0.0025405895430594683,
          "train_metric": 0.00031129317358136177,
          "validation_metric": 0.00025587889831513166,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.5219435730014084
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.003468473551038187,
          "train_loss": 0.003211130853742361,
          "validation_loss": 0.002483213786035776,
          "train_metric": 0.0003234133473597467,
          "validation_metric": 0.00025010015815496445,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.5404950650008686
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0034491548212827183,
          "train_loss": 0.003110517282038927,
          "validation_loss": 0.0027305898256599903,
          "train_metric": 0.00031327991746366024,
          "validation_metric": 0.0002750149869825691,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.557994682001663
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.00341058707999764,
          "train_loss": 0.0030036356765776873,
          "validation_loss": 0.0026305029168725014,
          "train_metric": 0.0003025151672773063,
          "validation_metric": 0.0002649346424732357,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.5766433960016
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0036833696722169407,
          "train_loss": 0.00715977605432272,
          "validation_loss": 0.006156512536108494,
          "train_metric": 0.0007211064221337438,
          "validation_metric": 0.0006200614152476192,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.5936168170010205
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.004822219103516545,
          "train_loss": 0.005682012997567654,
          "validation_loss": 0.004764600656926632,
          "train_metric": 0.0005722715286538005,
          "validation_metric": 0.00047987315338104963,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.610186349000287
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003978816857852507,
          "train_loss": 0.0031570992432534695,
          "validation_loss": 0.0026085444260388613,
          "train_metric": 0.0003179714549332857,
          "validation_metric": 0.0002627230132929981,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.6276879840006586
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003126280462311115,
          "train_loss": 0.0029575303196907043,
          "validation_loss": 0.0025101618375629187,
          "train_metric": 0.000297871622024104,
          "validation_metric": 0.000252814352279529,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.645241980000719
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.003000586621055845,
          "train_loss": 0.0034779913257807493,
          "validation_loss": 0.002593284472823143,
          "train_metric": 0.00035029061837121844,
          "validation_metric": 0.00026118612731806934,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.662448472001415
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.003201823419658467,
          "train_loss": 0.0029774813447147608,
          "validation_loss": 0.002398156560957432,
          "train_metric": 0.0002998810086864978,
          "validation_metric": 0.0002415335620753467,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.6801195750012994
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004045840214530472,
          "train_loss": 0.004592662211507559,
          "validation_loss": 0.0032023629173636436,
          "train_metric": 0.0004625561705324799,
          "validation_metric": 0.000322530249832198,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.6974826420009776
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.003701841131260153,
          "train_loss": 0.0028661314863711596,
          "validation_loss": 0.0023970170877873898,
          "train_metric": 0.00028866625507362187,
          "validation_metric": 0.00024141877656802535,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.715680932000396
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0032009144488256425,
          "train_loss": 0.002971579786390066,
          "validation_loss": 0.002543577691540122,
          "train_metric": 0.00029928662115707994,
          "validation_metric": 0.0002561798319220543,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.7332934040005057
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0033831810069386847,
          "train_loss": 0.0028433888219296932,
          "validation_loss": 0.002360927639529109,
          "train_metric": 0.000286375725409016,
          "validation_metric": 0.00023778397007845342,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.750825534001706
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0034718084025371354,
          "train_loss": 0.00358563638292253,
          "validation_loss": 0.003119276836514473,
          "train_metric": 0.00036113214446231723,
          "validation_metric": 0.0003141621418762952,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.7677390160006325
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.00364026908937376,
          "train_loss": 0.0030280323699116707,
          "validation_loss": 0.002602670341730118,
          "train_metric": 0.00030497234547510743,
          "validation_metric": 0.0002621314488351345,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.7842397800013714
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0034630412337719463,
          "train_loss": 0.0029487404972314835,
          "validation_loss": 0.0023792257998138666,
          "train_metric": 0.00029698637081310153,
          "validation_metric": 0.0002396269264863804,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.8005794690016046
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0031155328470049426,
          "train_loss": 0.0028732072096318007,
          "validation_loss": 0.002532368525862694,
          "train_metric": 0.0002893789205700159,
          "validation_metric": 0.0002550508943386376,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.8172653740002715
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0031480775796808302,
          "train_loss": 0.0027690185233950615,
          "validation_loss": 0.002394314855337143,
          "train_metric": 0.00027888541808351874,
          "validation_metric": 0.0002411466557532549,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.8345855780007696
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.003161994754918851,
          "train_loss": 0.002765270182862878,
          "validation_loss": 0.0023818134795874357,
          "train_metric": 0.000278507883194834,
          "validation_metric": 0.00023988753673620522,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.852248118000716
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.002794301137328148,
          "train_loss": 0.0026434308383613825,
          "validation_loss": 0.0022947294637560844,
          "train_metric": 0.0002662366605363786,
          "validation_metric": 0.00023111674818210304,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.8703708900011407
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0033306965306110214,
          "train_loss": 0.002610274590551853,
          "validation_loss": 0.0021499297581613064,
          "train_metric": 0.000262897287029773,
          "validation_metric": 0.0002165330370189622,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.888293756001076
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.003099310375546338,
          "train_loss": 0.0026231734082102776,
          "validation_loss": 0.0022100061178207397,
          "train_metric": 0.00026419645291753113,
          "validation_metric": 0.0002225837088190019,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.905303923000247
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0029334629034565296,
          "train_loss": 0.0028142475057393312,
          "validation_loss": 0.002330898307263851,
          "train_metric": 0.00028344072052277625,
          "validation_metric": 0.00023475951456930488,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.922424658001546
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0029454486866598018,
          "train_loss": 0.0026651681400835514,
          "validation_loss": 0.0023774609435349703,
          "train_metric": 0.0002684259961824864,
          "validation_metric": 0.00023944914573803544,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.9395170530006
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.003623248519943445,
          "train_loss": 0.0032946672290563583,
          "validation_loss": 0.0030858528334647417,
          "train_metric": 0.000331826857291162,
          "validation_metric": 0.00031079576001502573,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.9566875740001706
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0033483493243693374,
          "train_loss": 0.003119024448096752,
          "validation_loss": 0.0023656764533370733,
          "train_metric": 0.000314136705128476,
          "validation_metric": 0.00023826223332434893,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.973780925000028
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0035234598399256356,
          "train_loss": 0.002566365757957101,
          "validation_loss": 0.0022103488445281982,
          "train_metric": 0.0002584749599918723,
          "validation_metric": 0.0002226182259619236,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.9905795770009718
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0033360855231876485,
          "train_loss": 0.002734483452513814,
          "validation_loss": 0.00236922362819314,
          "train_metric": 0.0002754071610979736,
          "validation_metric": 0.00023861952649895102,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.0075796880009875
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0035321497343829833,
          "train_loss": 0.0040888856165111065,
          "validation_loss": 0.0036982623860239983,
          "train_metric": 0.0004118176584597677,
          "validation_metric": 0.000372475478798151,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.0244559290003963
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.003944880336348433,
          "train_loss": 0.00267544062808156,
          "validation_loss": 0.002169861923903227,
          "train_metric": 0.0002694605791475624,
          "validation_metric": 0.00021854056103620678,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.041398689001653
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.002734667796175927,
          "train_loss": 0.002526236232370138,
          "validation_loss": 0.002251574071124196,
          "train_metric": 0.00025443322374485433,
          "validation_metric": 0.0002267703239340335,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.0581915950006078
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.002611969721328933,
          "train_loss": 0.0027287136763334274,
          "validation_loss": 0.0025383131578564644,
          "train_metric": 0.000274826044915244,
          "validation_metric": 0.00025564958923496306,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.075190815001406
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0028576991262525553,
          "train_loss": 0.002436836715787649,
          "validation_loss": 0.00210479274392128,
          "train_metric": 0.0002454292553011328,
          "validation_metric": 0.00021198703325353563,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.092909333001444
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.00274297802025103,
          "train_loss": 0.003218618221580982,
          "validation_loss": 0.002196247223764658,
          "train_metric": 0.0003241674858145416,
          "validation_metric": 0.00022119797358755022,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.109766691000914
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0028490365730249323,
          "train_loss": 0.00361606664955616,
          "validation_loss": 0.0023912547621876,
          "train_metric": 0.0003641969815362245,
          "validation_metric": 0.00024083843163680285,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.1264045310017536
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.002846019786375109,
          "train_loss": 0.0024282371159642935,
          "validation_loss": 0.002011409495025873,
          "train_metric": 0.00024456315441057086,
          "validation_metric": 0.00020258179574739188,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.144423439000093
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0027837877059937455,
          "train_loss": 0.0025886453222483397,
          "validation_loss": 0.0023770828265696764,
          "train_metric": 0.0002607188653200865,
          "validation_metric": 0.00023941107792779803,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.1755968050001684
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0025080675550270826,
          "train_loss": 0.0023071656469255686,
          "validation_loss": 0.0019889415707439184,
          "train_metric": 0.0002323692460777238,
          "validation_metric": 0.00020031891472171992,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.1935672940016957
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0024350797903025523,
          "train_loss": 0.002585792448371649,
          "validation_loss": 0.0019787440542131662,
          "train_metric": 0.0002604315523058176,
          "validation_metric": 0.00019929185509681702,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2116745710009127
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0024724585491640028,
          "train_loss": 0.0029579512774944305,
          "validation_loss": 0.0021118861623108387,
          "train_metric": 0.0002979140554089099,
          "validation_metric": 0.00021270141587592661,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.228894723000849
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.003456006379565224,
          "train_loss": 0.006605537608265877,
          "validation_loss": 0.0044284602627158165,
          "train_metric": 0.000665285624563694,
          "validation_metric": 0.00044601832632906735,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2458223780013213
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.008367638438357972,
          "train_loss": 0.00667746365070343,
          "validation_loss": 0.003961273934692144,
          "train_metric": 0.0006725296261720359,
          "validation_metric": 0.0003989649994764477,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2624645170017175
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.003778901125770062,
          "train_loss": 0.002751436550170183,
          "validation_loss": 0.0023794749286025763,
          "train_metric": 0.00027711462462320924,
          "validation_metric": 0.00023965199943631887,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2791078690006543
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0027989275549771264,
          "train_loss": 0.0022551422007381916,
          "validation_loss": 0.001931679667904973,
          "train_metric": 0.00022712969803251326,
          "validation_metric": 0.00019455168512649834,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2963972330016986
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0023428753156622406,
          "train_loss": 0.0025034884456545115,
          "validation_loss": 0.002303030574694276,
          "train_metric": 0.0002521421993151307,
          "validation_metric": 0.00023195282847154886,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.3128517090008245
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.002781721450446639,
          "train_loss": 0.0025290122721344233,
          "validation_loss": 0.002239036373794079,
          "train_metric": 0.00025471285334788263,
          "validation_metric": 0.0002255075378343463,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.3295583900016936
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0024593927591922693,
          "train_loss": 0.0028805267065763474,
          "validation_loss": 0.002573669422417879,
          "train_metric": 0.00029011612059548497,
          "validation_metric": 0.00025921055930666625,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.345959831000073
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.002692918249522336,
          "train_loss": 0.002867524977773428,
          "validation_loss": 0.0024963475298136473,
          "train_metric": 0.0002888066228479147,
          "validation_metric": 0.00025142295635305345,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.3619759060002252
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0022859658129164018,
          "train_loss": 0.0021482130978256464,
          "validation_loss": 0.0018289401195943356,
          "train_metric": 0.00021636018936987966,
          "validation_metric": 0.00018420415290165693,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.3796970660005172
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0021850819248356856,
          "train_loss": 0.0021388516761362553,
          "validation_loss": 0.0017434998881071806,
          "train_metric": 0.00021541734167840332,
          "validation_metric": 0.0001755989360390231,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.3971644610010117
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.002214500607806258,
          "train_loss": 0.0021644423250108957,
          "validation_loss": 0.0017437917413190007,
          "train_metric": 0.0002179947041440755,
          "validation_metric": 0.00017562831635586917,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.413926965000428
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.00247939799373853,
          "train_loss": 0.0021963987965136766,
          "validation_loss": 0.0019269349286332726,
          "train_metric": 0.00022121323854662478,
          "validation_metric": 0.000194073814782314,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.430263300000661
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.002343061478313757,
          "train_loss": 0.00215665390715003,
          "validation_loss": 0.0018901220755651593,
          "train_metric": 0.0002172102831536904,
          "validation_metric": 0.00019036619050893933,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.446987817000263
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0022076006353017874,
          "train_loss": 0.002215523738414049,
          "validation_loss": 0.0017410183791071177,
          "train_metric": 0.00022313944646157324,
          "validation_metric": 0.00017534899234306067,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.464049411000815
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0026584236802591477,
          "train_loss": 0.0024428488686680794,
          "validation_loss": 0.002219004090875387,
          "train_metric": 0.0002460347895976156,
          "validation_metric": 0.00022348995844367892,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.480524589000197
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.002309982417500578,
          "train_loss": 0.0021334204357117414,
          "validation_loss": 0.0016990805743262172,
          "train_metric": 0.00021487030608113855,
          "validation_metric": 0.00017112516798079014,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.4976157950004563
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0021716779519920237,
          "train_loss": 0.0021026008762419224,
          "validation_loss": 0.0017640161095187068,
          "train_metric": 0.0002117662806995213,
          "validation_metric": 0.00017766522068995982,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.51404282600015
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.002145257472875528,
          "train_loss": 0.002097312593832612,
          "validation_loss": 0.001718519371934235,
          "train_metric": 0.00021123365149833262,
          "validation_metric": 0.00017308298265561461,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.53056264100087
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0021248970660963096,
          "train_loss": 0.00214741681702435,
          "validation_loss": 0.001894770422950387,
          "train_metric": 0.00021627993555739522,
          "validation_metric": 0.00019083434017375112,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.5470383589999983
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.002222987557615852,
          "train_loss": 0.0021922558080404997,
          "validation_loss": 0.00173779611941427,
          "train_metric": 0.0002207959769293666,
          "validation_metric": 0.00017502445552963763,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.5634690140013845
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0021547073192778043,
          "train_loss": 0.002065409207716584,
          "validation_loss": 0.00172293814830482,
          "train_metric": 0.00020802044309675694,
          "validation_metric": 0.00017352800932712853,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.580004926001493
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0021463388002302963,
          "train_loss": 0.0020682329777628183,
          "validation_loss": 0.001750419964082539,
          "train_metric": 0.00020830486027989537,
          "validation_metric": 0.00017629590001888573,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.5961168680005358
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0021474468885571696,
          "train_loss": 0.0021727553103119135,
          "validation_loss": 0.0017065901774913073,
          "train_metric": 0.00021883199224248528,
          "validation_metric": 0.0001718815037747845,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.612894812000377
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0025694335272419266,
          "train_loss": 0.0020520214457064867,
          "validation_loss": 0.0017194639658555388,
          "train_metric": 0.0002066720771836117,
          "validation_metric": 0.00017317810852546245,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.6298528680017625
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0023227856509038247,
          "train_loss": 0.002083564642816782,
          "validation_loss": 0.0018177415477111936,
          "train_metric": 0.0002098489785566926,
          "validation_metric": 0.00018307626305613667,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.645736742000736
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0023307058872887865,
          "train_loss": 0.002074659802019596,
          "validation_loss": 0.0017126485472545028,
          "train_metric": 0.00020895215857308358,
          "validation_metric": 0.00017249168013222516,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.66237276800166
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.002171964959416073,
          "train_loss": 0.002070782706141472,
          "validation_loss": 0.0017837577033787966,
          "train_metric": 0.0002085616288240999,
          "validation_metric": 0.00017965353617910296,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.679228328001045
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0021622197273245547,
          "train_loss": 0.002202106174081564,
          "validation_loss": 0.001693084486760199,
          "train_metric": 0.00022178808285389096,
          "validation_metric": 0.00017052126349881291,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.696496140000818
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0023952178889885545,
          "train_loss": 0.002250529592856765,
          "validation_loss": 0.0017067443113774061,
          "train_metric": 0.00022666508448310196,
          "validation_metric": 0.00017189703066833317,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7134136730001046
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002377044384047622,
          "train_loss": 0.0020230694208294153,
          "validation_loss": 0.001710841548629105,
          "train_metric": 0.00020375613530632108,
          "validation_metric": 0.0001723096938803792,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7298443880008563
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0021345306631701533,
          "train_loss": 0.0020333535503596067,
          "validation_loss": 0.0017521529225632548,
          "train_metric": 0.00020479191152844578,
          "validation_metric": 0.0001764704065863043,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.746792368001479
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0022078818001318723,
          "train_loss": 0.002161021577194333,
          "validation_loss": 0.0019191030878573656,
          "train_metric": 0.00021765020210295916,
          "validation_metric": 0.00019328499911352992,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7631990990012127
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0020879262538073817,
          "train_loss": 0.0021222124341875315,
          "validation_loss": 0.001697065308690071,
          "train_metric": 0.00021374148491304368,
          "validation_metric": 0.0001709222124191001,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7796497010003804
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.00211915520412731,
          "train_loss": 0.0020109836477786303,
          "validation_loss": 0.001657347078435123,
          "train_metric": 0.00020253888214938343,
          "validation_metric": 0.00016692193457856774,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7969513230000302
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0020602322692866437,
          "train_loss": 0.0020009458530694246,
          "validation_loss": 0.0017017739592120051,
          "train_metric": 0.000201527975150384,
          "validation_metric": 0.00017139645933639258,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.8136399510003685
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0020506933196884347,
          "train_loss": 0.0019919415935873985,
          "validation_loss": 0.0016718355473130941,
          "train_metric": 0.0002006210561376065,
          "validation_metric": 0.00016838117153383791,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.829930357000194
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.002316417379915947,
          "train_loss": 0.0021723490208387375,
          "validation_loss": 0.0019376910058781505,
          "train_metric": 0.00021879102860111743,
          "validation_metric": 0.000195157146663405,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.846403523000845
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0022752305267204065,
          "train_loss": 0.0019885736983269453,
          "validation_loss": 0.0017217487329617143,
          "train_metric": 0.0002002818655455485,
          "validation_metric": 0.00017340823251288384,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.862560460000168
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.002089910278300522,
          "train_loss": 0.002138064941391349,
          "validation_loss": 0.0016606017015874386,
          "train_metric": 0.00021533806284423918,
          "validation_metric": 0.00016724970191717148,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.8787936750013614
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0021891966644034255,
          "train_loss": 0.0020296385046094656,
          "validation_loss": 0.0016011763364076614,
          "train_metric": 0.00020441773813217878,
          "validation_metric": 0.00016126461559906602,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.8964690010016056
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002008050132644712,
          "train_loss": 0.0019713956862688065,
          "validation_loss": 0.0016440065810456872,
          "train_metric": 0.0001985517592402175,
          "validation_metric": 0.00016557831258978695,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.9137188010008686
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0021013054029026534,
          "train_loss": 0.0021338998340070248,
          "validation_loss": 0.0016616865759715438,
          "train_metric": 0.0002149186038877815,
          "validation_metric": 0.0001673589868005365,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.93054512100025
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0020646093907998875,
          "train_loss": 0.0020024296827614307,
          "validation_loss": 0.0017333761788904667,
          "train_metric": 0.00020167739421594888,
          "validation_metric": 0.00017457929789088666,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.947014712000964
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0020180058709229343,
          "train_loss": 0.001955227227881551,
          "validation_loss": 0.0016796872951090336,
          "train_metric": 0.00019692332716658711,
          "validation_metric": 0.00016917195171117783,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.9636984470016614
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.001989331140066497,
          "train_loss": 0.002429925138130784,
          "validation_loss": 0.0017676708521321416,
          "train_metric": 0.00024473314988426864,
          "validation_metric": 0.00017803334048949182,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.9808823040002608
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.002300674541402259,
          "train_loss": 0.0019638047087937593,
          "validation_loss": 0.001695740153081715,
          "train_metric": 0.00019778723071794957,
          "validation_metric": 0.00017078874225262552,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.9977825840014702
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0020501979961409234,
          "train_loss": 0.001944313058629632,
          "validation_loss": 0.0015939971199259162,
          "train_metric": 0.00019582410459406674,
          "validation_metric": 0.00016054156003519893,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.0150552120012435
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0019896450194210047,
          "train_loss": 0.0019348043715581298,
          "validation_loss": 0.0016473334981128573,
          "train_metric": 0.00019486641394905746,
          "validation_metric": 0.00016591339954175055,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.032053656001153
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0019458953938737977,
          "train_loss": 0.0019273520447313786,
          "validation_loss": 0.0016013088170439005,
          "train_metric": 0.0001941158261615783,
          "validation_metric": 0.00016127797425724566,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.048823018001713
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0019549276985344477,
          "train_loss": 0.0019271585624665022,
          "validation_loss": 0.0016486074309796095,
          "train_metric": 0.00019409635569900274,
          "validation_metric": 0.00016604168922640383,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.065663756000504
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.002016080781686469,
          "train_loss": 0.0019496947061270475,
          "validation_loss": 0.0016999848885461688,
          "train_metric": 0.00019636611978057772,
          "validation_metric": 0.00017121623386628926,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.083017258000837
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.002230808056992828,
          "train_loss": 0.0021412756759673357,
          "validation_loss": 0.0019287427421659231,
          "train_metric": 0.00021566145005635917,
          "validation_metric": 0.00019425591744948179,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.100585726000645
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.002078054028970655,
          "train_loss": 0.0019053614232689142,
          "validation_loss": 0.0015723296673968434,
          "train_metric": 0.0001919010392157361,
          "validation_metric": 0.0001583592820679769,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.119094158000735
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0019431264845479745,
          "train_loss": 0.00195978581905365,
          "validation_loss": 0.0017177679110318422,
          "train_metric": 0.00019738245464395732,
          "validation_metric": 0.00017300728359259665,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.136096563001047
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.001995489470573375,
          "train_loss": 0.00197138381190598,
          "validation_loss": 0.001544205006211996,
          "train_metric": 0.000198550580535084,
          "validation_metric": 0.00015552666445728391,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.153486134000559
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0020394663079059683,
          "train_loss": 0.0019981309305876493,
          "validation_loss": 0.0015480315778404474,
          "train_metric": 0.00020124440197832882,
          "validation_metric": 0.00015591208648402244,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.1701099290003185
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0019659853278426453,
          "train_loss": 0.0019725863821804523,
          "validation_loss": 0.001770749338902533,
          "train_metric": 0.00019867168157361448,
          "validation_metric": 0.0001783433835953474,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.1870010280017596
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.001972240937902825,
          "train_loss": 0.0021399217657744884,
          "validation_loss": 0.0016033294377848506,
          "train_metric": 0.00021552508405875415,
          "validation_metric": 0.00016148148279171437,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.204034970000066
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0019395659110159613,
          "train_loss": 0.0018924152245745063,
          "validation_loss": 0.0016076392494142056,
          "train_metric": 0.00019059715850744396,
          "validation_metric": 0.0001619155373191461,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.220772103000854
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0019182832547812723,
          "train_loss": 0.0018953767139464617,
          "validation_loss": 0.001542690908536315,
          "train_metric": 0.00019089541456196457,
          "validation_metric": 0.00015537417493760586,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.238288954000382
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0019462956661300268,
          "train_loss": 0.0019187855068594217,
          "validation_loss": 0.0016586733981966972,
          "train_metric": 0.00019325304310768843,
          "validation_metric": 0.00016705552116036415,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.254798081001354
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0019971773581346497,
          "train_loss": 0.0019065082306042314,
          "validation_loss": 0.001694396836683154,
          "train_metric": 0.0001920165232149884,
          "validation_metric": 0.00017065343854483217,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.271375417001764
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0019162740209139884,
          "train_loss": 0.0020168519113212824,
          "validation_loss": 0.0015275515615940094,
          "train_metric": 0.000203129937290214,
          "validation_metric": 0.00015384939615614712,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.289571094001076
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.001934786603669636,
          "train_loss": 0.0018598813330754638,
          "validation_loss": 0.001589444000273943,
          "train_metric": 0.0001873204455478117,
          "validation_metric": 0.0001600829855306074,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.306748841001536
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0019136869814246893,
          "train_loss": 0.0018748817965388298,
          "validation_loss": 0.0015114285051822662,
          "train_metric": 0.0001888312108349055,
          "validation_metric": 0.00015222554793581367,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.324091593000048
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0019383116195967887,
          "train_loss": 0.0018412458011880517,
          "validation_loss": 0.0015433483058586717,
          "train_metric": 0.00018544355407357216,
          "validation_metric": 0.0001554403715999797,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.340782360000958
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.002037589098108583,
          "train_loss": 0.0019803678151220083,
          "validation_loss": 0.0017683159094303846,
          "train_metric": 0.00019945540407206863,
          "validation_metric": 0.00017809828568715602,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.357623199000955
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.001966494739463087,
          "train_loss": 0.0019282185239717364,
          "validation_loss": 0.001538028009235859,
          "train_metric": 0.00019420312310103327,
          "validation_metric": 0.00015490454097744077,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.3742962810010795
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0019512799372023437,
          "train_loss": 0.0018311787862330675,
          "validation_loss": 0.001551593653857708,
          "train_metric": 0.00018442962027620524,
          "validation_metric": 0.00015627082029823214,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.391100297001685
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0018869199775508605,
          "train_loss": 0.0019753281958401203,
          "validation_loss": 0.0015291995368897915,
          "train_metric": 0.00019894786237273365,
          "validation_metric": 0.00015401537530124187,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.408016777000739
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0018707562248891918,
          "train_loss": 0.0018732337048277259,
          "validation_loss": 0.001643979805521667,
          "train_metric": 0.00018866524624172598,
          "validation_metric": 0.0001655756204854697,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.424591261000387
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.001950766694790218,
          "train_loss": 0.0021558075677603483,
          "validation_loss": 0.001563705736771226,
          "train_metric": 0.00021712505258619785,
          "validation_metric": 0.00015749070735182613,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.4410746570010815
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.002030489999015117,
          "train_loss": 0.0018239234341308475,
          "validation_loss": 0.0015624334337189794,
          "train_metric": 0.00018369888130109757,
          "validation_metric": 0.00015736257773824036,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.45772000100078
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.002220501053670887,
          "train_loss": 0.002060580300167203,
          "validation_loss": 0.001880903379060328,
          "train_metric": 0.00020753411808982491,
          "validation_metric": 0.00018943769100587815,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.474376740001389
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0019798120338236913,
          "train_loss": 0.0018295967020094395,
          "validation_loss": 0.001493909046985209,
          "train_metric": 0.0001842703204602003,
          "validation_metric": 0.00015046104090288281,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.492203696001525
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0020198843285470502,
          "train_loss": 0.0019221253460273147,
          "validation_loss": 0.0017104195430874825,
          "train_metric": 0.00019358942518010736,
          "validation_metric": 0.00017226718773599714,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.5095651440005895
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0017113705889642006,
          "train_loss": 0.0020021821837872267,
          "validation_loss": 0.001479346421547234,
          "train_metric": 0.0002016524667851627,
          "validation_metric": 0.00014899435336701572,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.526771185001053
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.0019057006284128875,
          "train_loss": 0.0017929929308593273,
          "validation_loss": 0.0015248189447447658,
          "train_metric": 0.0001805837091524154,
          "validation_metric": 0.00015357417578343302,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.543820932000017
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0018881010000768583,
          "train_loss": 0.0018534568371251225,
          "validation_loss": 0.0016549284337088466,
          "train_metric": 0.00018667338008526713,
          "validation_metric": 0.00016667833551764488,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.5606146110003465
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.0019226969761803048,
          "train_loss": 0.0018620999762788415,
          "validation_loss": 0.0014837994240224361,
          "train_metric": 0.0001875439047580585,
          "validation_metric": 0.00014944285794626921,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.577721807001581
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.001879545729025267,
          "train_loss": 0.001825918909162283,
          "validation_loss": 0.0016064562369138002,
          "train_metric": 0.00018389987235423177,
          "validation_metric": 0.00016179640078917146,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.594567908001409
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.0017889308219309896,
          "train_loss": 0.0017932786140590906,
          "validation_loss": 0.001438958221115172,
          "train_metric": 0.00018061246373690665,
          "validation_metric": 0.00014492659829556942,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.612246017000871
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.0020046699792146683,
          "train_loss": 0.001927848905324936,
          "validation_loss": 0.0014235218986868858,
          "train_metric": 0.0001941658993018791,
          "validation_metric": 0.00014337192988023162,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.630016819000957
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.0018735782650765032,
          "train_loss": 0.0017608138732612133,
          "validation_loss": 0.0014368332922458649,
          "train_metric": 0.00017734270659275353,
          "validation_metric": 0.00014471259783022106,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.647067152000091
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.0019831132121908013,
          "train_loss": 0.0019525757525116205,
          "validation_loss": 0.0014769736444577575,
          "train_metric": 0.00019665627041831613,
          "validation_metric": 0.00014875536726322025,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.663935372000196
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.0019062790015595965,
          "train_loss": 0.0017410919535905123,
          "validation_loss": 0.0015003010630607605,
          "train_metric": 0.0001753564429236576,
          "validation_metric": 0.00015110484673641622,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.680412079000234
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.001759751405188581,
          "train_loss": 0.0018476179102435708,
          "validation_loss": 0.0014133801450952888,
          "train_metric": 0.0001860853226389736,
          "validation_metric": 0.00014235047274269164,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.697641490000024
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.0019553595175239025,
          "train_loss": 0.0017468681326135993,
          "validation_loss": 0.0014328197576105595,
          "train_metric": 0.00017593815573491156,
          "validation_metric": 0.00014430837472900748,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.714359094001338
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.0022211173018149566,
          "train_loss": 0.002403627848252654,
          "validation_loss": 0.0022247503511607647,
          "train_metric": 0.00024208459944929928,
          "validation_metric": 0.00022406871721614152,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.731223536000471
        }
      ],
      "best": {
        "epoch": 238,
        "updates": 3808
      },
      "best_validation_metric": 0.00014235047274269164,
      "config": {
        "label": "Step decay",
        "batch": 64,
        "lr": 0.003,
        "schedule": "step",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.011931071734873261
    },
    "449a997d9d64_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 1.0181641317903996,
          "train_loss": 1.0090246200561523,
          "validation_loss": 0.7197027206420898,
          "train_metric": 0.10162526369094849,
          "validation_metric": 0.07248582690954208,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.044723048000378185
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 1.00135001167655,
          "train_loss": 0.992499589920044,
          "validation_loss": 0.7088721990585327,
          "train_metric": 0.09996093064546585,
          "validation_metric": 0.07139500975608826,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.06571865300065838
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.9843178056180477,
          "train_loss": 0.9750196933746338,
          "validation_loss": 0.6974018812179565,
          "train_metric": 0.09820041060447693,
          "validation_metric": 0.07023975253105164,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.0835526300015772
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.9662981517612934,
          "train_loss": 0.9560590982437134,
          "validation_loss": 0.6848512887954712,
          "train_metric": 0.09629076719284058,
          "validation_metric": 0.06897571682929993,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.10143446800066158
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.9463809989392757,
          "train_loss": 0.9354147911071777,
          "validation_loss": 0.6712557077407837,
          "train_metric": 0.09421156346797943,
          "validation_metric": 0.06760641932487488,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.11933041400152433
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.9245813973248005,
          "train_loss": 0.9122766852378845,
          "validation_loss": 0.6558070182800293,
          "train_metric": 0.0918811708688736,
          "validation_metric": 0.06605048477649689,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.13704708200020832
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.9005637988448143,
          "train_loss": 0.8858925104141235,
          "validation_loss": 0.638243556022644,
          "train_metric": 0.08922384679317474,
          "validation_metric": 0.06428155303001404,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.15504614500059688
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.8728776536881924,
          "train_loss": 0.8568716049194336,
          "validation_loss": 0.6188462376594543,
          "train_metric": 0.08630096912384033,
          "validation_metric": 0.06232793256640434,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.17256235200147785
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.8426884859800339,
          "train_loss": 0.8241733312606812,
          "validation_loss": 0.5968865156173706,
          "train_metric": 0.08300771564245224,
          "validation_metric": 0.060116227716207504,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.19035224100116466
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.8077005296945572,
          "train_loss": 0.7894571423530579,
          "validation_loss": 0.5734796524047852,
          "train_metric": 0.07951122522354126,
          "validation_metric": 0.05775877460837364,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.20863908900173556
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.7713866718113422,
          "train_loss": 0.749400794506073,
          "validation_loss": 0.5465593338012695,
          "train_metric": 0.07547689974308014,
          "validation_metric": 0.05504745617508888,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.22646232200168015
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.7303116731345654,
          "train_loss": 0.7066187858581543,
          "validation_loss": 0.5175202488899231,
          "train_metric": 0.07116806507110596,
          "validation_metric": 0.052122749388217926,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.24374616500062984
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.6856365837156773,
          "train_loss": 0.6614526510238647,
          "validation_loss": 0.486768901348114,
          "train_metric": 0.06661909073591232,
          "validation_metric": 0.049025580286979675,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.2785959740012913
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.6391964983195066,
          "train_loss": 0.6124037504196167,
          "validation_loss": 0.45329439640045166,
          "train_metric": 0.061679061502218246,
          "validation_metric": 0.04565415158867836,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.601979201001086
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.5893467646092176,
          "train_loss": 0.5618070960044861,
          "validation_loss": 0.4185435175895691,
          "train_metric": 0.056583158671855927,
          "validation_metric": 0.04215417057275772,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.648167742001533
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.5373027306050062,
          "train_loss": 0.5108941197395325,
          "validation_loss": 0.3831537365913391,
          "train_metric": 0.05145538970828056,
          "validation_metric": 0.03858984261751175,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6726372380016983
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.48562802374362946,
          "train_loss": 0.45845893025398254,
          "validation_loss": 0.3466431498527527,
          "train_metric": 0.0461742989718914,
          "validation_metric": 0.03491262346506119,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.6998184340009175
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.4341329596936703,
          "train_loss": 0.40655386447906494,
          "validation_loss": 0.31028279662132263,
          "train_metric": 0.04094661399722099,
          "validation_metric": 0.031250547617673874,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7256832410002971
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.3832380585372448,
          "train_loss": 0.3583773076534271,
          "validation_loss": 0.2761671841144562,
          "train_metric": 0.03609445318579674,
          "validation_metric": 0.02781454473733902,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7468915160006873
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.3367975428700447,
          "train_loss": 0.31300249695777893,
          "validation_loss": 0.24374714493751526,
          "train_metric": 0.03152446076273918,
          "validation_metric": 0.024549320340156555,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.7650422530005017
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.2939921170473099,
          "train_loss": 0.27279847860336304,
          "validation_loss": 0.21482539176940918,
          "train_metric": 0.0274752639234066,
          "validation_metric": 0.02163642644882202,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.78291422500115
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.2559880493208766,
          "train_loss": 0.2390541434288025,
          "validation_loss": 0.18994052708148956,
          "train_metric": 0.024076659232378006,
          "validation_metric": 0.01913011260330677,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8060381770010281
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.22530293744057417,
          "train_loss": 0.2103351354598999,
          "validation_loss": 0.1686915159225464,
          "train_metric": 0.02118418551981449,
          "validation_metric": 0.016989992931485176,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8422267970017856
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.19958616700023413,
          "train_loss": 0.18804068863391876,
          "validation_loss": 0.1518644243478775,
          "train_metric": 0.01893877051770687,
          "validation_metric": 0.015295227989554405,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.8791597900017223
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.179584383033216,
          "train_loss": 0.17146405577659607,
          "validation_loss": 0.13922449946403503,
          "train_metric": 0.017269231379032135,
          "validation_metric": 0.014022179879248142,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9264900570015016
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.16526575526222587,
          "train_loss": 0.15868869423866272,
          "validation_loss": 0.1293826699256897,
          "train_metric": 0.01598254404962063,
          "validation_metric": 0.013030949980020523,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 0.9696522480007843
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.15439275605604053,
          "train_loss": 0.14970561861991882,
          "validation_loss": 0.12250709533691406,
          "train_metric": 0.01507780235260725,
          "validation_metric": 0.012338466942310333,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0155103000015515
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.14652589662000537,
          "train_loss": 0.14319540560245514,
          "validation_loss": 0.1174580305814743,
          "train_metric": 0.014422115869820118,
          "validation_metric": 0.011829942464828491,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.036843186000624
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.14084398047998548,
          "train_loss": 0.1383742094039917,
          "validation_loss": 0.11376190185546875,
          "train_metric": 0.0139365429058671,
          "validation_metric": 0.011457682587206364,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.054659529001583
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.13662891881540418,
          "train_loss": 0.1346127986907959,
          "validation_loss": 0.11092957854270935,
          "train_metric": 0.013557708822190762,
          "validation_metric": 0.011172421276569366,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0722223669999948
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.1332504074089229,
          "train_loss": 0.1315782219171524,
          "validation_loss": 0.10866496711969376,
          "train_metric": 0.013252075761556625,
          "validation_metric": 0.010944337584078312,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.0896899290000874
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.13046850683167577,
          "train_loss": 0.1289498209953308,
          "validation_loss": 0.10669705271720886,
          "train_metric": 0.012987354770302773,
          "validation_metric": 0.01074613630771637,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.107717307000712
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.12807933520525694,
          "train_loss": 0.12666551768779755,
          "validation_loss": 0.10498982667922974,
          "train_metric": 0.012757288292050362,
          "validation_metric": 0.010574190877377987,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1260589850007818
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.12573716952465475,
          "train_loss": 0.12463565170764923,
          "validation_loss": 0.10340660810470581,
          "train_metric": 0.012552846223115921,
          "validation_metric": 0.010414734482765198,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1435332230012136
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.1237924718298018,
          "train_loss": 0.12273932993412018,
          "validation_loss": 0.1018964871764183,
          "train_metric": 0.012361856177449226,
          "validation_metric": 0.010262642055749893,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1612293340003816
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.121854018419981,
          "train_loss": 0.12080737203359604,
          "validation_loss": 0.10041753202676773,
          "train_metric": 0.012167277745902538,
          "validation_metric": 0.010113687254488468,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.1785834790007357
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.11993204918690026,
          "train_loss": 0.11900446563959122,
          "validation_loss": 0.09903397411108017,
          "train_metric": 0.01198569405823946,
          "validation_metric": 0.009974339045584202,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2010479740001756
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.11821794789284468,
          "train_loss": 0.11725421249866486,
          "validation_loss": 0.09763326495885849,
          "train_metric": 0.011809414252638817,
          "validation_metric": 0.009833265095949173,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2216091150003194
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.11668987479060888,
          "train_loss": 0.11555871367454529,
          "validation_loss": 0.0963173434138298,
          "train_metric": 0.011638649739325047,
          "validation_metric": 0.009700728580355644,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2470990159999928
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.1148666413500905,
          "train_loss": 0.11393845826387405,
          "validation_loss": 0.09494313597679138,
          "train_metric": 0.011475464329123497,
          "validation_metric": 0.009562324732542038,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.2651233010001306
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.11325082904659212,
          "train_loss": 0.11246871203184128,
          "validation_loss": 0.09365823119878769,
          "train_metric": 0.011327437125146389,
          "validation_metric": 0.009432914666831493,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.284730245000901
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.11174584832042456,
          "train_loss": 0.11090318113565445,
          "validation_loss": 0.09239872545003891,
          "train_metric": 0.011169763281941414,
          "validation_metric": 0.009306061081588268,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3028629800010094
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.11024895985610783,
          "train_loss": 0.10942181199789047,
          "validation_loss": 0.09119133651256561,
          "train_metric": 0.01102056447416544,
          "validation_metric": 0.009184456430375576,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3207862000017485
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.10879282024689019,
          "train_loss": 0.10793640464544296,
          "validation_loss": 0.08994779735803604,
          "train_metric": 0.010870959609746933,
          "validation_metric": 0.009059213101863861,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3405162650014972
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.10731082153506577,
          "train_loss": 0.10651964694261551,
          "validation_loss": 0.08872167766094208,
          "train_metric": 0.010728269815444946,
          "validation_metric": 0.008935722522437572,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3690558460002649
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.10590213676914573,
          "train_loss": 0.1051020398736,
          "validation_loss": 0.0875159502029419,
          "train_metric": 0.010585492476820946,
          "validation_metric": 0.008814285509288311,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.3873324760006653
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.10453967750072479,
          "train_loss": 0.10374396294355392,
          "validation_loss": 0.08634193986654282,
          "train_metric": 0.010448711924254894,
          "validation_metric": 0.00869604293256998,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.405580409000322
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.10313244466669858,
          "train_loss": 0.1024014875292778,
          "validation_loss": 0.08520184457302094,
          "train_metric": 0.010313502512872219,
          "validation_metric": 0.008581217378377914,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.4232206190008583
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.10182218160480261,
          "train_loss": 0.10104809701442719,
          "validation_loss": 0.08405205607414246,
          "train_metric": 0.010177193209528923,
          "validation_metric": 0.008465414866805077,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.441463051000028
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.10054683801718056,
          "train_loss": 0.09976523369550705,
          "validation_loss": 0.082924485206604,
          "train_metric": 0.010047989897429943,
          "validation_metric": 0.00835185032337904,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.5612006610008393
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.09918995900079608,
          "train_loss": 0.09851589798927307,
          "validation_loss": 0.08180892467498779,
          "train_metric": 0.00992216169834137,
          "validation_metric": 0.008239495567977428,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6081708720012102
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0979508135933429,
          "train_loss": 0.09727303683757782,
          "validation_loss": 0.08073524385690689,
          "train_metric": 0.009796984493732452,
          "validation_metric": 0.008131357841193676,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6518670630011911
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.09673515078611672,
          "train_loss": 0.09603121876716614,
          "validation_loss": 0.07966114580631256,
          "train_metric": 0.009671914391219616,
          "validation_metric": 0.008023180067539215,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.6893778840003506
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.09551479713991284,
          "train_loss": 0.09479688107967377,
          "validation_loss": 0.0785631611943245,
          "train_metric": 0.00954759493470192,
          "validation_metric": 0.007912594825029373,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7070431820011436
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.09429742093198001,
          "train_loss": 0.0936059057712555,
          "validation_loss": 0.07751600444316864,
          "train_metric": 0.00942764338105917,
          "validation_metric": 0.007807128597050905,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7245244510013436
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.09304456668905914,
          "train_loss": 0.09243926405906677,
          "validation_loss": 0.07652387768030167,
          "train_metric": 0.009310143068432808,
          "validation_metric": 0.007707204669713974,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7422697620004328
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.09194117528386414,
          "train_loss": 0.09124128520488739,
          "validation_loss": 0.07546890527009964,
          "train_metric": 0.009189488366246223,
          "validation_metric": 0.007600952405482531,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7601100420015428
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.09081339277327061,
          "train_loss": 0.09011714905500412,
          "validation_loss": 0.07446164637804031,
          "train_metric": 0.009076269343495369,
          "validation_metric": 0.007499503903090954,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7774498740000126
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.08973115216940641,
          "train_loss": 0.08896683156490326,
          "validation_loss": 0.07344426214694977,
          "train_metric": 0.008960412815213203,
          "validation_metric": 0.007397038396447897,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.7953086250017805
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.08852235646918416,
          "train_loss": 0.08790366351604462,
          "validation_loss": 0.07246571034193039,
          "train_metric": 0.008853334933519363,
          "validation_metric": 0.007298481650650501,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8130070410006738
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.08746814355254173,
          "train_loss": 0.08683499693870544,
          "validation_loss": 0.0715341717004776,
          "train_metric": 0.008745702914893627,
          "validation_metric": 0.007204659748822451,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8306736220001767
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.08646230935119092,
          "train_loss": 0.08575782179832458,
          "validation_loss": 0.07059451937675476,
          "train_metric": 0.00863721314817667,
          "validation_metric": 0.0071100215427577496,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8484356130011292
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.0854126000776887,
          "train_loss": 0.08469394594430923,
          "validation_loss": 0.06965243816375732,
          "train_metric": 0.008530063554644585,
          "validation_metric": 0.007015138864517212,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8663257470016106
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.084258092334494,
          "train_loss": 0.08367147296667099,
          "validation_loss": 0.06872906535863876,
          "train_metric": 0.008427084423601627,
          "validation_metric": 0.006922140251845121,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.8841838960015593
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.08324715727940202,
          "train_loss": 0.08267682790756226,
          "validation_loss": 0.0678100436925888,
          "train_metric": 0.008326908573508263,
          "validation_metric": 0.0068295784294605255,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9015807200012205
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.08227651054039598,
          "train_loss": 0.08168444782495499,
          "validation_loss": 0.06696369498968124,
          "train_metric": 0.00822695903480053,
          "validation_metric": 0.006744338199496269,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9218751190001058
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.08139558439143002,
          "train_loss": 0.08074869215488434,
          "validation_loss": 0.06610191613435745,
          "train_metric": 0.008132711984217167,
          "validation_metric": 0.006657542660832405,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9588399750009557
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.08031122200191021,
          "train_loss": 0.07974547892808914,
          "validation_loss": 0.06522586941719055,
          "train_metric": 0.008031672798097134,
          "validation_metric": 0.006569310557097197,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9763537510007154
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.07937867171131074,
          "train_loss": 0.07884196937084198,
          "validation_loss": 0.06442420929670334,
          "train_metric": 0.007940675131976604,
          "validation_metric": 0.006488570012152195,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 1.9942556640016846
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.07858691504225135,
          "train_loss": 0.07790637761354446,
          "validation_loss": 0.06358152627944946,
          "train_metric": 0.007846444845199585,
          "validation_metric": 0.006403698585927486,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0117098030004854
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.07752638671081513,
          "train_loss": 0.0770343542098999,
          "validation_loss": 0.06281416863203049,
          "train_metric": 0.007758617866784334,
          "validation_metric": 0.006326412782073021,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0292556160002277
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.07673805532976985,
          "train_loss": 0.07614309340715408,
          "validation_loss": 0.062015682458877563,
          "train_metric": 0.0076688542030751705,
          "validation_metric": 0.006245993077754974,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0469131540012313
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.07578345760703087,
          "train_loss": 0.07531502097845078,
          "validation_loss": 0.061290107667446136,
          "train_metric": 0.007585453800857067,
          "validation_metric": 0.006172915454953909,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.064661104001061
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.07495633163489401,
          "train_loss": 0.07447028160095215,
          "validation_loss": 0.06053609400987625,
          "train_metric": 0.007500373758375645,
          "validation_metric": 0.006096973549574614,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0819957090006938
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.07411659485660493,
          "train_loss": 0.07364438474178314,
          "validation_loss": 0.05979417264461517,
          "train_metric": 0.007417193613946438,
          "validation_metric": 0.006022250279784203,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.0997156980010914
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.07333636819384992,
          "train_loss": 0.07284481823444366,
          "validation_loss": 0.05909651145339012,
          "train_metric": 0.007336663547903299,
          "validation_metric": 0.005951984785497189,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.127534667000873
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.07268078834749758,
          "train_loss": 0.07203298062086105,
          "validation_loss": 0.05835258960723877,
          "train_metric": 0.007254897151142359,
          "validation_metric": 0.005877058953046799,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.1456980020011542
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.07174397935159504,
          "train_loss": 0.07128211110830307,
          "validation_loss": 0.057702891528606415,
          "train_metric": 0.00717927236109972,
          "validation_metric": 0.005811624228954315,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.163440337000793
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0710003151325509,
          "train_loss": 0.07055007666349411,
          "validation_loss": 0.05705637112259865,
          "train_metric": 0.00710554514080286,
          "validation_metric": 0.005746508948504925,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.1813746390016604
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.07024559448473155,
          "train_loss": 0.06984436511993408,
          "validation_loss": 0.05645763501524925,
          "train_metric": 0.0070344689302146435,
          "validation_metric": 0.00568620627745986,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.1988525290016696
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.06959768745582551,
          "train_loss": 0.06913153827190399,
          "validation_loss": 0.05583386868238449,
          "train_metric": 0.0069626751355826855,
          "validation_metric": 0.005623382981866598,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.216095805000805
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.06893030623905361,
          "train_loss": 0.06849200278520584,
          "validation_loss": 0.0553012378513813,
          "train_metric": 0.006898263469338417,
          "validation_metric": 0.005569737404584885,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.233446323001772
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0683209173148498,
          "train_loss": 0.06781435012817383,
          "validation_loss": 0.054660629481077194,
          "train_metric": 0.006830012891441584,
          "validation_metric": 0.005505218170583248,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.251494942000136
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0675723449094221,
          "train_loss": 0.06714756786823273,
          "validation_loss": 0.05406143143773079,
          "train_metric": 0.006762857083231211,
          "validation_metric": 0.00544486939907074,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.269120181001199
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.06697710929438472,
          "train_loss": 0.06655360013246536,
          "validation_loss": 0.053593672811985016,
          "train_metric": 0.006703035440295935,
          "validation_metric": 0.005397758446633816,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.2867537190013536
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.06629720190539956,
          "train_loss": 0.06590428203344345,
          "validation_loss": 0.05300436541438103,
          "train_metric": 0.0066376375034451485,
          "validation_metric": 0.005338405258953571,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.404411251000056
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.06574253889266402,
          "train_loss": 0.06531552970409393,
          "validation_loss": 0.05246255174279213,
          "train_metric": 0.006578341126441956,
          "validation_metric": 0.005283836275339127,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4225604250004835
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.06511697836685926,
          "train_loss": 0.06475049257278442,
          "validation_loss": 0.052024129778146744,
          "train_metric": 0.0065214321948587894,
          "validation_metric": 0.005239680875092745,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.4402947840007982
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.06460500753019005,
          "train_loss": 0.06426170468330383,
          "validation_loss": 0.0516604408621788,
          "train_metric": 0.006472203880548477,
          "validation_metric": 0.0052030510269105434,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.478277849000733
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.06399743515066803,
          "train_loss": 0.06368453800678253,
          "validation_loss": 0.0511382557451725,
          "train_metric": 0.006414072588086128,
          "validation_metric": 0.005150457378476858,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.520525735000774
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.06346305296756327,
          "train_loss": 0.06317756325006485,
          "validation_loss": 0.050651632249355316,
          "train_metric": 0.006363013293594122,
          "validation_metric": 0.005101446527987719,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.5673874570002226
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0630023037083447,
          "train_loss": 0.0626397579908371,
          "validation_loss": 0.050169914960861206,
          "train_metric": 0.006308847106993198,
          "validation_metric": 0.0050529297441244125,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.585548507000567
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.06253378605470061,
          "train_loss": 0.06217129901051521,
          "validation_loss": 0.049861032515764236,
          "train_metric": 0.006261665839701891,
          "validation_metric": 0.005021820776164532,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.603709588000129
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.06198464194312692,
          "train_loss": 0.06169375404715538,
          "validation_loss": 0.04943009465932846,
          "train_metric": 0.0062135690823197365,
          "validation_metric": 0.004978417884558439,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.6210072780013434
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.06160609575454146,
          "train_loss": 0.06121928244829178,
          "validation_loss": 0.04901433736085892,
          "train_metric": 0.006165782455354929,
          "validation_metric": 0.004936544690281153,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.638701381001738
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.06108633056282997,
          "train_loss": 0.06078461557626724,
          "validation_loss": 0.048687294125556946,
          "train_metric": 0.0061220042407512665,
          "validation_metric": 0.0049036056734621525,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.655361011000423
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.06062686594668776,
          "train_loss": 0.06037785857915878,
          "validation_loss": 0.04837433248758316,
          "train_metric": 0.006081037223339081,
          "validation_metric": 0.004872085526585579,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.672603295000954
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.06025839620269835,
          "train_loss": 0.05993939936161041,
          "validation_loss": 0.0479697659611702,
          "train_metric": 0.006036876700818539,
          "validation_metric": 0.0048313383013010025,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.6903304680017754
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.059804884949699044,
          "train_loss": 0.05954980477690697,
          "validation_loss": 0.047665514051914215,
          "train_metric": 0.005997638218104839,
          "validation_metric": 0.004800695925951004,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7083358770014456
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.059426975436508656,
          "train_loss": 0.05918135493993759,
          "validation_loss": 0.047409892082214355,
          "train_metric": 0.005960530135780573,
          "validation_metric": 0.004774951376020908,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7262204450016725
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.059073072858154774,
          "train_loss": 0.05882460996508598,
          "validation_loss": 0.04711513593792915,
          "train_metric": 0.00592459924519062,
          "validation_metric": 0.004745263606309891,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.7504239010013407
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.05886441748589277,
          "train_loss": 0.05844079330563545,
          "validation_loss": 0.046707041561603546,
          "train_metric": 0.005885943304747343,
          "validation_metric": 0.004704162012785673,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.793006775000322
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.05846624542027712,
          "train_loss": 0.05813883990049362,
          "validation_loss": 0.046598486602306366,
          "train_metric": 0.005855531431734562,
          "validation_metric": 0.004693229217082262,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.836006139001256
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.05801626946777105,
          "train_loss": 0.05774508789181709,
          "validation_loss": 0.04616626352071762,
          "train_metric": 0.005815873853862286,
          "validation_metric": 0.004649697337299585,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.879236237000441
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.05765987036284059,
          "train_loss": 0.05741459131240845,
          "validation_loss": 0.04591233655810356,
          "train_metric": 0.005782587919384241,
          "validation_metric": 0.00462412228807807,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.9262700080016657
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.057295265258289874,
          "train_loss": 0.057095058262348175,
          "validation_loss": 0.04568784683942795,
          "train_metric": 0.005750405136495829,
          "validation_metric": 0.00460151256993413,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 2.9690347660016414
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.05711951432749629,
          "train_loss": 0.05680885165929794,
          "validation_loss": 0.04550371319055557,
          "train_metric": 0.005721579305827618,
          "validation_metric": 0.004582967143505812,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0136700380007824
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.05671391426585615,
          "train_loss": 0.05649891868233681,
          "validation_loss": 0.045198213309049606,
          "train_metric": 0.005690363701432943,
          "validation_metric": 0.004552198573946953,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.031098228000701
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.056506922584958375,
          "train_loss": 0.05621550977230072,
          "validation_loss": 0.04494912922382355,
          "train_metric": 0.005661820527166128,
          "validation_metric": 0.004527111072093248,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0489817100005894
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0561273277271539,
          "train_loss": 0.05592207610607147,
          "validation_loss": 0.044794898480176926,
          "train_metric": 0.005632266867905855,
          "validation_metric": 0.004511578474193811,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.066515282000182
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.05591402028221637,
          "train_loss": 0.05570107325911522,
          "validation_loss": 0.04469753056764603,
          "train_metric": 0.005610008258372545,
          "validation_metric": 0.00450177164748311,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.0842482860007294
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.055761834722943604,
          "train_loss": 0.055397339165210724,
          "validation_loss": 0.044339556246995926,
          "train_metric": 0.005579417105764151,
          "validation_metric": 0.004465717822313309,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1017259850013943
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.055423695128411055,
          "train_loss": 0.05512489750981331,
          "validation_loss": 0.04415540397167206,
          "train_metric": 0.005551978014409542,
          "validation_metric": 0.004447170067578554,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1200172290009505
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0550101506523788,
          "train_loss": 0.054913878440856934,
          "validation_loss": 0.044063206762075424,
          "train_metric": 0.005530724301934242,
          "validation_metric": 0.004437885247170925,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.137606237
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.05484564718790352,
          "train_loss": 0.05463949218392372,
          "validation_loss": 0.04384768381714821,
          "train_metric": 0.005503089167177677,
          "validation_metric": 0.004416178446263075,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1551685750000615
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.054682755609974265,
          "train_loss": 0.054400522261857986,
          "validation_loss": 0.04367803782224655,
          "train_metric": 0.005479021463543177,
          "validation_metric": 0.004399091936647892,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.172532101001707
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0543772941455245,
          "train_loss": 0.05416710674762726,
          "validation_loss": 0.04342951253056526,
          "train_metric": 0.0054555125534534454,
          "validation_metric": 0.004374061711132526,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.1910100060013065
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.054119368083775043,
          "train_loss": 0.053935714066028595,
          "validation_loss": 0.043276309967041016,
          "train_metric": 0.005432207603007555,
          "validation_metric": 0.00435863109305501,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.209106602000247
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.053952262620441616,
          "train_loss": 0.053707946091890335,
          "validation_loss": 0.043057844042778015,
          "train_metric": 0.0054092672653496265,
          "validation_metric": 0.004336628131568432,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.2275575190014933
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.05363189650233835,
          "train_loss": 0.053497523069381714,
          "validation_loss": 0.04297981411218643,
          "train_metric": 0.005388073623180389,
          "validation_metric": 0.0043287696316838264,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.2458922660007374
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.05357589584309608,
          "train_loss": 0.053280964493751526,
          "validation_loss": 0.042783819139003754,
          "train_metric": 0.005366263445466757,
          "validation_metric": 0.004309029318392277,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.263454101001116
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.05326322535984218,
          "train_loss": 0.053073152899742126,
          "validation_loss": 0.04259404167532921,
          "train_metric": 0.005345333367586136,
          "validation_metric": 0.004289915319532156,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.3847061400010716
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.05309635098092258,
          "train_loss": 0.05285811424255371,
          "validation_loss": 0.04245227202773094,
          "train_metric": 0.0053236763924360275,
          "validation_metric": 0.004275637213140726,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4058817340010137
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.052993528079241514,
          "train_loss": 0.05265622213482857,
          "validation_loss": 0.04233243316411972,
          "train_metric": 0.005303341895341873,
          "validation_metric": 0.0042635672725737095,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4236122790007357
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.05266004474833608,
          "train_loss": 0.05247735232114792,
          "validation_loss": 0.04224778711795807,
          "train_metric": 0.005285326857119799,
          "validation_metric": 0.004255041480064392,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.44159184900127
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0524467877112329,
          "train_loss": 0.052291788160800934,
          "validation_loss": 0.042020853608846664,
          "train_metric": 0.00526663800701499,
          "validation_metric": 0.004232186358422041,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.459607334001703
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.05228726449422538,
          "train_loss": 0.05209272354841232,
          "validation_loss": 0.041875120252370834,
          "train_metric": 0.0052465880289673805,
          "validation_metric": 0.004217508714646101,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4783091960016463
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.052079706569202244,
          "train_loss": 0.051910582929849625,
          "validation_loss": 0.04182814061641693,
          "train_metric": 0.005228244233876467,
          "validation_metric": 0.00421277666464448,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.4965184470001986
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.05194075335748494,
          "train_loss": 0.05173776298761368,
          "validation_loss": 0.041716091334819794,
          "train_metric": 0.005210838280618191,
          "validation_metric": 0.004201491363346577,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.514769925001019
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.051680607022717595,
          "train_loss": 0.05155225843191147,
          "validation_loss": 0.04146574065089226,
          "train_metric": 0.00519215501844883,
          "validation_metric": 0.004176277667284012,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5332738990000507
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.051517561776563525,
          "train_loss": 0.051348693668842316,
          "validation_loss": 0.04133840650320053,
          "train_metric": 0.005171653348952532,
          "validation_metric": 0.004163452889770269,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.55176959200071
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.051480535068549216,
          "train_loss": 0.05118953064084053,
          "validation_loss": 0.0411936491727829,
          "train_metric": 0.005155622493475676,
          "validation_metric": 0.004148873500525951,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.5696427850016335
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.05125152016989887,
          "train_loss": 0.051021989434957504,
          "validation_loss": 0.041187308728694916,
          "train_metric": 0.005138748325407505,
          "validation_metric": 0.004148234613239765,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.587287110000034
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.05111500155180693,
          "train_loss": 0.05085580050945282,
          "validation_loss": 0.04093850031495094,
          "train_metric": 0.005122009664773941,
          "validation_metric": 0.0041231755167245865,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.6046262410000054
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.05084383673965931,
          "train_loss": 0.050666213035583496,
          "validation_loss": 0.04092059284448624,
          "train_metric": 0.005102915223687887,
          "validation_metric": 0.004121372476220131,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.622828184001264
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.050604436779394746,
          "train_loss": 0.05047939345240593,
          "validation_loss": 0.04070889949798584,
          "train_metric": 0.005084099248051643,
          "validation_metric": 0.0041000512428581715,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.64029696200123
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.050630192970857024,
          "train_loss": 0.05031678453087807,
          "validation_loss": 0.040581606328487396,
          "train_metric": 0.0050677224062383175,
          "validation_metric": 0.0040872301906347275,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.657770657000583
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.05029567494057119,
          "train_loss": 0.05013474449515343,
          "validation_loss": 0.040477629750967026,
          "train_metric": 0.00504938792437315,
          "validation_metric": 0.004076758865267038,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.675828054001613
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.050164065789431334,
          "train_loss": 0.049975063651800156,
          "validation_loss": 0.04037650674581528,
          "train_metric": 0.0050333053804934025,
          "validation_metric": 0.004066574387252331,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.69572010200136
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.050050349556840956,
          "train_loss": 0.049839507788419724,
          "validation_loss": 0.040367744863033295,
          "train_metric": 0.0050196521915495396,
          "validation_metric": 0.004065691493451595,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7277958400009084
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.04979700071271509,
          "train_loss": 0.04964667186141014,
          "validation_loss": 0.040105514228343964,
          "train_metric": 0.005000231321901083,
          "validation_metric": 0.004039280116558075,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7465077090000705
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.04969384241849184,
          "train_loss": 0.049475785344839096,
          "validation_loss": 0.03998566418886185,
          "train_metric": 0.004983019549399614,
          "validation_metric": 0.004027209710329771,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.772646353001619
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.049624410341493785,
          "train_loss": 0.04931703954935074,
          "validation_loss": 0.03983529657125473,
          "train_metric": 0.00496703153476119,
          "validation_metric": 0.004012065473943949,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.7906411860003573
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.049323557468596846,
          "train_loss": 0.04914627596735954,
          "validation_loss": 0.03974037989974022,
          "train_metric": 0.004949832335114479,
          "validation_metric": 0.004002505913376808,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8083862060011597
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.049230595701374114,
          "train_loss": 0.048981014639139175,
          "validation_loss": 0.03963606804609299,
          "train_metric": 0.004933188669383526,
          "validation_metric": 0.003992000129073858,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8271947860012006
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.049027823959477246,
          "train_loss": 0.048820700496435165,
          "validation_loss": 0.03951655700802803,
          "train_metric": 0.004917042329907417,
          "validation_metric": 0.003979962784796953,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.8457423930012737
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.048810044885613024,
          "train_loss": 0.048675015568733215,
          "validation_loss": 0.03947664052248001,
          "train_metric": 0.004902369808405638,
          "validation_metric": 0.003975942265242338,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.893535214001531
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.04867168748751283,
          "train_loss": 0.048500075936317444,
          "validation_loss": 0.03927890211343765,
          "train_metric": 0.004884750582277775,
          "validation_metric": 0.003956027328968048,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9154988790014613
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.048479085438884795,
          "train_loss": 0.048346832394599915,
          "validation_loss": 0.03918490558862686,
          "train_metric": 0.0048693157732486725,
          "validation_metric": 0.0039465599693357944,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9336747790002846
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0484043681062758,
          "train_loss": 0.04820353537797928,
          "validation_loss": 0.038985952734947205,
          "train_metric": 0.004854883998632431,
          "validation_metric": 0.003926522098481655,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9511901650002983
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.048545173136517406,
          "train_loss": 0.04808531701564789,
          "validation_loss": 0.039082616567611694,
          "train_metric": 0.004842977039515972,
          "validation_metric": 0.003936258610337973,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.9675965160004125
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.04852082428988069,
          "train_loss": 0.047898706048727036,
          "validation_loss": 0.03875140845775604,
          "train_metric": 0.0048241824842989445,
          "validation_metric": 0.003902899567037821,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 3.984709495000061
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.04787223017774522,
          "train_loss": 0.047707702964544296,
          "validation_loss": 0.038713134825229645,
          "train_metric": 0.0048049446195364,
          "validation_metric": 0.0038990448229014874,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.002060392000203
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.04770116484723985,
          "train_loss": 0.04757242649793625,
          "validation_loss": 0.03865542262792587,
          "train_metric": 0.004791320767253637,
          "validation_metric": 0.0038932322058826685,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.019918274001611
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.04753037367481738,
          "train_loss": 0.04739970713853836,
          "validation_loss": 0.038478124886751175,
          "train_metric": 0.0047739241272211075,
          "validation_metric": 0.003875375958159566,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.038543036000192
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.047457824693992734,
          "train_loss": 0.04723712429404259,
          "validation_loss": 0.0383346825838089,
          "train_metric": 0.004757550545036793,
          "validation_metric": 0.0038609295152127743,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.056198209000286
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.04725286108441651,
          "train_loss": 0.04708537831902504,
          "validation_loss": 0.038153503090143204,
          "train_metric": 0.0047422670759260654,
          "validation_metric": 0.003842680947855115,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.074202927000442
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.047044627950526774,
          "train_loss": 0.04691240191459656,
          "validation_loss": 0.03809290751814842,
          "train_metric": 0.004724845755845308,
          "validation_metric": 0.0038365779910236597,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.189221133001411
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.047071619192138314,
          "train_loss": 0.046750251203775406,
          "validation_loss": 0.03794693574309349,
          "train_metric": 0.004708513617515564,
          "validation_metric": 0.0038218763656914234,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.208164640000177
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.04698745836503804,
          "train_loss": 0.046639882028102875,
          "validation_loss": 0.0379936620593071,
          "train_metric": 0.004697397351264954,
          "validation_metric": 0.0038265823386609554,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.226045143001102
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.046575854532420635,
          "train_loss": 0.04646851867437363,
          "validation_loss": 0.03767402470111847,
          "train_metric": 0.004680139012634754,
          "validation_metric": 0.003794389311224222,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.245435158001783
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.04660443670582026,
          "train_loss": 0.04629535600543022,
          "validation_loss": 0.037554845213890076,
          "train_metric": 0.00466269813477993,
          "validation_metric": 0.003782386425882578,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.26345115300137
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.046344871749170125,
          "train_loss": 0.046122197061777115,
          "validation_loss": 0.037532057613134384,
          "train_metric": 0.004645258188247681,
          "validation_metric": 0.003780091181397438,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.281839135001064
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.04621870326809585,
          "train_loss": 0.04598867893218994,
          "validation_loss": 0.037482187151908875,
          "train_metric": 0.004631811287254095,
          "validation_metric": 0.0037750687915831804,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.29939977200047
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.04599758342374116,
          "train_loss": 0.04581339657306671,
          "validation_loss": 0.03721247985959053,
          "train_metric": 0.004614157136529684,
          "validation_metric": 0.003747904673218727,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.318133141001454
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.04642121319193393,
          "train_loss": 0.04571852087974548,
          "validation_loss": 0.037325866520404816,
          "train_metric": 0.0046046022325754166,
          "validation_metric": 0.0037593243177980185,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.335065163000763
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0455853781895712,
          "train_loss": 0.045580390840768814,
          "validation_loss": 0.03699113428592682,
          "train_metric": 0.0045906901359558105,
          "validation_metric": 0.0037256113719195127,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.382182817000285
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.04552717681508511,
          "train_loss": 0.045331571251153946,
          "validation_loss": 0.036867134273052216,
          "train_metric": 0.00456562964245677,
          "validation_metric": 0.0037131228018552065,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.410722536000321
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.04534459894057363,
          "train_loss": 0.045171089470386505,
          "validation_loss": 0.036767419427633286,
          "train_metric": 0.004549466539174318,
          "validation_metric": 0.003703080117702484,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.4287600360003125
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.045196443679742515,
          "train_loss": 0.0450202114880085,
          "validation_loss": 0.03669754043221474,
          "train_metric": 0.004534271080046892,
          "validation_metric": 0.0036960418801754713,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.453481685000952
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.04510426660999656,
          "train_loss": 0.04488922655582428,
          "validation_loss": 0.036492858082056046,
          "train_metric": 0.004521078430116177,
          "validation_metric": 0.003675426822155714,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.471717614000227
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.044857512577436864,
          "train_loss": 0.04472288861870766,
          "validation_loss": 0.03650890663266182,
          "train_metric": 0.004504325333982706,
          "validation_metric": 0.003677043132483959,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.488186735999989
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.04476918431464583,
          "train_loss": 0.04453820362687111,
          "validation_loss": 0.0362788550555706,
          "train_metric": 0.004485724493861198,
          "validation_metric": 0.003653873223811388,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.5051257140003145
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.04453985020518303,
          "train_loss": 0.044374506920576096,
          "validation_loss": 0.036155302077531815,
          "train_metric": 0.004469238221645355,
          "validation_metric": 0.0036414291244000196,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.522344630000589
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.04456206306349486,
          "train_loss": 0.04430744796991348,
          "validation_loss": 0.03626619279384613,
          "train_metric": 0.0044624838046729565,
          "validation_metric": 0.0036525980103760958,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.538903192000362
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.04433649219572544,
          "train_loss": 0.044131141155958176,
          "validation_loss": 0.03587932512164116,
          "train_metric": 0.0044447267428040504,
          "validation_metric": 0.0036136340349912643,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.555726870001308
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.044433753355406225,
          "train_loss": 0.04391010105609894,
          "validation_loss": 0.03576716035604477,
          "train_metric": 0.004422464873641729,
          "validation_metric": 0.0036023377906531096,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.573571341001298
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.043877648655325174,
          "train_loss": 0.04378734156489372,
          "validation_loss": 0.03582132235169411,
          "train_metric": 0.004410101100802422,
          "validation_metric": 0.0036077918484807014,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.590158976001476
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.043807267560623586,
          "train_loss": 0.043587855994701385,
          "validation_loss": 0.035559698939323425,
          "train_metric": 0.0043900092132389545,
          "validation_metric": 0.0035814421717077494,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.607426939001016
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.043617622228339314,
          "train_loss": 0.04342929273843765,
          "validation_loss": 0.035476502031087875,
          "train_metric": 0.004374039359390736,
          "validation_metric": 0.0035730632953345776,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.624625561000357
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.043427959200926125,
          "train_loss": 0.04326856881380081,
          "validation_loss": 0.0353538952767849,
          "train_metric": 0.004357851576060057,
          "validation_metric": 0.0035607146564871073,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.642457835001551
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.04329990327823907,
          "train_loss": 0.04311525821685791,
          "validation_loss": 0.035193927586078644,
          "train_metric": 0.004342410713434219,
          "validation_metric": 0.003544603241607547,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.660386792000281
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.04327343846671283,
          "train_loss": 0.04299384355545044,
          "validation_loss": 0.03521713986992836,
          "train_metric": 0.004330182448029518,
          "validation_metric": 0.0035469410941004753,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.677371755000422
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.043089765356853604,
          "train_loss": 0.042829252779483795,
          "validation_loss": 0.03494898974895477,
          "train_metric": 0.00431360537186265,
          "validation_metric": 0.0035199341364204884,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.694802094001716
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.04290728527121246,
          "train_loss": 0.04264739900827408,
          "validation_loss": 0.03484678268432617,
          "train_metric": 0.004295289516448975,
          "validation_metric": 0.00350963999517262,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.712578875001782
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.04272519645746797,
          "train_loss": 0.04250921681523323,
          "validation_loss": 0.03468656539916992,
          "train_metric": 0.004281373228877783,
          "validation_metric": 0.003493503900244832,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.730292385000212
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.04260211775545031,
          "train_loss": 0.04232204705476761,
          "validation_loss": 0.03460671007633209,
          "train_metric": 0.004262521862983704,
          "validation_metric": 0.00348546146415174,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.747799243001282
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0423372567165643,
          "train_loss": 0.04216494783759117,
          "validation_loss": 0.034487947821617126,
          "train_metric": 0.004246699158102274,
          "validation_metric": 0.003473500022664666,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.764982388000135
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.04224193876143545,
          "train_loss": 0.04200681298971176,
          "validation_loss": 0.034357503056526184,
          "train_metric": 0.004230772610753775,
          "validation_metric": 0.0034603618551045656,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.782374879001509
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.04201902693603188,
          "train_loss": 0.04185960069298744,
          "validation_loss": 0.03431570529937744,
          "train_metric": 0.004215945955365896,
          "validation_metric": 0.0034561518114060163,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.799795398001152
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.041940008639357984,
          "train_loss": 0.041739583015441895,
          "validation_loss": 0.034273792058229446,
          "train_metric": 0.004203857854008675,
          "validation_metric": 0.0034519308246672153,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.817671735001568
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.041788994683884084,
          "train_loss": 0.04161227494478226,
          "validation_loss": 0.03394753113389015,
          "train_metric": 0.004191035870462656,
          "validation_metric": 0.003419071203097701,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.834661396000229
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.04163356253411621,
          "train_loss": 0.04139441251754761,
          "validation_loss": 0.033842943608760834,
          "train_metric": 0.004169093910604715,
          "validation_metric": 0.003408537246286869,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.852469251000002
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.04140104097314179,
          "train_loss": 0.041318830102682114,
          "validation_loss": 0.03397144377231598,
          "train_metric": 0.004161480814218521,
          "validation_metric": 0.0034214789047837257,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.967613504000838
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.04129871755139902,
          "train_loss": 0.04108458757400513,
          "validation_loss": 0.03361250087618828,
          "train_metric": 0.004137889016419649,
          "validation_metric": 0.0033853277564048767,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 4.985587270000906
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.041382950148545206,
          "train_loss": 0.040914386510849,
          "validation_loss": 0.033490508794784546,
          "train_metric": 0.004120747558772564,
          "validation_metric": 0.003373041283339262,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.003337875001307
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.040964833344332874,
          "train_loss": 0.040794845670461655,
          "validation_loss": 0.033529918640851974,
          "train_metric": 0.004108707420527935,
          "validation_metric": 0.0033770103473216295,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.020013992001623
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.040967336972244084,
          "train_loss": 0.04061376303434372,
          "validation_loss": 0.03334326297044754,
          "train_metric": 0.004090470261871815,
          "validation_metric": 0.0033582113683223724,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.03708167700097
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.04060949000995606,
          "train_loss": 0.04045633226633072,
          "validation_loss": 0.033213548362255096,
          "train_metric": 0.004074614029377699,
          "validation_metric": 0.00334514700807631,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.054111738001666
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.040831277961842716,
          "train_loss": 0.040327902883291245,
          "validation_loss": 0.03300703689455986,
          "train_metric": 0.00406167795881629,
          "validation_metric": 0.0033243480138480663,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.071349178000673
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.040302611072547734,
          "train_loss": 0.04024401679635048,
          "validation_loss": 0.033151429146528244,
          "train_metric": 0.00405323039740324,
          "validation_metric": 0.0033388910815119743,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.088290354000492
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.04025426192674786,
          "train_loss": 0.03999355435371399,
          "validation_loss": 0.032804105430841446,
          "train_metric": 0.004028004594147205,
          "validation_metric": 0.0033039094414561987,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.10652861500057
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.040024701156653464,
          "train_loss": 0.039837293326854706,
          "validation_loss": 0.03273894265294075,
          "train_metric": 0.004012266639620066,
          "validation_metric": 0.003297346644103527,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.1246235400012665
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.03986134333536029,
          "train_loss": 0.03968058526515961,
          "validation_loss": 0.03257138282060623,
          "train_metric": 0.00399648305028677,
          "validation_metric": 0.0032804706133902073,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.14219982800023
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.03968814457766712,
          "train_loss": 0.039532747119665146,
          "validation_loss": 0.03244894742965698,
          "train_metric": 0.003981593064963818,
          "validation_metric": 0.0032681392040103674,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.160724307001146
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.03960691089741886,
          "train_loss": 0.03941961005330086,
          "validation_loss": 0.03229254484176636,
          "train_metric": 0.0039701987989246845,
          "validation_metric": 0.0032523865811526775,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.178743962000226
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.03947937162593007,
          "train_loss": 0.03923317790031433,
          "validation_loss": 0.03228920325636864,
          "train_metric": 0.003951421473175287,
          "validation_metric": 0.003252050606533885,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.196698041001582
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.03919338167179376,
          "train_loss": 0.03907470032572746,
          "validation_loss": 0.03209328651428223,
          "train_metric": 0.0039354609325528145,
          "validation_metric": 0.003232318442314863,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.21443821000139
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.039251736365258694,
          "train_loss": 0.03891627490520477,
          "validation_loss": 0.03202260285615921,
          "train_metric": 0.0039195045828819275,
          "validation_metric": 0.0032251994125545025,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.232459706001464
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.03889142849948257,
          "train_loss": 0.03877364844083786,
          "validation_loss": 0.031914953142404556,
          "train_metric": 0.003905140096321702,
          "validation_metric": 0.003214356955140829,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.249926812000922
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.03883343713823706,
          "train_loss": 0.038617994636297226,
          "validation_loss": 0.03176802769303322,
          "train_metric": 0.003889462910592556,
          "validation_metric": 0.0031995594035834074,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.268140933001632
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.03856127068866044,
          "train_loss": 0.038471344858407974,
          "validation_loss": 0.03168685361742973,
          "train_metric": 0.003874692600220442,
          "validation_metric": 0.0031913835555315018,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.2868078840001544
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.03854372981004417,
          "train_loss": 0.038322001695632935,
          "validation_loss": 0.03149150684475899,
          "train_metric": 0.003859651740640402,
          "validation_metric": 0.00317170936614275,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.305688376000035
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0384409703547135,
          "train_loss": 0.038163166493177414,
          "validation_loss": 0.031448569148778915,
          "train_metric": 0.0038436544127762318,
          "validation_metric": 0.003167384769767523,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.3414234000010765
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.038568872259929776,
          "train_loss": 0.03806221857666969,
          "validation_loss": 0.03141385316848755,
          "train_metric": 0.0038334871642291546,
          "validation_metric": 0.0031638885848224163,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.38513425400015
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.038401618832722306,
          "train_loss": 0.03792383521795273,
          "validation_loss": 0.031125018373131752,
          "train_metric": 0.003819549921900034,
          "validation_metric": 0.0031347977928817272,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.403052900001057
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.037911965628154576,
          "train_loss": 0.03776305541396141,
          "validation_loss": 0.031184818595647812,
          "train_metric": 0.0038033565506339073,
          "validation_metric": 0.003140820888802409,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.419832644000053
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.03801422147080302,
          "train_loss": 0.037652552127838135,
          "validation_loss": 0.03110775165259838,
          "train_metric": 0.0037922270130366087,
          "validation_metric": 0.00313305901363492,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.436925976000566
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.03762285830453038,
          "train_loss": 0.03754624351859093,
          "validation_loss": 0.03078511171042919,
          "train_metric": 0.0037815202958881855,
          "validation_metric": 0.003100563772022724,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.453989149000336
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.037461596773937345,
          "train_loss": 0.03729165717959404,
          "validation_loss": 0.030765382573008537,
          "train_metric": 0.003755879122763872,
          "validation_metric": 0.0030985765624791384,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.47172445400065
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.03733016853220761,
          "train_loss": 0.037151288241147995,
          "validation_loss": 0.030677178874611855,
          "train_metric": 0.0037417414132505655,
          "validation_metric": 0.003089692909270525,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.490422861001207
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.037149643001612276,
          "train_loss": 0.03699636086821556,
          "validation_loss": 0.03051871433854103,
          "train_metric": 0.0037261382676661015,
          "validation_metric": 0.003073733067139983,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.5081787390008685
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.03712608094792813,
          "train_loss": 0.03692048788070679,
          "validation_loss": 0.03035522624850273,
          "train_metric": 0.003718496300280094,
          "validation_metric": 0.0030572672840207815,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.5260537570011365
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.03704832377843559,
          "train_loss": 0.03671412542462349,
          "validation_loss": 0.030318107455968857,
          "train_metric": 0.003697712207213044,
          "validation_metric": 0.0030535287223756313,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.543608916001176
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.036742684547789395,
          "train_loss": 0.036557797342538834,
          "validation_loss": 0.030172521248459816,
          "train_metric": 0.003681967733427882,
          "validation_metric": 0.0030388659797608852,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.561461613000574
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.03655797126702964,
          "train_loss": 0.03643607348203659,
          "validation_loss": 0.030014609917998314,
          "train_metric": 0.003669708501547575,
          "validation_metric": 0.0030229617841541767,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.578996628000823
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.036441274103708565,
          "train_loss": 0.03628736361861229,
          "validation_loss": 0.029993927106261253,
          "train_metric": 0.0036547300405800343,
          "validation_metric": 0.003020878415554762,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.596676347000539
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.036477713903877884,
          "train_loss": 0.036138031631708145,
          "validation_loss": 0.02979111857712269,
          "train_metric": 0.0036396903451532125,
          "validation_metric": 0.003000452183187008,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.614361078000002
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.03612482454627752,
          "train_loss": 0.036034781485795975,
          "validation_loss": 0.029823901131749153,
          "train_metric": 0.003629291197285056,
          "validation_metric": 0.0030037539545446634,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.631029780000972
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.03603033139370382,
          "train_loss": 0.03585970401763916,
          "validation_loss": 0.029634622856974602,
          "train_metric": 0.0036116582341492176,
          "validation_metric": 0.002984690712764859,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.7437349160009035
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.035897003079298884,
          "train_loss": 0.03574910759925842,
          "validation_loss": 0.029450852423906326,
          "train_metric": 0.003600519150495529,
          "validation_metric": 0.002966182306408882,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.762121912001021
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.03580060019157827,
          "train_loss": 0.035586029291152954,
          "validation_loss": 0.029361193999648094,
          "train_metric": 0.0035840943455696106,
          "validation_metric": 0.002957151737064123,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.779847180001525
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.035608359903562814,
          "train_loss": 0.03543609008193016,
          "validation_loss": 0.029275886714458466,
          "train_metric": 0.0035689936485141516,
          "validation_metric": 0.002948560519143939,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.798147453000638
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.03542902518529445,
          "train_loss": 0.0353236198425293,
          "validation_loss": 0.029250070452690125,
          "train_metric": 0.003557665506377816,
          "validation_metric": 0.002945960033684969,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.816026263000822
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.035297319816891104,
          "train_loss": 0.035170480608940125,
          "validation_loss": 0.02908402867615223,
          "train_metric": 0.003542242106050253,
          "validation_metric": 0.0029292369727045298,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.83452576900163
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.03524905553786084,
          "train_loss": 0.03505058214068413,
          "validation_loss": 0.028928512707352638,
          "train_metric": 0.003530166344717145,
          "validation_metric": 0.002913573756814003,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.8521315030011465
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.0351979824481532,
          "train_loss": 0.0349198579788208,
          "validation_loss": 0.028828438371419907,
          "train_metric": 0.0035169997718185186,
          "validation_metric": 0.002903494518250227,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.8695380090011895
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.03489759913645685,
          "train_loss": 0.03484275937080383,
          "validation_loss": 0.028893711045384407,
          "train_metric": 0.0035092351026833057,
          "validation_metric": 0.0029100687243044376,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.885891872001594
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.03479310043621808,
          "train_loss": 0.03463776782155037,
          "validation_loss": 0.02864881232380867,
          "train_metric": 0.003488589311018586,
          "validation_metric": 0.0028854035772383213,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.903580771000634
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.03480511764064431,
          "train_loss": 0.03455234691500664,
          "validation_loss": 0.028494907543063164,
          "train_metric": 0.0034799857530742884,
          "validation_metric": 0.002869902877137065,
          "lr_first": 0.0001,
          "lr_last": 0.0001,
          "seconds": 5.9217804960007925
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 3840
      },
      "best_validation_metric": 0.002869902877137065,
      "config": {
        "label": "LR 0.0001",
        "batch": 64,
        "lr": 0.0001,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.05357147447230723
    },
    "32d8a0bb9dc3_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.5785819953307509,
          "train_loss": 0.24066151678562164,
          "validation_loss": 0.15968576073646545,
          "train_metric": 0.024238549172878265,
          "validation_metric": 0.016082964837551117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.018309380999198765
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.20281902188435197,
          "train_loss": 0.14825578033924103,
          "validation_loss": 0.12119951099157333,
          "train_metric": 0.014931779354810715,
          "validation_metric": 0.012206770479679108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03625433900015196
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13218980515375733,
          "train_loss": 0.11559955775737762,
          "validation_loss": 0.09988148510456085,
          "train_metric": 0.011642763391137123,
          "validation_metric": 0.010059698484838009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05579813299846137
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10902172443456948,
          "train_loss": 0.09608157724142075,
          "validation_loss": 0.08105382323265076,
          "train_metric": 0.009676986373960972,
          "validation_metric": 0.008163443766534328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07352632499896572
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08852994372136891,
          "train_loss": 0.07870391011238098,
          "validation_loss": 0.06390558928251266,
          "train_metric": 0.00792676955461502,
          "validation_metric": 0.0064363377168774605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09128805699947407
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.0738776980433613,
          "train_loss": 0.06744672358036041,
          "validation_loss": 0.05580252408981323,
          "train_metric": 0.006792987231165171,
          "validation_metric": 0.005620225332677364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10955410199858306
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.0662799112033099,
          "train_loss": 0.062105804681777954,
          "validation_loss": 0.051005952060222626,
          "train_metric": 0.006255070213228464,
          "validation_metric": 0.0051371329464018345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1280609639998147
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06061952211894095,
          "train_loss": 0.058522891253232956,
          "validation_loss": 0.04748249426484108,
          "train_metric": 0.005894211120903492,
          "validation_metric": 0.0047822631895542145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14528609899934963
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0586743209278211,
          "train_loss": 0.05518710985779762,
          "validation_loss": 0.04696771502494812,
          "train_metric": 0.00555824302136898,
          "validation_metric": 0.0047304159961640835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1631110849994002
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05452180642168969,
          "train_loss": 0.05012526735663414,
          "validation_loss": 0.04216490313410759,
          "train_metric": 0.005048433784395456,
          "validation_metric": 0.004246694967150688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18098375999943528
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0489687561057508,
          "train_loss": 0.0455508716404438,
          "validation_loss": 0.0387922078371048,
          "train_metric": 0.004587717354297638,
          "validation_metric": 0.003907009027898312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19895195699973556
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.045328833162784576,
          "train_loss": 0.04298621788620949,
          "validation_loss": 0.03603599593043327,
          "train_metric": 0.004329414572566748,
          "validation_metric": 0.0036294138990342617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21663078099845734
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.040962925646454096,
          "train_loss": 0.03840823099017143,
          "validation_loss": 0.03274441510438919,
          "train_metric": 0.003868336323648691,
          "validation_metric": 0.003297897521406412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23480337799992412
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.035591799300163984,
          "train_loss": 0.03237031772732735,
          "validation_loss": 0.027760149911046028,
          "train_metric": 0.003260219469666481,
          "validation_metric": 0.0027959002181887627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.252499109999917
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.034175259876064956,
          "train_loss": 0.030200820416212082,
          "validation_loss": 0.02461184561252594,
          "train_metric": 0.0030417158268392086,
          "validation_metric": 0.0024788149166852236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2699393000002601
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.029044358525425196,
          "train_loss": 0.02824326977133751,
          "validation_loss": 0.02484940178692341,
          "train_metric": 0.0028445585630834103,
          "validation_metric": 0.0025027410592883825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2870993809992797
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.027968548878561705,
          "train_loss": 0.026699528098106384,
          "validation_loss": 0.022850092500448227,
          "train_metric": 0.0026890786830335855,
          "validation_metric": 0.0023013774771243334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3051657409996551
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029929061303846538,
          "train_loss": 0.028845902532339096,
          "validation_loss": 0.023522071540355682,
          "train_metric": 0.002905253553763032,
          "validation_metric": 0.002369056921452284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32290090899914503
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.0253515514486935,
          "train_loss": 0.022620845586061478,
          "validation_loss": 0.019433408975601196,
          "train_metric": 0.0022782885935157537,
          "validation_metric": 0.001957261934876442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3412675359995774
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02046690188581124,
          "train_loss": 0.019571326673030853,
          "validation_loss": 0.016824472695589066,
          "train_metric": 0.0019711521454155445,
          "validation_metric": 0.00169449916575104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35895646599965403
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.021430058928672224,
          "train_loss": 0.019809599965810776,
          "validation_loss": 0.01710662618279457,
          "train_metric": 0.001995150465518236,
          "validation_metric": 0.0017229169607162476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37639894299900334
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.019933807488996536,
          "train_loss": 0.018322203308343887,
          "validation_loss": 0.015159186907112598,
          "train_metric": 0.0018453451339155436,
          "validation_metric": 0.0015267777489498258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3936508539991337
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.0184865053743124,
          "train_loss": 0.018727105110883713,
          "validation_loss": 0.016270481050014496,
          "train_metric": 0.0018861254211515188,
          "validation_metric": 0.001638703397475183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4107961409990821
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019489634316414595,
          "train_loss": 0.018848121166229248,
          "validation_loss": 0.015607197768986225,
          "train_metric": 0.0018983136396855116,
          "validation_metric": 0.0015718997456133366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4279762779988232
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.025991787144448608,
          "train_loss": 0.022055301815271378,
          "validation_loss": 0.01751544512808323,
          "train_metric": 0.0022213291376829147,
          "validation_metric": 0.0017640916630625725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44517856999846117
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.020158647326752543,
          "train_loss": 0.019878527149558067,
          "validation_loss": 0.015789136290550232,
          "train_metric": 0.0020020923111587763,
          "validation_metric": 0.0015902240993455052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.461334800998884
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01832235194160603,
          "train_loss": 0.015337939374148846,
          "validation_loss": 0.013187260366976261,
          "train_metric": 0.0015447810292243958,
          "validation_metric": 0.001328172511421144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4787587999999232
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015697864320827648,
          "train_loss": 0.01441369578242302,
          "validation_loss": 0.012476164847612381,
          "train_metric": 0.0014516946393996477,
          "validation_metric": 0.0012565534561872482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4962098359992524
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.01796175292111002,
          "train_loss": 0.015583343803882599,
          "validation_loss": 0.012778116390109062,
          "train_metric": 0.0015694972826167941,
          "validation_metric": 0.0012869652127847075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5131836489999841
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.017315161181613803,
          "train_loss": 0.026303283870220184,
          "validation_loss": 0.022693904116749763,
          "train_metric": 0.0026491701137274504,
          "validation_metric": 0.002285646740347147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5297965869995096
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.018650115351192653,
          "train_loss": 0.015033909119665623,
          "validation_loss": 0.012586742639541626,
          "train_metric": 0.0015141601907089353,
          "validation_metric": 0.0012676906771957874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5467597810002189
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01525422916165553,
          "train_loss": 0.012996863573789597,
          "validation_loss": 0.011027900502085686,
          "train_metric": 0.0013089965796098113,
          "validation_metric": 0.0011106899473816156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6821172069994645
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.017277867038501427,
          "train_loss": 0.018468111753463745,
          "validation_loss": 0.01605597883462906,
          "train_metric": 0.0018600407056510448,
          "validation_metric": 0.001617099391296506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6992667769991385
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.017836909275501966,
          "train_loss": 0.015464769676327705,
          "validation_loss": 0.013479636050760746,
          "train_metric": 0.001557554816827178,
          "validation_metric": 0.0013576196506619453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7153285089989367
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014926508301869035,
          "train_loss": 0.01525525376200676,
          "validation_loss": 0.012201250530779362,
          "train_metric": 0.0015364533755928278,
          "validation_metric": 0.00122886523604393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7322110009990865
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.013606283508124761,
          "train_loss": 0.01189349964261055,
          "validation_loss": 0.010086850263178349,
          "train_metric": 0.0011978696566075087,
          "validation_metric": 0.0010159105295315385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7639562459989975
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013865286426153034,
          "train_loss": 0.01421459298580885,
          "validation_loss": 0.012337259016931057,
          "train_metric": 0.0014316417509689927,
          "validation_metric": 0.00124256347771734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8100025489984546
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.013906719570513815,
          "train_loss": 0.01308034174144268,
          "validation_loss": 0.011206195689737797,
          "train_metric": 0.0013174042105674744,
          "validation_metric": 0.0011286470107734203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8502410459987004
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.015102692559594288,
          "train_loss": 0.012156970798969269,
          "validation_loss": 0.009952702559530735,
          "train_metric": 0.0012244055978953838,
          "validation_metric": 0.00100239971652627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8928852779990848
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.011600760830333456,
          "train_loss": 0.01074554119259119,
          "validation_loss": 0.009170734323561192,
          "train_metric": 0.0010822515469044447,
          "validation_metric": 0.0009236427722498775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9358432939989143
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011461123736808077,
          "train_loss": 0.011186208575963974,
          "validation_loss": 0.008815637789666653,
          "train_metric": 0.001126633957028389,
          "validation_metric": 0.000887878704816103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9869133319989487
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.012731426904792897,
          "train_loss": 0.011028323322534561,
          "validation_loss": 0.008532376028597355,
          "train_metric": 0.0011107323225587606,
          "validation_metric": 0.0008593496750108898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0086293509993993
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011592802344239317,
          "train_loss": 0.01758863590657711,
          "validation_loss": 0.0152521887794137,
          "train_metric": 0.0017714631976559758,
          "validation_metric": 0.0015361446421593428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0258270180002
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014004348166054115,
          "train_loss": 0.011498655192553997,
          "validation_loss": 0.009649231098592281,
          "train_metric": 0.0011581024155020714,
          "validation_metric": 0.0009718352230265737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.043349858999136
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.011971019470365718,
          "train_loss": 0.010247869417071342,
          "validation_loss": 0.008546342141926289,
          "train_metric": 0.0010321278823539615,
          "validation_metric": 0.0008607562049292028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.059912261998761
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010324565839255229,
          "train_loss": 0.009410005994141102,
          "validation_loss": 0.008049655705690384,
          "train_metric": 0.0009477414423599839,
          "validation_metric": 0.0008107318426482379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0773429529999703
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0099878313485533,
          "train_loss": 0.009574130177497864,
          "validation_loss": 0.008396321907639503,
          "train_metric": 0.0009642713703215122,
          "validation_metric": 0.0008456468349322677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0947061379993102
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010388583701569587,
          "train_loss": 0.010846927762031555,
          "validation_loss": 0.009024742990732193,
          "train_metric": 0.0010924628004431725,
          "validation_metric": 0.0009089390514418483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.111331218999112
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.011729866499081254,
          "train_loss": 0.008839637972414494,
          "validation_loss": 0.007451624143868685,
          "train_metric": 0.0008902959525585175,
          "validation_metric": 0.0007505002431571484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1288038179991418
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009075396141270176,
          "train_loss": 0.008698005229234695,
          "validation_loss": 0.007412043400108814,
          "train_metric": 0.0008760312921367586,
          "validation_metric": 0.0007465138914994895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1462222869995458
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008586802636273205,
          "train_loss": 0.009239615872502327,
          "validation_loss": 0.00846143253147602,
          "train_metric": 0.0009305804269388318,
          "validation_metric": 0.000852204451803118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1630175829996006
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.009399648464750499,
          "train_loss": 0.008207714185118675,
          "validation_loss": 0.007094361819326878,
          "train_metric": 0.0008266508812084794,
          "validation_metric": 0.0007145180716179311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1803962329995557
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.009622707089874893,
          "train_loss": 0.008783034980297089,
          "validation_loss": 0.007847893983125687,
          "train_metric": 0.0008845950360409915,
          "validation_metric": 0.0007904112571850419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.197921006998513
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009374020868563093,
          "train_loss": 0.008376436308026314,
          "validation_loss": 0.007370653096586466,
          "train_metric": 0.0008436439675278962,
          "validation_metric": 0.0007423452916555107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.214429552999718
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008409974529058672,
          "train_loss": 0.008557048626244068,
          "validation_loss": 0.006746719125658274,
          "train_metric": 0.0008618345018476248,
          "validation_metric": 0.0006795047665946186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2327947469984792
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.008126632194034755,
          "train_loss": 0.007593871559947729,
          "validation_loss": 0.006299670785665512,
          "train_metric": 0.000764826952945441,
          "validation_metric": 0.0006344797438941896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.251010651998513
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008103969594230875,
          "train_loss": 0.007627340033650398,
          "validation_loss": 0.006149530876427889,
          "train_metric": 0.0007681977585889399,
          "validation_metric": 0.0006193582667037845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2694520439999906
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007666467514354736,
          "train_loss": 0.007350016385316849,
          "validation_loss": 0.0060503422282636166,
          "train_metric": 0.0007402666378766298,
          "validation_metric": 0.0006093683768995106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2878364989992406
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.008778885996434838,
          "train_loss": 0.007884827442467213,
          "validation_loss": 0.007127765100449324,
          "train_metric": 0.0007941309013403952,
          "validation_metric": 0.0007178823580034077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3052664159986307
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007124831900000572,
          "train_loss": 0.008507896214723587,
          "validation_loss": 0.006483363453298807,
          "train_metric": 0.0008568841149099171,
          "validation_metric": 0.0006529804668389261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.321794688999944
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.00802242913050577,
          "train_loss": 0.0072455033659935,
          "validation_loss": 0.005634790752083063,
          "train_metric": 0.000729740597307682,
          "validation_metric": 0.0005675156135112047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3390371059995232
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007499659288441762,
          "train_loss": 0.006905286107212305,
          "validation_loss": 0.005780530162155628,
          "train_metric": 0.0006954751443117857,
          "validation_metric": 0.0005821938975714147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3561101249997591
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006983088416745886,
          "train_loss": 0.00676384475082159,
          "validation_loss": 0.005742666777223349,
          "train_metric": 0.0006812297506257892,
          "validation_metric": 0.0005783804226666689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.373353238999698
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007186421637015883,
          "train_loss": 0.006360976025462151,
          "validation_loss": 0.0053828442469239235,
          "train_metric": 0.0006406542379409075,
          "validation_metric": 0.000542140391189605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3908619369994994
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008146895095705986,
          "train_loss": 0.00710770720615983,
          "validation_loss": 0.0064037274569272995,
          "train_metric": 0.0007158622611314058,
          "validation_metric": 0.0006449599750339985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4082301709986496
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0070507670170627534,
          "train_loss": 0.006784606724977493,
          "validation_loss": 0.005462770350277424,
          "train_metric": 0.0006833208026364446,
          "validation_metric": 0.0005501902778632939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.425167072999102
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007892993970017415,
          "train_loss": 0.007619489450007677,
          "validation_loss": 0.005710892844945192,
          "train_metric": 0.0007674070075154305,
          "validation_metric": 0.0005751801654696465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4422443059993384
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0075215069082332775,
          "train_loss": 0.008999602869153023,
          "validation_loss": 0.007687784265726805,
          "train_metric": 0.0009064070181921124,
          "validation_metric": 0.0007742854650132358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.559662360999937
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.008329692267579958,
          "train_loss": 0.006699476856738329,
          "validation_loss": 0.005957436747848988,
          "train_metric": 0.000674746697768569,
          "validation_metric": 0.0006000111461617053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5773548090000986
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006823713432822842,
          "train_loss": 0.006718658842146397,
          "validation_loss": 0.005331420339643955,
          "train_metric": 0.0006766787264496088,
          "validation_metric": 0.0005369611317291856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.611699975999727
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0063610283978050575,
          "train_loss": 0.006560879293829203,
          "validation_loss": 0.005013317335397005,
          "train_metric": 0.0006607878021895885,
          "validation_metric": 0.0005049229366704822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6557318439990922
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006628050803556107,
          "train_loss": 0.005550465080887079,
          "validation_loss": 0.00450581219047308,
          "train_metric": 0.0005590225337073207,
          "validation_metric": 0.00045380889787338674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7030151679991832
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006065987341571599,
          "train_loss": 0.006213136482983828,
          "validation_loss": 0.004682971630245447,
          "train_metric": 0.000625764369033277,
          "validation_metric": 0.0004716518160421401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7451380339989555
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0066800750500988215,
          "train_loss": 0.007923943921923637,
          "validation_loss": 0.00671512121334672,
          "train_metric": 0.0007980706868693233,
          "validation_metric": 0.0006763224373571575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7655111690000922
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.008286909156595357,
          "train_loss": 0.005879192613065243,
          "validation_loss": 0.005012104287743568,
          "train_metric": 0.0005921307601965964,
          "validation_metric": 0.0005048008169978857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7827975349991902
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.008240004448452964,
          "train_loss": 0.006389837712049484,
          "validation_loss": 0.005709755700081587,
          "train_metric": 0.0006435610703192651,
          "validation_metric": 0.0005750657292082906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.799601986000198
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007031992237898521,
          "train_loss": 0.0061054527759552,
          "validation_loss": 0.005275707691907883,
          "train_metric": 0.000614918943028897,
          "validation_metric": 0.0005313499132171273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8168210229996475
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006232404222828336,
          "train_loss": 0.005883560515940189,
          "validation_loss": 0.004626329056918621,
          "train_metric": 0.0005925706936977804,
          "validation_metric": 0.00046594697050750256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8334380889991735
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006698730117932428,
          "train_loss": 0.00830888468772173,
          "validation_loss": 0.005257878918200731,
          "train_metric": 0.0008368405397050083,
          "validation_metric": 0.0005295543232932687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8510676850000891
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.007175867009209469,
          "train_loss": 0.005164492409676313,
          "validation_loss": 0.0041005248203873634,
          "train_metric": 0.0005201487801969051,
          "validation_metric": 0.00041298987343907356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8686532359988632
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005006608822441194,
          "train_loss": 0.004682530649006367,
          "validation_loss": 0.0038298754952847958,
          "train_metric": 0.00047160725807771087,
          "validation_metric": 0.0003857310803141445,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.886314893999952
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0047530331066809595,
          "train_loss": 0.004614436533302069,
          "validation_loss": 0.0037280500400811434,
          "train_metric": 0.0004647492023650557,
          "validation_metric": 0.0003754755889531225,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9037822289992619
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.004831954880501144,
          "train_loss": 0.004826447926461697,
          "validation_loss": 0.003774649929255247,
          "train_metric": 0.00048610224621370435,
          "validation_metric": 0.0003801689308602363,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.921270425998955
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.004723478225059807,
          "train_loss": 0.004758256953209639,
          "validation_loss": 0.0041597019881010056,
          "train_metric": 0.0004792342078872025,
          "validation_metric": 0.00041894998867064714,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9381873369984532
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.004544722309219651,
          "train_loss": 0.004982247017323971,
          "validation_loss": 0.0037444012705236673,
          "train_metric": 0.0005017936346121132,
          "validation_metric": 0.00037712237099185586,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9547276409994083
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005111032500280999,
          "train_loss": 0.005444197915494442,
          "validation_loss": 0.004954630974680185,
          "train_metric": 0.0005483196582645178,
          "validation_metric": 0.000499012297950685,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9713958779993845
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.004937594123475719,
          "train_loss": 0.004555682651698589,
          "validation_loss": 0.0038721475284546614,
          "train_metric": 0.00045883169514127076,
          "validation_metric": 0.0003899885050486773,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 1.9916192079999746
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.004684804502176121,
          "train_loss": 0.004587079398334026,
          "validation_loss": 0.0036314702592790127,
          "train_metric": 0.0004619938845280558,
          "validation_metric": 0.00036574839032255113,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0055083939987526
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004726370840216987,
          "train_loss": 0.004458734765648842,
          "validation_loss": 0.003617749083787203,
          "train_metric": 0.0004490674764383584,
          "validation_metric": 0.00036436639493331313,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0192739689991868
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004901242042251397,
          "train_loss": 0.0044781980104744434,
          "validation_loss": 0.0036862227134406567,
          "train_metric": 0.0004510277067311108,
          "validation_metric": 0.00037126283859834075,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.0325677179989725
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004720698798337253,
          "train_loss": 0.004403026774525642,
          "validation_loss": 0.003754256758838892,
          "train_metric": 0.00044345666537992656,
          "validation_metric": 0.00037811495712958276,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.047575038999639
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004310628821258433,
          "train_loss": 0.004451462998986244,
          "validation_loss": 0.003464131848886609,
          "train_metric": 0.0004483350785449147,
          "validation_metric": 0.00034889468224719167,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.061757024999679
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004575315018882975,
          "train_loss": 0.00425539817661047,
          "validation_loss": 0.0035492475144565105,
          "train_metric": 0.00042858810047619045,
          "validation_metric": 0.0003574672155082226,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.075220961998639
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.00452937743102666,
          "train_loss": 0.004221615381538868,
          "validation_loss": 0.0034551930148154497,
          "train_metric": 0.0004251856589689851,
          "validation_metric": 0.00034799438435584307,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.087867058999109
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004493092026677914,
          "train_loss": 0.00483277253806591,
          "validation_loss": 0.0036720172502100468,
          "train_metric": 0.0004867391544394195,
          "validation_metric": 0.0003698321233969182,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1011274619995675
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004618228325853124,
          "train_loss": 0.004230763763189316,
          "validation_loss": 0.0034471999388188124,
          "train_metric": 0.00042610702803358436,
          "validation_metric": 0.00034718934330157936,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.114791920999778
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.004309263611503411,
          "train_loss": 0.004273563623428345,
          "validation_loss": 0.0033762678503990173,
          "train_metric": 0.00043041774188168347,
          "validation_metric": 0.0003400453133508563,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.128472929000054
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.00461614959203871,
          "train_loss": 0.004263418726623058,
          "validation_loss": 0.0036157681606709957,
          "train_metric": 0.000429395935498178,
          "validation_metric": 0.0003641669172793627,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1414465620000556
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0054794011084595695,
          "train_loss": 0.004880278371274471,
          "validation_loss": 0.004295249003916979,
          "train_metric": 0.0004915237077511847,
          "validation_metric": 0.00043260178063064814,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.153839362999861
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.005487519665621221,
          "train_loss": 0.004076802637428045,
          "validation_loss": 0.0033675276208668947,
          "train_metric": 0.0004106006817892194,
          "validation_metric": 0.0003391650679986924,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.168310435999956
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.004555686784442514,
          "train_loss": 0.004284693859517574,
          "validation_loss": 0.0036846445873379707,
          "train_metric": 0.00043153867591172457,
          "validation_metric": 0.0003711039025802165,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.1821958019991143
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.004428768792422488,
          "train_loss": 0.00415347795933485,
          "validation_loss": 0.003527863882482052,
          "train_metric": 0.00041832315037027,
          "validation_metric": 0.0003553135320544243,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.195510817999093
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.004350576615252066,
          "train_loss": 0.004084523301571608,
          "validation_loss": 0.003460848471149802,
          "train_metric": 0.0004113782779313624,
          "validation_metric": 0.0003485639754217118,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.208216213999549
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004157324983680155,
          "train_loss": 0.003971362952142954,
          "validation_loss": 0.003251558169722557,
          "train_metric": 0.0003999811306130141,
          "validation_metric": 0.00032748503144830465,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.3708831679996365
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004175209855020512,
          "train_loss": 0.004431312903761864,
          "validation_loss": 0.0033320626243948936,
          "train_metric": 0.00044630561023950577,
          "validation_metric": 0.0003355931839905679,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.4029899849992944
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004625317051250022,
          "train_loss": 0.004035807680338621,
          "validation_loss": 0.003301678691059351,
          "train_metric": 0.00040647172136232257,
          "validation_metric": 0.0003325329744257033,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.4342637749996356
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.005357643487513997,
          "train_loss": 0.005911142099648714,
          "validation_loss": 0.005460995715111494,
          "train_metric": 0.0005953486543148756,
          "validation_metric": 0.0005500114639289677,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.465032979998796
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.004505643082666211,
          "train_loss": 0.0041570779867470264,
          "validation_loss": 0.0036136629059910774,
          "train_metric": 0.0004186857258901,
          "validation_metric": 0.0003639548958744854,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.4868284610001865
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.00414285555598326,
          "train_loss": 0.0040346537716686726,
          "validation_loss": 0.0031985845416784286,
          "train_metric": 0.0004063555388711393,
          "validation_metric": 0.00032214971724897623,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.5151138309993257
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0043376561588956974,
          "train_loss": 0.004382317420095205,
          "validation_loss": 0.0033771467860788107,
          "train_metric": 0.00044137099757790565,
          "validation_metric": 0.0003401338472031057,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.5345743660000153
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004649544935091399,
          "train_loss": 0.004025734029710293,
          "validation_loss": 0.003075607120990753,
          "train_metric": 0.00040545716183260083,
          "validation_metric": 0.0003097639128100127,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.5671760179993726
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.00445506906544324,
          "train_loss": 0.003900369396433234,
          "validation_loss": 0.0031657032668590546,
          "train_metric": 0.0003928309306502342,
          "validation_metric": 0.0003188379923813045,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.597701758999392
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004050074276165105,
          "train_loss": 0.003767502959817648,
          "validation_loss": 0.0030195522122085094,
          "train_metric": 0.00037944907671771944,
          "validation_metric": 0.0003041182062588632,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.6205845969998336
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004267545518814586,
          "train_loss": 0.003808323061093688,
          "validation_loss": 0.002971723210066557,
          "train_metric": 0.00038356034201569855,
          "validation_metric": 0.00029930108576081693,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.6540471449989127
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004638338701624889,
          "train_loss": 0.005681032780557871,
          "validation_loss": 0.0040354421362280846,
          "train_metric": 0.0005721728084608912,
          "validation_metric": 0.0004064349050167948,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.684960822998619
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0044425743544707075,
          "train_loss": 0.0037500052712857723,
          "validation_loss": 0.003042899537831545,
          "train_metric": 0.00037768681067973375,
          "validation_metric": 0.00030646967934444547,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.717201182998906
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003921757874195464,
          "train_loss": 0.0039705513045191765,
          "validation_loss": 0.0029558129608631134,
          "train_metric": 0.00039989937795326114,
          "validation_metric": 0.0002976986870635301,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.750441303000116
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0038853222722536884,
          "train_loss": 0.003649438265711069,
          "validation_loss": 0.0030386028811335564,
          "train_metric": 0.00036755806650035083,
          "validation_metric": 0.0003060369344893843,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.7829790389987465
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0038052260715630837,
          "train_loss": 0.0037370685022324324,
          "validation_loss": 0.0028323684819042683,
          "train_metric": 0.00037638386129401624,
          "validation_metric": 0.00028526579262688756,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.8061206189995573
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0037643134928657673,
          "train_loss": 0.003615944180637598,
          "validation_loss": 0.0029903582762926817,
          "train_metric": 0.0003641846706159413,
          "validation_metric": 0.00030117796268314123,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.836114830999577
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0039062385840225033,
          "train_loss": 0.004116878844797611,
          "validation_loss": 0.003594105364754796,
          "train_metric": 0.0004146369465161115,
          "validation_metric": 0.00036198514862917364,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.8584483609993185
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.003987392934504896,
          "train_loss": 0.0036326227709650993,
          "validation_loss": 0.003002345096319914,
          "train_metric": 0.00036586442729458213,
          "validation_metric": 0.000302385218674317,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.884546198998578
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.003662230883492157,
          "train_loss": 0.003590257605537772,
          "validation_loss": 0.003034493885934353,
          "train_metric": 0.00036159763112664223,
          "validation_metric": 0.00030562313622795045,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.9151653979988623
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.00366309021046618,
          "train_loss": 0.003448683535680175,
          "validation_loss": 0.002868000417947769,
          "train_metric": 0.00034733873326331377,
          "validation_metric": 0.000288854498649016,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.9378236359989387
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.003565052174963057,
          "train_loss": 0.003488078247755766,
          "validation_loss": 0.0027197387535125017,
          "train_metric": 0.0003513063711579889,
          "validation_metric": 0.0002739221090450883,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 2.9705383199998323
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0037210083537502214,
          "train_loss": 0.003615891095250845,
          "validation_loss": 0.003037869930267334,
          "train_metric": 0.0003641793446149677,
          "validation_metric": 0.000305963127175346,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.0018728819995886
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0035212177826906554,
          "train_loss": 0.00351353595033288,
          "validation_loss": 0.002690921537578106,
          "train_metric": 0.0003538704477250576,
          "validation_metric": 0.00027101978776045144,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.033789039000112
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.003711458179168403,
          "train_loss": 0.003405582392588258,
          "validation_loss": 0.0028012623079121113,
          "train_metric": 0.0003429977805353701,
          "validation_metric": 0.000282132881693542,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.058815226999286
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.004038293176563457,
          "train_loss": 0.004503526724874973,
          "validation_loss": 0.004038428422063589,
          "train_metric": 0.00045357871567830443,
          "validation_metric": 0.00040673575131222606,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.0872834399997373
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.004482574149733409,
          "train_loss": 0.003901484189555049,
          "validation_loss": 0.003346751444041729,
          "train_metric": 0.00039294318412430584,
          "validation_metric": 0.0003370725316926837,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.117472985999484
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.003631255094660446,
          "train_loss": 0.0034384557511657476,
          "validation_loss": 0.0029292767867445946,
          "train_metric": 0.0003463086613919586,
          "validation_metric": 0.0002950260532088578,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.13808434099883
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.003634557138866512,
          "train_loss": 0.004088644869625568,
          "validation_loss": 0.0035466260742396116,
          "train_metric": 0.0004117932985536754,
          "validation_metric": 0.0003572031855583191,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.155664017000163
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.003633753276517382,
          "train_loss": 0.0035659384448081255,
          "validation_loss": 0.002690419089049101,
          "train_metric": 0.00035914831096306443,
          "validation_metric": 0.0002709691761992872,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.1699414830000023
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.003543649047060171,
          "train_loss": 0.003349532838910818,
          "validation_loss": 0.0028550028800964355,
          "train_metric": 0.00033735265606082976,
          "validation_metric": 0.00028754540835507214,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.186438098999133
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.003504595995764248,
          "train_loss": 0.0032376162707805634,
          "validation_loss": 0.0026840772479772568,
          "train_metric": 0.0003260808880440891,
          "validation_metric": 0.00027033043443225324,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2121392019998893
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004163791782048065,
          "train_loss": 0.004283751826733351,
          "validation_loss": 0.0031648240983486176,
          "train_metric": 0.0004314437974244356,
          "validation_metric": 0.00031874951673671603,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2270924459990056
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003917385445674881,
          "train_loss": 0.0034126187674701214,
          "validation_loss": 0.002875839825719595,
          "train_metric": 0.0003437064297031611,
          "validation_metric": 0.00028964405646547675,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2397893689994817
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.00381513629690744,
          "train_loss": 0.004379733465611935,
          "validation_loss": 0.003951994236558676,
          "train_metric": 0.000441110780229792,
          "validation_metric": 0.0003980303881689906,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.2524512319996575
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0038485121258418076,
          "train_loss": 0.004417017567902803,
          "validation_loss": 0.00383564829826355,
          "train_metric": 0.00044486584374681115,
          "validation_metric": 0.0003863124584313482,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.266020719000153
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.003524625390127767,
          "train_loss": 0.0031914273276925087,
          "validation_loss": 0.00244126352481544,
          "train_metric": 0.00032142887357622385,
          "validation_metric": 0.0002458750968798995,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.423624270999426
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0032931721216300502,
          "train_loss": 0.0031335544772446156,
          "validation_loss": 0.002514241263270378,
          "train_metric": 0.0003156001039315015,
          "validation_metric": 0.00025322518195025623,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.468686718999379
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.003341185307363048,
          "train_loss": 0.0033401024993509054,
          "validation_loss": 0.0028360779397189617,
          "train_metric": 0.0003364028816577047,
          "validation_metric": 0.0002856393693946302,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.502999506999913
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0031420288287336007,
          "train_loss": 0.004387508612126112,
          "validation_loss": 0.003044320037588477,
          "train_metric": 0.0004418937605805695,
          "validation_metric": 0.0003066127828788012,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.5156292149986257
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.003492709667625604,
          "train_loss": 0.0044592865742743015,
          "validation_loss": 0.0039874836802482605,
          "train_metric": 0.0004491230065468699,
          "validation_metric": 0.00040160477510653436,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.5279254050001327
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.004293229045288172,
          "train_loss": 0.0035472847521305084,
          "validation_loss": 0.0029675839468836784,
          "train_metric": 0.00035726951318793,
          "validation_metric": 0.0002988842024933547,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.540013297999394
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.004702866077423096,
          "train_loss": 0.003585374914109707,
          "validation_loss": 0.003011337947100401,
          "train_metric": 0.00036110583459958434,
          "validation_metric": 0.0003032909007743001,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.5520406649993674
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0033236273302463815,
          "train_loss": 0.0034219257067888975,
          "validation_loss": 0.0024908361956477165,
          "train_metric": 0.00034464383497834206,
          "validation_metric": 0.0002508678880985826,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.5644721150001715
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.003575744296540506,
          "train_loss": 0.0036436051595956087,
          "validation_loss": 0.002639387035742402,
          "train_metric": 0.00036697054747492075,
          "validation_metric": 0.0002658293815329671,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.576437174999228
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.003304269070213195,
          "train_loss": 0.0032200049608945847,
          "validation_loss": 0.002796364948153496,
          "train_metric": 0.000324307125993073,
          "validation_metric": 0.00028163965907879174,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.5909249129999807
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0031906527547107544,
          "train_loss": 0.0030243895016610622,
          "validation_loss": 0.002450580708682537,
          "train_metric": 0.00030460540438070893,
          "validation_metric": 0.0002468135207891464,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.6033754279997083
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.003116214298643172,
          "train_loss": 0.0040935310535132885,
          "validation_loss": 0.003691093297675252,
          "train_metric": 0.0004122854443266988,
          "validation_metric": 0.00037175341276451945,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.615438656999686
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0037108174365130253,
          "train_loss": 0.0033132852986454964,
          "validation_loss": 0.0028483800124377012,
          "train_metric": 0.00033370195887982845,
          "validation_metric": 0.0002868784358724952,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.626874888999737
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0032649154745740816,
          "train_loss": 0.0032895072363317013,
          "validation_loss": 0.002858857624232769,
          "train_metric": 0.0003313070919830352,
          "validation_metric": 0.00028793365345336497,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.63841896699887
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0031424571825482417,
          "train_loss": 0.0028846068307757378,
          "validation_loss": 0.0022548495326191187,
          "train_metric": 0.0002905270375777036,
          "validation_metric": 0.0002271002158522606,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.650706933998663
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.00304165403576917,
          "train_loss": 0.0029053424950689077,
          "validation_loss": 0.0022576642222702503,
          "train_metric": 0.00029261544113978744,
          "validation_metric": 0.00022738368716090918,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.6625277509992884
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.003120211371424375,
          "train_loss": 0.00407961243763566,
          "validation_loss": 0.0029099888633936644,
          "train_metric": 0.00041088362922891974,
          "validation_metric": 0.00029308340162970126,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.6742917310002667
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0038932722418394405,
          "train_loss": 0.0037094061262905598,
          "validation_loss": 0.0027539364527910948,
          "train_metric": 0.0003735978389158845,
          "validation_metric": 0.00027736640186049044,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.685516909999933
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.004235325250192545,
          "train_loss": 0.004167043603956699,
          "validation_loss": 0.003041912568733096,
          "train_metric": 0.0004196894005872309,
          "validation_metric": 0.0003063703188672662,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.696995866999714
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.00339452793559758,
          "train_loss": 0.004343399778008461,
          "validation_loss": 0.00326365465298295,
          "train_metric": 0.0004374513518996537,
          "validation_metric": 0.0003287033177912235,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.708405862998916
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.003911161944415653,
          "train_loss": 0.003030705964192748,
          "validation_loss": 0.002518812892958522,
          "train_metric": 0.00030524161411449313,
          "validation_metric": 0.00025368560454808176,
          "lr_first": 0.0009,
          "lr_last": 0.0009,
          "seconds": 3.719852718999391
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.003508894093101844,
          "train_loss": 0.0030665872618556023,
          "validation_loss": 0.00229842821136117,
          "train_metric": 0.00030885543674230576,
          "validation_metric": 0.00023148923355620354,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.732026091998705
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0029590882040793076,
          "train_loss": 0.002738995011895895,
          "validation_loss": 0.0021904893219470978,
          "train_metric": 0.0002758615300990641,
          "validation_metric": 0.00022061806521378458,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7442732459985564
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0028967426587769296,
          "train_loss": 0.0027513280510902405,
          "validation_loss": 0.002232820261269808,
          "train_metric": 0.00027710371068678796,
          "validation_metric": 0.00022488151444122195,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7562868959994375
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0028388001992425416,
          "train_loss": 0.00273332092911005,
          "validation_loss": 0.0021836955565959215,
          "train_metric": 0.00027529007638804615,
          "validation_metric": 0.00021993379050400108,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7680549629985762
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0027890087803825736,
          "train_loss": 0.002730238251388073,
          "validation_loss": 0.002154580084607005,
          "train_metric": 0.00027497956762090325,
          "validation_metric": 0.00021700141951441765,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.7801220509991253
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0027651901764329523,
          "train_loss": 0.002718048170208931,
          "validation_loss": 0.0021637985482811928,
          "train_metric": 0.00027375182253308594,
          "validation_metric": 0.00021792987536173314,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.791877563999151
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.002750997053226456,
          "train_loss": 0.0027078574057668447,
          "validation_loss": 0.002192836022004485,
          "train_metric": 0.00027272544684819877,
          "validation_metric": 0.00022085444652475417,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.8036314629989647
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0028583766652445775,
          "train_loss": 0.0027513063978403807,
          "validation_loss": 0.002277075545862317,
          "train_metric": 0.00027710149879567325,
          "validation_metric": 0.0002293387078680098,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.815846621999299
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0027559098780329805,
          "train_loss": 0.0026954703498631716,
          "validation_loss": 0.002181601943448186,
          "train_metric": 0.0002714778820518404,
          "validation_metric": 0.0002197229623561725,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.8287168810002186
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0027539794136828277,
          "train_loss": 0.0026911620516330004,
          "validation_loss": 0.002178539289161563,
          "train_metric": 0.0002710439439397305,
          "validation_metric": 0.00021941447630524635,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.839924694999354
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.002791362385323737,
          "train_loss": 0.00268525630235672,
          "validation_loss": 0.002148301573470235,
          "train_metric": 0.0002704491780605167,
          "validation_metric": 0.00021636905148625374,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.8519736749985896
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.002741131069342373,
          "train_loss": 0.00273197703063488,
          "validation_loss": 0.0021334164775907993,
          "train_metric": 0.0002751547144725919,
          "validation_metric": 0.00021486992773134261,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.864576991998547
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0027848558602272533,
          "train_loss": 0.00268984935246408,
          "validation_loss": 0.0022051138803362846,
          "train_metric": 0.0002709117834456265,
          "validation_metric": 0.0002220909809693694,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.876525333000245
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0027924484602408484,
          "train_loss": 0.002677302807569504,
          "validation_loss": 0.0021812659688293934,
          "train_metric": 0.00026964815333485603,
          "validation_metric": 0.00021968911460135132,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.887908343000163
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0028720221598632634,
          "train_loss": 0.0027346282731741667,
          "validation_loss": 0.002151301596313715,
          "train_metric": 0.00027542177122086287,
          "validation_metric": 0.000216671236557886,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.8997488529985276
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0030697241745656356,
          "train_loss": 0.0027501676231622696,
          "validation_loss": 0.00227718916721642,
          "train_metric": 0.0002769867714960128,
          "validation_metric": 0.00022935016022529453,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 3.9917082989995833
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.003129137840005569,
          "train_loss": 0.0028239262755960226,
          "validation_loss": 0.002185560530051589,
          "train_metric": 0.00028441555332392454,
          "validation_metric": 0.0002201216557295993,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.0038802639992355
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.002825466715876246,
          "train_loss": 0.0027627686504274607,
          "validation_loss": 0.002300407737493515,
          "train_metric": 0.0002782559604384005,
          "validation_metric": 0.0002316886675544083,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.018771446999381
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.002714617246965645,
          "train_loss": 0.002745156642049551,
          "validation_loss": 0.002133432310074568,
          "train_metric": 0.0002764820819720626,
          "validation_metric": 0.00021487148478627205,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.032302010999047
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0029122400810592808,
          "train_loss": 0.0026689006481319666,
          "validation_loss": 0.002093044575303793,
          "train_metric": 0.00026880193036049604,
          "validation_metric": 0.00021080378792248666,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.044877143000122
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002717001782002626,
          "train_loss": 0.0026678191497921944,
          "validation_loss": 0.0021504275500774384,
          "train_metric": 0.00026869302382692695,
          "validation_metric": 0.0002165831538150087,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.057390314999793
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.002714655416639289,
          "train_loss": 0.002642781473696232,
          "validation_loss": 0.002116643823683262,
          "train_metric": 0.0002661712933331728,
          "validation_metric": 0.00021318062499631196,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.069043265999426
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0028233479933987837,
          "train_loss": 0.0026386636309325695,
          "validation_loss": 0.0021148724481463432,
          "train_metric": 0.0002657565346453339,
          "validation_metric": 0.00021300220396369696,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.080992679999326
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0027389666356611997,
          "train_loss": 0.0026560865808278322,
          "validation_loss": 0.0021806752774864435,
          "train_metric": 0.00026751129189506173,
          "validation_metric": 0.0002196295972680673,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.093214589000127
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.002672456543223234,
          "train_loss": 0.0026090641040354967,
          "validation_loss": 0.0020933339837938547,
          "train_metric": 0.00026277542929165065,
          "validation_metric": 0.00021083293540868908,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.105568327999208
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.002660261256096419,
          "train_loss": 0.0026160208508372307,
          "validation_loss": 0.0021117483265697956,
          "train_metric": 0.00026347607490606606,
          "validation_metric": 0.00021268757700454444,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.117731956999705
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.002717546587518882,
          "train_loss": 0.0025975266471505165,
          "validation_loss": 0.0020872403401881456,
          "train_metric": 0.00026161340065300465,
          "validation_metric": 0.0002102192083839327,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.130645631999869
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0027944294051849283,
          "train_loss": 0.0026052864268422127,
          "validation_loss": 0.0021195521112531424,
          "train_metric": 0.0002623949258122593,
          "validation_metric": 0.00021347354049794376,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.142481392998889
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0026776099402923137,
          "train_loss": 0.0025928188115358353,
          "validation_loss": 0.002062567975372076,
          "train_metric": 0.0002611392119433731,
          "validation_metric": 0.00020773430878762156,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.154701962999752
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0026150248777412344,
          "train_loss": 0.002587886992841959,
          "validation_loss": 0.002075116615742445,
          "train_metric": 0.00026064246776513755,
          "validation_metric": 0.00020899817172903568,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.167003361999377
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0026886422565439716,
          "train_loss": 0.002607128117233515,
          "validation_loss": 0.0020635495893657207,
          "train_metric": 0.00026258043362759054,
          "validation_metric": 0.00020783315994776785,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.179009838999264
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.0026996324450010434,
          "train_loss": 0.002619128907099366,
          "validation_loss": 0.00214870972558856,
          "train_metric": 0.0002637890283949673,
          "validation_metric": 0.00021641018975060433,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.191393834998962
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0027391475159674883,
          "train_loss": 0.002691226778551936,
          "validation_loss": 0.002076806267723441,
          "train_metric": 0.00027105052140541375,
          "validation_metric": 0.0002091682981699705,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.203479541998604
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.002681559995835414,
          "train_loss": 0.0026294472627341747,
          "validation_loss": 0.0021805951837450266,
          "train_metric": 0.0002648283261805773,
          "validation_metric": 0.000219621550058946,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.2154730949987425
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.002619715731270844,
          "train_loss": 0.002665199339389801,
          "validation_loss": 0.0020715617574751377,
          "train_metric": 0.0002684291102923453,
          "validation_metric": 0.00020864009275101125,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.230732533998889
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.002678971522982465,
          "train_loss": 0.002614332363009453,
          "validation_loss": 0.002157805720344186,
          "train_metric": 0.00026330596301704645,
          "validation_metric": 0.00021732630557380617,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.244043609000073
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.002612678727018647,
          "train_loss": 0.0026081379037350416,
          "validation_loss": 0.0020445489790290594,
          "train_metric": 0.00026268212241120636,
          "validation_metric": 0.00020591948123183101,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.256496513999082
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.002866339847969357,
          "train_loss": 0.0026941217947751284,
          "validation_loss": 0.0022511815186589956,
          "train_metric": 0.00027134205447509885,
          "validation_metric": 0.00022673075727652758,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.268149072999222
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.002694720336876344,
          "train_loss": 0.00257632858119905,
          "validation_loss": 0.002031969837844372,
          "train_metric": 0.0002594784018583596,
          "validation_metric": 0.00020465257694013417,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.2805773860000045
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0025983898412960116,
          "train_loss": 0.002533211372792721,
          "validation_loss": 0.0020324161741882563,
          "train_metric": 0.0002551357611082494,
          "validation_metric": 0.00020469751325435936,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.294240546998481
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0025942290012608282,
          "train_loss": 0.0025311328936368227,
          "validation_loss": 0.002029937459155917,
          "train_metric": 0.0002549264463596046,
          "validation_metric": 0.00020444787514861673,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.306483598998966
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0027618346430244856,
          "train_loss": 0.0026617010589689016,
          "validation_loss": 0.0022309008054435253,
          "train_metric": 0.0002680768084246665,
          "validation_metric": 0.00022468814859166741,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.318179931999111
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0026365865305706393,
          "train_loss": 0.0025579684879630804,
          "validation_loss": 0.0020068625453859568,
          "train_metric": 0.0002576292317826301,
          "validation_metric": 0.00020212384697515517,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.330822908999835
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.002629137110488955,
          "train_loss": 0.002538927597925067,
          "validation_loss": 0.0019961586222052574,
          "train_metric": 0.00025571149308234453,
          "validation_metric": 0.0002010458119912073,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.343283390999204
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.00268510354362661,
          "train_loss": 0.0025248629972338676,
          "validation_loss": 0.0020093503408133984,
          "train_metric": 0.0002542949514463544,
          "validation_metric": 0.00020237441640347242,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.354933691998667
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0025652794065536,
          "train_loss": 0.002505990443751216,
          "validation_loss": 0.002003981964662671,
          "train_metric": 0.00025239420938305557,
          "validation_metric": 0.00020183372544124722,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.366350443999181
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0025623406399972737,
          "train_loss": 0.002561295637860894,
          "validation_loss": 0.0021329945884644985,
          "train_metric": 0.000257964275078848,
          "validation_metric": 0.00021482740703504533,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.377936419999969
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0025890316828736104,
          "train_loss": 0.0025409001391381025,
          "validation_loss": 0.0019789617508649826,
          "train_metric": 0.00025591018493287265,
          "validation_metric": 0.00019931378483306617,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.390038122999613
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.00251686336559942,
          "train_loss": 0.0024857495445758104,
          "validation_loss": 0.0020084655843675137,
          "train_metric": 0.00025035563157871366,
          "validation_metric": 0.00020228527137078345,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.401901688999715
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0026083278462465387,
          "train_loss": 0.0025046132504940033,
          "validation_loss": 0.0020048825535923243,
          "train_metric": 0.00025225544231943786,
          "validation_metric": 0.0002019244129769504,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.413632333998976
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0028220077692822088,
          "train_loss": 0.002570404438301921,
          "validation_loss": 0.0021030521020293236,
          "train_metric": 0.00025888177333399653,
          "validation_metric": 0.00021181171177886426,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.428891085999567
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.002714548263611505,
          "train_loss": 0.0025194655172526836,
          "validation_loss": 0.00200667604804039,
          "train_metric": 0.00025375132099725306,
          "validation_metric": 0.00020210504590068012,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.521685259998776
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0025766202343220357,
          "train_loss": 0.002530014142394066,
          "validation_loss": 0.0019734185189008713,
          "train_metric": 0.0002548137854319066,
          "validation_metric": 0.0001987555151572451,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.534626933998879
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.00256531478225952,
          "train_loss": 0.0025388707872480154,
          "validation_loss": 0.00196094554848969,
          "train_metric": 0.00025570575962774456,
          "validation_metric": 0.0001974992483155802,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.5469228490001115
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.002509244965040125,
          "train_loss": 0.002498035319149494,
          "validation_loss": 0.0020572158973664045,
          "train_metric": 0.0002515930391382426,
          "validation_metric": 0.00020719524763990194,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.558854040998995
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0025389361471752636,
          "train_loss": 0.002462056465446949,
          "validation_loss": 0.0020037118811160326,
          "train_metric": 0.00024796929210424423,
          "validation_metric": 0.00020180649880785495,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.570204943000135
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.002535691179218702,
          "train_loss": 0.002488157246261835,
          "validation_loss": 0.002033993834629655,
          "train_metric": 0.00025059806648641825,
          "validation_metric": 0.0002048564056167379,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.581897021000259
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0025547645200276747,
          "train_loss": 0.0025884227361530066,
          "validation_loss": 0.001998220570385456,
          "train_metric": 0.00026069648447446525,
          "validation_metric": 0.00020125345326960087,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.59361158899992
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0027472369620227255,
          "train_loss": 0.002556795487180352,
          "validation_loss": 0.002111603505909443,
          "train_metric": 0.0002575110411271453,
          "validation_metric": 0.00021267296688165516,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.605220842999188
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.002595630088762846,
          "train_loss": 0.0024785627610981464,
          "validation_loss": 0.0019497129833325744,
          "train_metric": 0.0002496317611075938,
          "validation_metric": 0.00019636798242572695,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.6173765109997476
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.002812055310641881,
          "train_loss": 0.002574366517364979,
          "validation_loss": 0.0019803340546786785,
          "train_metric": 0.00025928078684955835,
          "validation_metric": 0.00019945198437198997,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.632671780998862
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.002659091867826646,
          "train_loss": 0.0024864706210792065,
          "validation_loss": 0.002042614622041583,
          "train_metric": 0.00025042821653187275,
          "validation_metric": 0.00020572467474266887,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.6442663189991436
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.0024875979506759904,
          "train_loss": 0.002419011201709509,
          "validation_loss": 0.0019591792952269316,
          "train_metric": 0.0002436339418636635,
          "validation_metric": 0.0001973213511519134,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.657373550999182
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0025263197712774854,
          "train_loss": 0.0024786991998553276,
          "validation_loss": 0.0019223669078201056,
          "train_metric": 0.0002496455272193998,
          "validation_metric": 0.00019361375598236918,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.669752698999218
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.002571788922068663,
          "train_loss": 0.0024255765601992607,
          "validation_loss": 0.0019428916275501251,
          "train_metric": 0.0002442951372358948,
          "validation_metric": 0.00019568092830013484,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.682523580999259
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.002524110095691867,
          "train_loss": 0.002419686410576105,
          "validation_loss": 0.001960781402885914,
          "train_metric": 0.0002437019720673561,
          "validation_metric": 0.00019748271733988076,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.694572065000102
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.002506953969714232,
          "train_loss": 0.002435436937958002,
          "validation_loss": 0.0019306456670165062,
          "train_metric": 0.00024528830545023084,
          "validation_metric": 0.00019444756617303938,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.706678880000254
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.00254276136183762,
          "train_loss": 0.002407628810033202,
          "validation_loss": 0.0019247066229581833,
          "train_metric": 0.00024248755653388798,
          "validation_metric": 0.00019384940969757736,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.718739813999491
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.002636208853800781,
          "train_loss": 0.002621079096570611,
          "validation_loss": 0.0020209839567542076,
          "train_metric": 0.00026398547925055027,
          "validation_metric": 0.0002035461220657453,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.730815629998688
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0025763841113075614,
          "train_loss": 0.0025887987576425076,
          "validation_loss": 0.002167294267565012,
          "train_metric": 0.000260734319454059,
          "validation_metric": 0.00021828194439876825,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.742741144998945
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.0025055346450244542,
          "train_loss": 0.002388628199696541,
          "validation_loss": 0.0019235862419009209,
          "train_metric": 0.00024057387781795114,
          "validation_metric": 0.0001937365741468966,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.754395719999593
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.002426501756417565,
          "train_loss": 0.0024412956554442644,
          "validation_loss": 0.0019035791046917439,
          "train_metric": 0.00024587835650891066,
          "validation_metric": 0.00019172154134139419,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.766754186999606
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.002542096390243387,
          "train_loss": 0.0023737966548651457,
          "validation_loss": 0.0019070453708991408,
          "train_metric": 0.00023908009461592883,
          "validation_metric": 0.0001920706417877227,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.778939507999894
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.0024506454428774305,
          "train_loss": 0.002418645191937685,
          "validation_loss": 0.0018789051100611687,
          "train_metric": 0.0002435970673104748,
          "validation_metric": 0.00018923645257018507,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.791488942999422
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.0024823407038638834,
          "train_loss": 0.0023879902437329292,
          "validation_loss": 0.0018735299818217754,
          "train_metric": 0.0002405096311122179,
          "validation_metric": 0.00018869504856411368,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.804001751999749
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.002462386626575608,
          "train_loss": 0.002348274691030383,
          "validation_loss": 0.0018922103336080909,
          "train_metric": 0.00023650962975807488,
          "validation_metric": 0.00019057648023590446,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.815877528999408
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.002452360622555716,
          "train_loss": 0.0024014210794121027,
          "validation_loss": 0.0018813086207956076,
          "train_metric": 0.00024186234804801643,
          "validation_metric": 0.00018947852368000895,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.8308108749988605
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.0024828158202581108,
          "train_loss": 0.0023722657933831215,
          "validation_loss": 0.0018838964169844985,
          "train_metric": 0.00023892591707408428,
          "validation_metric": 0.000189739148481749,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.842884658999537
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.0024194509751396254,
          "train_loss": 0.002354611176997423,
          "validation_loss": 0.0018732913304120302,
          "train_metric": 0.00023714781855233014,
          "validation_metric": 0.00018867106700781733,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.856801067999186
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.0024512262716598343,
          "train_loss": 0.0024630744010210037,
          "validation_loss": 0.0020748015958815813,
          "train_metric": 0.00024807179579511285,
          "validation_metric": 0.00020896640489809215,
          "lr_first": 0.00027,
          "lr_last": 0.00027,
          "seconds": 4.868787856999916
        }
      ],
      "best": {
        "epoch": 239,
        "updates": 3824
      },
      "best_validation_metric": 0.00018867106700781733,
      "config": {
        "label": "Step decay",
        "batch": 64,
        "lr": 0.003,
        "schedule": "step",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.0137357586979321
    },
    "9967999fa0f8_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.19682097947224975,
          "train_loss": 0.6091052293777466,
          "validation_loss": 0.4449542760848999,
          "train_metric": 0.06134684756398201,
          "validation_metric": 0.04481416940689087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01599734800038277
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.058354404056444764,
          "train_loss": 0.14781013131141663,
          "validation_loss": 0.11435435712337494,
          "train_metric": 0.014886895194649696,
          "validation_metric": 0.011517352424561977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.032026864999352256
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.057119667646475136,
          "train_loss": 0.039981674402952194,
          "validation_loss": 0.029736973345279694,
          "train_metric": 0.0040268078446388245,
          "validation_metric": 0.0029949992895126343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04759623400059354
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.0530130424303934,
          "train_loss": 0.027812983840703964,
          "validation_loss": 0.022474143654108047,
          "train_metric": 0.002801222261041403,
          "validation_metric": 0.0022635136265307665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.062980938999317
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.06610384373925626,
          "train_loss": 0.042283426970243454,
          "validation_loss": 0.03110162727534771,
          "train_metric": 0.004258631728589535,
          "validation_metric": 0.0031324424780905247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07808006999948702
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.048984937951900065,
          "train_loss": 0.022569144144654274,
          "validation_loss": 0.017640821635723114,
          "train_metric": 0.002273081336170435,
          "validation_metric": 0.001776718650944531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09427715600031661
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05377026624046266,
          "train_loss": 0.027647489681839943,
          "validation_loss": 0.02143549732863903,
          "train_metric": 0.002784553449600935,
          "validation_metric": 0.0021589049138128757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10913115000039397
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04934446350671351,
          "train_loss": 0.034094005823135376,
          "validation_loss": 0.024037521332502365,
          "train_metric": 0.003433823585510254,
          "validation_metric": 0.0024209709372371435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12385154299954593
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0650314548111055,
          "train_loss": 0.05951915681362152,
          "validation_loss": 0.0484737828373909,
          "train_metric": 0.005994551349431276,
          "validation_metric": 0.004882101435214281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13815756999974838
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05463054869323969,
          "train_loss": 0.03044225089251995,
          "validation_loss": 0.025889094918966293,
          "train_metric": 0.0030660321936011314,
          "validation_metric": 0.0026074545457959175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15230291099942406
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.048171582631766796,
          "train_loss": 0.022896233946084976,
          "validation_loss": 0.01884184218943119,
          "train_metric": 0.002306024543941021,
          "validation_metric": 0.0018976811552420259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16682364500047697
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.038186646008398384,
          "train_loss": 0.03902680426836014,
          "validation_loss": 0.0334593690931797,
          "train_metric": 0.003930637147277594,
          "validation_metric": 0.0033699055202305317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18117373899985978
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.0537871653214097,
          "train_loss": 0.02045082487165928,
          "validation_loss": 0.017042066901922226,
          "train_metric": 0.0020597318653017282,
          "validation_metric": 0.0017164144665002823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19660680100059835
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.04304345918353647,
          "train_loss": 0.01814754493534565,
          "validation_loss": 0.014318203553557396,
          "train_metric": 0.00182775454595685,
          "validation_metric": 0.0014420769875869155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2123296730005677
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.0418543143896386,
          "train_loss": 0.01800830475986004,
          "validation_loss": 0.013800390064716339,
          "train_metric": 0.0018137304577976465,
          "validation_metric": 0.001389924786053598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22725196899955336
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.04753694770624861,
          "train_loss": 0.02508183941245079,
          "validation_loss": 0.0194687657058239,
          "train_metric": 0.002526151016354561,
          "validation_metric": 0.001960823079571128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24189801800093846
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03891421266598627,
          "train_loss": 0.041287861764431,
          "validation_loss": 0.03361939266324043,
          "train_metric": 0.0041583627462387085,
          "validation_metric": 0.003386022290214896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25610588600102346
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.037983305810485035,
          "train_loss": 0.01780412346124649,
          "validation_loss": 0.014118591323494911,
          "train_metric": 0.0017931663896888494,
          "validation_metric": 0.0014219727599993348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27026157200089074
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.03596429375465959,
          "train_loss": 0.013087221421301365,
          "validation_loss": 0.00977838784456253,
          "train_metric": 0.0013180967653170228,
          "validation_metric": 0.0009848434710875154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2848834190008347
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.035559753072448075,
          "train_loss": 0.014981775544583797,
          "validation_loss": 0.010946257971227169,
          "train_metric": 0.0015089093940332532,
          "validation_metric": 0.001102467067539692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29928255699996953
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.027796606125775725,
          "train_loss": 0.014671159908175468,
          "validation_loss": 0.011582918465137482,
          "train_metric": 0.0014776256866753101,
          "validation_metric": 0.001166589092463255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3134882470003504
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.039438113337382674,
          "train_loss": 0.012985244393348694,
          "validation_loss": 0.010294057428836823,
          "train_metric": 0.0013078262563794851,
          "validation_metric": 0.0010367798386141658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3273098989993741
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.02526115847285837,
          "train_loss": 0.012070661410689354,
          "validation_loss": 0.008994771167635918,
          "train_metric": 0.0012157128658145666,
          "validation_metric": 0.0009059203439392149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34212257100080024
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.033100065775215626,
          "train_loss": 0.019964782521128654,
          "validation_loss": 0.01513717882335186,
          "train_metric": 0.0020107801537960768,
          "validation_metric": 0.0015245613176375628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44064894000075583
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.03584234288427979,
          "train_loss": 0.011362169869244099,
          "validation_loss": 0.008706995286047459,
          "train_metric": 0.001144356094300747,
          "validation_metric": 0.0008769367705099285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4564172200007306
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.036358065670356154,
          "train_loss": 0.01233716681599617,
          "validation_loss": 0.00868823193013668,
          "train_metric": 0.0012425543973222375,
          "validation_metric": 0.00087504688417539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47171619499931694
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.03932574312784709,
          "train_loss": 0.030448850244283676,
          "validation_loss": 0.027755504474043846,
          "train_metric": 0.0030666966922581196,
          "validation_metric": 0.0027954326942563057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48562769000091066
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.028520736552309245,
          "train_loss": 0.015581130050122738,
          "validation_loss": 0.012823951430618763,
          "train_metric": 0.0015692743472754955,
          "validation_metric": 0.001291581429541111,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5002128110008925
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.04234867275226861,
          "train_loss": 0.053118858486413956,
          "validation_loss": 0.047223448753356934,
          "train_metric": 0.00534993689507246,
          "validation_metric": 0.004756172653287649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5143435200006934
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.03642092365771532,
          "train_loss": 0.01375625655055046,
          "validation_loss": 0.01091674529016018,
          "train_metric": 0.00138547969982028,
          "validation_metric": 0.0010994946351274848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5288010750009562
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.04914502857718617,
          "train_loss": 0.01271960511803627,
          "validation_loss": 0.010201000608503819,
          "train_metric": 0.0012810719199478626,
          "validation_metric": 0.001027407357469201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5430793340001401
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.026065201964229345,
          "train_loss": 0.014119801111519337,
          "validation_loss": 0.011303075589239597,
          "train_metric": 0.0014220945304259658,
          "validation_metric": 0.0011384044773876667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5573893040000257
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.030540888023097068,
          "train_loss": 0.010499193333089352,
          "validation_loss": 0.008200853131711483,
          "train_metric": 0.0010574405314400792,
          "validation_metric": 0.0008259598398581147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5722907130002568
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.032828566909302026,
          "train_loss": 0.013440784066915512,
          "validation_loss": 0.0096448864787817,
          "train_metric": 0.0013537065824493766,
          "validation_metric": 0.0009713975596241653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5865179850006825
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.031062791880685836,
          "train_loss": 0.010150975547730923,
          "validation_loss": 0.008124027401208878,
          "train_metric": 0.0010223690187558532,
          "validation_metric": 0.0008182222372852266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6024655860001076
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.02767683024285361,
          "train_loss": 0.06293740123510361,
          "validation_loss": 0.05288161337375641,
          "train_metric": 0.006338824052363634,
          "validation_metric": 0.0053260428830981255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.616493392000848
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.04154013784136623,
          "train_loss": 0.029071200639009476,
          "validation_loss": 0.02533576637506485,
          "train_metric": 0.0029279449954628944,
          "validation_metric": 0.002551725599914789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6308627960006561
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.03561796236317605,
          "train_loss": 0.016969874501228333,
          "validation_loss": 0.013431817293167114,
          "train_metric": 0.0017091436311602592,
          "validation_metric": 0.001352803548797965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6456843899995874
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.046444684034213424,
          "train_loss": 0.016665874049067497,
          "validation_loss": 0.014252765104174614,
          "train_metric": 0.001678525935858488,
          "validation_metric": 0.0014354862505570054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6604478799999924
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.04787315125577152,
          "train_loss": 0.013596932403743267,
          "validation_loss": 0.011985552497208118,
          "train_metric": 0.0013694331282749772,
          "validation_metric": 0.0012071409728378057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.674914451999939
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.035440784791717306,
          "train_loss": 0.012962195090949535,
          "validation_loss": 0.010506550781428814,
          "train_metric": 0.001305504934862256,
          "validation_metric": 0.0010581813985481858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6896220640010142
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.025460661767283455,
          "train_loss": 0.012650489807128906,
          "validation_loss": 0.009119914844632149,
          "train_metric": 0.001274110865779221,
          "validation_metric": 0.0009185243397951126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7040066860008665
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.0398795697838068,
          "train_loss": 0.014821094460785389,
          "validation_loss": 0.012090771459043026,
          "train_metric": 0.0014927263837307692,
          "validation_metric": 0.0012177383759990335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7190825979996589
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.04464480030583218,
          "train_loss": 0.02799677476286888,
          "validation_loss": 0.024175232276320457,
          "train_metric": 0.002819732530042529,
          "validation_metric": 0.002434840891510248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7332328009997582
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.029584824573248625,
          "train_loss": 0.012369277887046337,
          "validation_loss": 0.009318667463958263,
          "train_metric": 0.0012457884149625897,
          "validation_metric": 0.0009385420707985759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7478503610000189
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.027532318199519068,
          "train_loss": 0.020694365724921227,
          "validation_loss": 0.016881922259926796,
          "train_metric": 0.002084260107949376,
          "validation_metric": 0.0017002852400764823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7627661940005055
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.03643512004055083,
          "train_loss": 0.010674141347408295,
          "validation_loss": 0.008448445238173008,
          "train_metric": 0.001075060456059873,
          "validation_metric": 0.0008508963510394096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7772534630003065
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.040426963416393846,
          "train_loss": 0.0178165752440691,
          "validation_loss": 0.014517176896333694,
          "train_metric": 0.0017944202991202474,
          "validation_metric": 0.0014621168375015259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7916995819996373
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.031613720464520156,
          "train_loss": 0.017063280567526817,
          "validation_loss": 0.0132959159091115,
          "train_metric": 0.0017185511533170938,
          "validation_metric": 0.0013391160173341632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8058622450007533
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.035274466150440276,
          "train_loss": 0.018741758540272713,
          "validation_loss": 0.0158102847635746,
          "train_metric": 0.001887601101770997,
          "validation_metric": 0.001592353917658329,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8202249179994396
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.025931313866749406,
          "train_loss": 0.014625156298279762,
          "validation_loss": 0.01199561171233654,
          "train_metric": 0.0014729922404512763,
          "validation_metric": 0.001208153902553022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8352703079999628
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.019343027379363775,
          "train_loss": 0.012571475468575954,
          "validation_loss": 0.010472558438777924,
          "train_metric": 0.0012661528307944536,
          "validation_metric": 0.0010547577403485775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8493712760009657
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.03264982491964474,
          "train_loss": 0.012063810601830482,
          "validation_loss": 0.008899045176804066,
          "train_metric": 0.0012150228722020984,
          "validation_metric": 0.0008962792344391346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8638270110004669
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.040746721846517175,
          "train_loss": 0.015441740863025188,
          "validation_loss": 0.012129580602049828,
          "train_metric": 0.0015552357072010636,
          "validation_metric": 0.0012216470204293728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8788338310005201
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.02784228848759085,
          "train_loss": 0.012928735464811325,
          "validation_loss": 0.01112405862659216,
          "train_metric": 0.00130213494412601,
          "validation_metric": 0.0011203745380043983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8936191220000183
        }
      ],
      "best": {
        "epoch": 35,
        "updates": 560
      },
      "best_validation_metric": 0.0008182222372852266,
      "config": {
        "label": "BatchNorm",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "batch",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.028604584200530282
    },
    "8ac87378bb5b_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 64,
          "samples_seen": 1024,
          "online_loss": 0.2624188729969319,
          "train_loss": 0.1171494722366333,
          "validation_loss": 0.09625931084156036,
          "train_metric": 0.011798867024481297,
          "validation_metric": 0.009694884531199932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04733017699982156
        },
        {
          "epoch": 2,
          "updates": 128,
          "samples_seen": 2048,
          "online_loss": 0.08470732023124583,
          "train_loss": 0.07202138006687164,
          "validation_loss": 0.05905001237988472,
          "train_metric": 0.007253730203956366,
          "validation_metric": 0.005947301164269447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09161780499925953
        },
        {
          "epoch": 3,
          "updates": 192,
          "samples_seen": 3072,
          "online_loss": 0.06356297535239719,
          "train_loss": 0.054105907678604126,
          "validation_loss": 0.04593408852815628,
          "train_metric": 0.00544934906065464,
          "validation_metric": 0.004626313224434853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13563267899917264
        },
        {
          "epoch": 4,
          "updates": 256,
          "samples_seen": 4096,
          "online_loss": 0.054570174688706174,
          "train_loss": 0.04432539641857147,
          "validation_loss": 0.038256630301475525,
          "train_metric": 0.004464291501790285,
          "validation_metric": 0.003853068221360445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1789789689992176
        },
        {
          "epoch": 5,
          "updates": 320,
          "samples_seen": 5120,
          "online_loss": 0.04212082936282968,
          "train_loss": 0.03483256697654724,
          "validation_loss": 0.03159569203853607,
          "train_metric": 0.003508209250867367,
          "validation_metric": 0.0031822023447602987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22242334200018377
        },
        {
          "epoch": 6,
          "updates": 384,
          "samples_seen": 6144,
          "online_loss": 0.03481512506186846,
          "train_loss": 0.025710269808769226,
          "validation_loss": 0.021541653200984,
          "train_metric": 0.002589444164186716,
          "validation_metric": 0.002169596031308174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26763055700030236
        },
        {
          "epoch": 7,
          "updates": 448,
          "samples_seen": 7168,
          "online_loss": 0.02729321763399639,
          "train_loss": 0.020683294162154198,
          "validation_loss": 0.017299188300967216,
          "train_metric": 0.002083145547658205,
          "validation_metric": 0.0017423109384253621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31293350299893063
        },
        {
          "epoch": 8,
          "updates": 512,
          "samples_seen": 8192,
          "online_loss": 0.022401315309252823,
          "train_loss": 0.029240833595395088,
          "validation_loss": 0.024743523448705673,
          "train_metric": 0.002945029642432928,
          "validation_metric": 0.0024920771829783916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.357081002999621
        },
        {
          "epoch": 9,
          "updates": 576,
          "samples_seen": 9216,
          "online_loss": 0.024232984520494938,
          "train_loss": 0.019727662205696106,
          "validation_loss": 0.01657680608332157,
          "train_metric": 0.0019868980161845684,
          "validation_metric": 0.0016695550875738263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4994835219986271
        },
        {
          "epoch": 10,
          "updates": 640,
          "samples_seen": 10240,
          "online_loss": 0.02759063059784239,
          "train_loss": 0.022793717682361603,
          "validation_loss": 0.018758468329906464,
          "train_metric": 0.0022956999018788338,
          "validation_metric": 0.0018892838852480054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5437342469995201
        },
        {
          "epoch": 11,
          "updates": 704,
          "samples_seen": 11264,
          "online_loss": 0.019792282218986657,
          "train_loss": 0.018156468868255615,
          "validation_loss": 0.015147591009736061,
          "train_metric": 0.0018286528065800667,
          "validation_metric": 0.001525609870441258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5872794469996734
        },
        {
          "epoch": 12,
          "updates": 768,
          "samples_seen": 12288,
          "online_loss": 0.01675870963936177,
          "train_loss": 0.012159856967628002,
          "validation_loss": 0.01066514290869236,
          "train_metric": 0.0012246962869539857,
          "validation_metric": 0.0010741541627794504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.631454374999521
        },
        {
          "epoch": 13,
          "updates": 832,
          "samples_seen": 13312,
          "online_loss": 0.014479773462880985,
          "train_loss": 0.012895269319415092,
          "validation_loss": 0.010621171444654465,
          "train_metric": 0.0012987643713131547,
          "validation_metric": 0.0010697254911065102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.675501470999734
        },
        {
          "epoch": 14,
          "updates": 896,
          "samples_seen": 14336,
          "online_loss": 0.01538535028885235,
          "train_loss": 0.015002022497355938,
          "validation_loss": 0.010967210866510868,
          "train_metric": 0.0015109487576410174,
          "validation_metric": 0.0011045774444937706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7176478159999533
        },
        {
          "epoch": 15,
          "updates": 960,
          "samples_seen": 15360,
          "online_loss": 0.016265414022200275,
          "train_loss": 0.0220046304166317,
          "validation_loss": 0.017810842022299767,
          "train_metric": 0.0022162257228046656,
          "validation_metric": 0.0017938429955393076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7600568919988291
        },
        {
          "epoch": 16,
          "updates": 1024,
          "samples_seen": 16384,
          "online_loss": 0.012747942846544902,
          "train_loss": 0.0103490324690938,
          "validation_loss": 0.007960377261042595,
          "train_metric": 0.0010423166677355766,
          "validation_metric": 0.0008017400978133082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8042524109987426
        },
        {
          "epoch": 17,
          "updates": 1088,
          "samples_seen": 17408,
          "online_loss": 0.012613498466635065,
          "train_loss": 0.00944472849369049,
          "validation_loss": 0.008201558142900467,
          "train_metric": 0.000951238558627665,
          "validation_metric": 0.000826030969619751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8488221269999485
        },
        {
          "epoch": 18,
          "updates": 1152,
          "samples_seen": 18432,
          "online_loss": 0.009903730539008393,
          "train_loss": 0.012951320968568325,
          "validation_loss": 0.008948936127126217,
          "train_metric": 0.0013044095830991864,
          "validation_metric": 0.0009013040107674897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8924572469986742
        },
        {
          "epoch": 19,
          "updates": 1216,
          "samples_seen": 19456,
          "online_loss": 0.0183119301473198,
          "train_loss": 0.014333060011267662,
          "validation_loss": 0.012964108027517796,
          "train_metric": 0.0014435735065490007,
          "validation_metric": 0.0013056974858045578,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9361211069990532
        },
        {
          "epoch": 20,
          "updates": 1280,
          "samples_seen": 20480,
          "online_loss": 0.010544464745180449,
          "train_loss": 0.007037613540887833,
          "validation_loss": 0.005865545943379402,
          "train_metric": 0.0007088026613928378,
          "validation_metric": 0.0005907563026994467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9801418079987343
        },
        {
          "epoch": 21,
          "updates": 1344,
          "samples_seen": 21504,
          "online_loss": 0.007833955711248564,
          "train_loss": 0.006359841674566269,
          "validation_loss": 0.005162452347576618,
          "train_metric": 0.0006405399763025343,
          "validation_metric": 0.0005199433071538806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0250932649996685
        },
        {
          "epoch": 22,
          "updates": 1408,
          "samples_seen": 22528,
          "online_loss": 0.007362566893789335,
          "train_loss": 0.006041003856807947,
          "validation_loss": 0.004798504989594221,
          "train_metric": 0.0006084277993068099,
          "validation_metric": 0.00048328781849704683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0691626989992074
        },
        {
          "epoch": 23,
          "updates": 1472,
          "samples_seen": 23552,
          "online_loss": 0.007419797226248193,
          "train_loss": 0.006737899035215378,
          "validation_loss": 0.004314305726438761,
          "train_metric": 0.00067861657589674,
          "validation_metric": 0.00043452103273011744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.112478970999291
        },
        {
          "epoch": 24,
          "updates": 1536,
          "samples_seen": 24576,
          "online_loss": 0.008811210521344037,
          "train_loss": 0.007847617380321026,
          "validation_loss": 0.006709723733365536,
          "train_metric": 0.0007903833175078034,
          "validation_metric": 0.0006757788360118866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.155836152998745
        },
        {
          "epoch": 25,
          "updates": 1600,
          "samples_seen": 25600,
          "online_loss": 0.01035352962117031,
          "train_loss": 0.008859946392476559,
          "validation_loss": 0.005991469603031874,
          "train_metric": 0.0008923413115553558,
          "validation_metric": 0.0006034388206899166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1984919549995539
        },
        {
          "epoch": 26,
          "updates": 1664,
          "samples_seen": 26624,
          "online_loss": 0.007424128142247355,
          "train_loss": 0.01274284441024065,
          "validation_loss": 0.010667464695870876,
          "train_metric": 0.001283412566408515,
          "validation_metric": 0.001074387808330357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2419333269990602
        },
        {
          "epoch": 27,
          "updates": 1728,
          "samples_seen": 27648,
          "online_loss": 0.008837456934543297,
          "train_loss": 0.013076290488243103,
          "validation_loss": 0.010273103602230549,
          "train_metric": 0.001316996174864471,
          "validation_metric": 0.0010346693452447653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2853876939989277
        },
        {
          "epoch": 28,
          "updates": 1792,
          "samples_seen": 28672,
          "online_loss": 0.015480354599276325,
          "train_loss": 0.010167221538722515,
          "validation_loss": 0.008841394446790218,
          "train_metric": 0.0010240052361041307,
          "validation_metric": 0.0008904729620553553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3286227869994036
        },
        {
          "epoch": 29,
          "updates": 1856,
          "samples_seen": 29696,
          "online_loss": 0.009996376227718429,
          "train_loss": 0.00548567995429039,
          "validation_loss": 0.0049051628448069096,
          "train_metric": 0.0005524977459572256,
          "validation_metric": 0.0004940300714224577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3727791349992913
        },
        {
          "epoch": 30,
          "updates": 1920,
          "samples_seen": 30720,
          "online_loss": 0.009597770405889605,
          "train_loss": 0.017969723790884018,
          "validation_loss": 0.013528532348573208,
          "train_metric": 0.0018098445143550634,
          "validation_metric": 0.0013625440187752247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4174167189994478
        },
        {
          "epoch": 31,
          "updates": 1984,
          "samples_seen": 31744,
          "online_loss": 0.008405326352203701,
          "train_loss": 0.006565312389284372,
          "validation_loss": 0.004434035625308752,
          "train_metric": 0.0006612342549487948,
          "validation_metric": 0.00044657988473773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4617080169991823
        },
        {
          "epoch": 32,
          "updates": 2048,
          "samples_seen": 32768,
          "online_loss": 0.009062504864232324,
          "train_loss": 0.005919359158724546,
          "validation_loss": 0.005023987032473087,
          "train_metric": 0.0005961760762147605,
          "validation_metric": 0.0005059975082986057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5064368049988843
        },
        {
          "epoch": 33,
          "updates": 2112,
          "samples_seen": 33792,
          "online_loss": 0.008290564923754573,
          "train_loss": 0.006588515359908342,
          "validation_loss": 0.004493733402341604,
          "train_metric": 0.0006635711179114878,
          "validation_metric": 0.00045259235776029527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5508231139992859
        },
        {
          "epoch": 34,
          "updates": 2176,
          "samples_seen": 34816,
          "online_loss": 0.007107449031536817,
          "train_loss": 0.004730650223791599,
          "validation_loss": 0.0034772916696965694,
          "train_metric": 0.000476453744340688,
          "validation_metric": 0.0003502200706861913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5954951219991926
        },
        {
          "epoch": 35,
          "updates": 2240,
          "samples_seen": 35840,
          "online_loss": 0.010305672046342806,
          "train_loss": 0.014995193108916283,
          "validation_loss": 0.01426345482468605,
          "train_metric": 0.001510260859504342,
          "validation_metric": 0.001436562859453261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6399688000001333
        },
        {
          "epoch": 36,
          "updates": 2304,
          "samples_seen": 36864,
          "online_loss": 0.010848984565654973,
          "train_loss": 0.0057310243137180805,
          "validation_loss": 0.004651216324418783,
          "train_metric": 0.0005772077711299062,
          "validation_metric": 0.00046845347969792783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.684275063998939
        },
        {
          "epoch": 37,
          "updates": 2368,
          "samples_seen": 37888,
          "online_loss": 0.007514664104746771,
          "train_loss": 0.006881095003336668,
          "validation_loss": 0.004433124791830778,
          "train_metric": 0.0006930386880412698,
          "validation_metric": 0.0004464880330488086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7273162269993918
        },
        {
          "epoch": 38,
          "updates": 2432,
          "samples_seen": 38912,
          "online_loss": 0.006451200300034543,
          "train_loss": 0.01110498420894146,
          "validation_loss": 0.007321333047002554,
          "train_metric": 0.0011184534523636103,
          "validation_metric": 0.0007373778498731554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7712818259988126
        },
        {
          "epoch": 39,
          "updates": 2496,
          "samples_seen": 39936,
          "online_loss": 0.0069781522870471235,
          "train_loss": 0.004393405746668577,
          "validation_loss": 0.0032984709832817316,
          "train_metric": 0.000442487740656361,
          "validation_metric": 0.00033220992190763354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8163077509998402
        },
        {
          "epoch": 40,
          "updates": 2560,
          "samples_seen": 40960,
          "online_loss": 0.006584257042050012,
          "train_loss": 0.009981204755604267,
          "validation_loss": 0.009126968681812286,
          "train_metric": 0.0010052702855318785,
          "validation_metric": 0.0009192348225042224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8601593280000088
        },
        {
          "epoch": 41,
          "updates": 2624,
          "samples_seen": 41984,
          "online_loss": 0.010061983021842025,
          "train_loss": 0.0044913082383573055,
          "validation_loss": 0.002628326416015625,
          "train_metric": 0.0004523480893112719,
          "validation_metric": 0.0002647154324222356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9054376160001993
        },
        {
          "epoch": 42,
          "updates": 2688,
          "samples_seen": 43008,
          "online_loss": 0.006987765137637325,
          "train_loss": 0.006876261439174414,
          "validation_loss": 0.0065422155894339085,
          "train_metric": 0.0006925518973730505,
          "validation_metric": 0.0006589079857803881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9499665799994546
        },
        {
          "epoch": 43,
          "updates": 2752,
          "samples_seen": 44032,
          "online_loss": 0.0062576308600910124,
          "train_loss": 0.004465088248252869,
          "validation_loss": 0.003558560973033309,
          "train_metric": 0.0004497073532547802,
          "validation_metric": 0.00035840520286001265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9948279479995108
        },
        {
          "epoch": 44,
          "updates": 2816,
          "samples_seen": 45056,
          "online_loss": 0.004728401679130911,
          "train_loss": 0.0043759532272815704,
          "validation_loss": 0.0028954697772860527,
          "train_metric": 0.0004407299857120961,
          "validation_metric": 0.00029162107966840267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0379198439986794
        },
        {
          "epoch": 45,
          "updates": 2880,
          "samples_seen": 46080,
          "online_loss": 0.006683852024252701,
          "train_loss": 0.014675063081085682,
          "validation_loss": 0.009597455151379108,
          "train_metric": 0.001478018588386476,
          "validation_metric": 0.0009666203986853361,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.082081302998631
        },
        {
          "epoch": 46,
          "updates": 2944,
          "samples_seen": 47104,
          "online_loss": 0.005745596100950934,
          "train_loss": 0.0037829275242984295,
          "validation_loss": 0.0027722904924303293,
          "train_metric": 0.00038100266829133034,
          "validation_metric": 0.0002792149316519499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.125992639999822
        },
        {
          "epoch": 47,
          "updates": 3008,
          "samples_seen": 48128,
          "online_loss": 0.006020910705501592,
          "train_loss": 0.008839322254061699,
          "validation_loss": 0.0070334430783987045,
          "train_metric": 0.0008902641711756587,
          "validation_metric": 0.0007083826349116862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1706211469991104
        },
        {
          "epoch": 48,
          "updates": 3072,
          "samples_seen": 49152,
          "online_loss": 0.004547376727032315,
          "train_loss": 0.00428582401946187,
          "validation_loss": 0.003157327650114894,
          "train_metric": 0.00043165250099264085,
          "validation_metric": 0.000317994476063177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.216350266999143
        },
        {
          "epoch": 49,
          "updates": 3136,
          "samples_seen": 50176,
          "online_loss": 0.006942263450582686,
          "train_loss": 0.005141069181263447,
          "validation_loss": 0.003975683357566595,
          "train_metric": 0.0005177896237000823,
          "validation_metric": 0.00040041626198217273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2624815119997947
        },
        {
          "epoch": 50,
          "updates": 3200,
          "samples_seen": 51200,
          "online_loss": 0.005434544714262302,
          "train_loss": 0.003877589013427496,
          "validation_loss": 0.002560761757194996,
          "train_metric": 0.00039053658838383853,
          "validation_metric": 0.0002579105494078249,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3095435459999862
        },
        {
          "epoch": 51,
          "updates": 3264,
          "samples_seen": 52224,
          "online_loss": 0.004601795514190599,
          "train_loss": 0.004701701458543539,
          "validation_loss": 0.0045182304456830025,
          "train_metric": 0.0004735381808131933,
          "validation_metric": 0.0004550596058834344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.38005981599963
        },
        {
          "epoch": 52,
          "updates": 3328,
          "samples_seen": 53248,
          "online_loss": 0.0046201620489227935,
          "train_loss": 0.004418829921633005,
          "validation_loss": 0.004040293861180544,
          "train_metric": 0.00044504841207526624,
          "validation_metric": 0.0004069235292263329,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.489849804998812
        },
        {
          "epoch": 53,
          "updates": 3392,
          "samples_seen": 54272,
          "online_loss": 0.008278696890101855,
          "train_loss": 0.005221362691372633,
          "validation_loss": 0.0035505935084074736,
          "train_metric": 0.0005258765886537731,
          "validation_metric": 0.00035760278115049005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.571757814999728
        },
        {
          "epoch": 54,
          "updates": 3456,
          "samples_seen": 55296,
          "online_loss": 0.004970595337681516,
          "train_loss": 0.005387685261666775,
          "validation_loss": 0.003142280038446188,
          "train_metric": 0.0005426278803497553,
          "validation_metric": 0.00031647892319597304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6611626209996757
        },
        {
          "epoch": 55,
          "updates": 3520,
          "samples_seen": 56320,
          "online_loss": 0.007255281524066959,
          "train_loss": 0.006570627447217703,
          "validation_loss": 0.003441527485847473,
          "train_metric": 0.0006617695325985551,
          "validation_metric": 0.0003466180060058832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.748053543999049
        },
        {
          "epoch": 56,
          "updates": 3584,
          "samples_seen": 57344,
          "online_loss": 0.0053205142030492425,
          "train_loss": 0.004314335994422436,
          "validation_loss": 0.0032241884618997574,
          "train_metric": 0.0004345240886323154,
          "validation_metric": 0.0003247284621465951,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.852848393999011
        },
        {
          "epoch": 57,
          "updates": 3648,
          "samples_seen": 58368,
          "online_loss": 0.009240228862836375,
          "train_loss": 0.003922522999346256,
          "validation_loss": 0.002622931031510234,
          "train_metric": 0.00039506214670836926,
          "validation_metric": 0.0002641720056999475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.062677531999725
        },
        {
          "epoch": 58,
          "updates": 3712,
          "samples_seen": 59392,
          "online_loss": 0.004579756043767702,
          "train_loss": 0.003659319831058383,
          "validation_loss": 0.001850242493674159,
          "train_metric": 0.00036855327198281884,
          "validation_metric": 0.00018634965817909688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2484849599986774
        },
        {
          "epoch": 59,
          "updates": 3776,
          "samples_seen": 60416,
          "online_loss": 0.003984138509167678,
          "train_loss": 0.0035012648440897465,
          "validation_loss": 0.0024417112581431866,
          "train_metric": 0.0003526346117723733,
          "validation_metric": 0.00024592020781710744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.433428593998542
        },
        {
          "epoch": 60,
          "updates": 3840,
          "samples_seen": 61440,
          "online_loss": 0.005328823183845088,
          "train_loss": 0.003341029165312648,
          "validation_loss": 0.0020822768565267324,
          "train_metric": 0.00033649621764197946,
          "validation_metric": 0.0002097192918881774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6180615249995753
        },
        {
          "epoch": 61,
          "updates": 3904,
          "samples_seen": 62464,
          "online_loss": 0.003959114797908114,
          "train_loss": 0.005015701055526733,
          "validation_loss": 0.004729214124381542,
          "train_metric": 0.0005051629850640893,
          "validation_metric": 0.000476309098303318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.856793432998529
        },
        {
          "epoch": 62,
          "updates": 3968,
          "samples_seen": 63488,
          "online_loss": 0.004089579194840098,
          "train_loss": 0.0027815019711852074,
          "validation_loss": 0.0016763299936428666,
          "train_metric": 0.00028014276176691055,
          "validation_metric": 0.00016883380885701627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.048176226999203
        },
        {
          "epoch": 63,
          "updates": 4032,
          "samples_seen": 64512,
          "online_loss": 0.004601376912660271,
          "train_loss": 0.008371973410248756,
          "validation_loss": 0.007237448822706938,
          "train_metric": 0.0008431944879703224,
          "validation_metric": 0.0007289292989298701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.232431151998753
        },
        {
          "epoch": 64,
          "updates": 4096,
          "samples_seen": 65536,
          "online_loss": 0.007482802419417567,
          "train_loss": 0.002640561433508992,
          "validation_loss": 0.0023297746665775776,
          "train_metric": 0.0002659476886037737,
          "validation_metric": 0.00023464634432457387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.416345867999553
        },
        {
          "epoch": 65,
          "updates": 4160,
          "samples_seen": 66560,
          "online_loss": 0.009836604079282552,
          "train_loss": 0.012143892236053944,
          "validation_loss": 0.009112924337387085,
          "train_metric": 0.001223088358528912,
          "validation_metric": 0.0009178203763440251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.60162219799895
        },
        {
          "epoch": 66,
          "updates": 4224,
          "samples_seen": 67584,
          "online_loss": 0.010706208254305238,
          "train_loss": 0.003347269259393215,
          "validation_loss": 0.002366446889936924,
          "train_metric": 0.0003371246566530317,
          "validation_metric": 0.00023833986779209226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.800743406998663
        },
        {
          "epoch": 67,
          "updates": 4288,
          "samples_seen": 68608,
          "online_loss": 0.0041712991005624644,
          "train_loss": 0.0029705215711146593,
          "validation_loss": 0.0019524048548191786,
          "train_metric": 0.00029918007203377783,
          "validation_metric": 0.0001966390700545162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.99598972899912
        },
        {
          "epoch": 68,
          "updates": 4352,
          "samples_seen": 69632,
          "online_loss": 0.005446394352475181,
          "train_loss": 0.00867989007383585,
          "validation_loss": 0.007559091784060001,
          "train_metric": 0.0008742067730054259,
          "validation_metric": 0.0007613240741193295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.178900054999758
        },
        {
          "epoch": 69,
          "updates": 4416,
          "samples_seen": 70656,
          "online_loss": 0.004712609716989391,
          "train_loss": 0.0024641137570142746,
          "validation_loss": 0.002070935443043709,
          "train_metric": 0.0002481765113770962,
          "validation_metric": 0.00020857703930232674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.361868727999536
        },
        {
          "epoch": 70,
          "updates": 4480,
          "samples_seen": 71680,
          "online_loss": 0.0053353536980011995,
          "train_loss": 0.003598368726670742,
          "validation_loss": 0.0026864390820264816,
          "train_metric": 0.0003624145465437323,
          "validation_metric": 0.00027056827093474567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.545836960998713
        },
        {
          "epoch": 71,
          "updates": 4544,
          "samples_seen": 72704,
          "online_loss": 0.006487410515092051,
          "train_loss": 0.00427223788574338,
          "validation_loss": 0.0034002484753727913,
          "train_metric": 0.0004302841261960566,
          "validation_metric": 0.0003424605820327997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.731307072999698
        },
        {
          "epoch": 72,
          "updates": 4608,
          "samples_seen": 73728,
          "online_loss": 0.0038497155237564584,
          "train_loss": 0.00345834088511765,
          "validation_loss": 0.003106774063780904,
          "train_metric": 0.0003483113832771778,
          "validation_metric": 0.0003129029064439237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.914763519000189
        },
        {
          "epoch": 73,
          "updates": 4672,
          "samples_seen": 74752,
          "online_loss": 0.005078895400856709,
          "train_loss": 0.013049841858446598,
          "validation_loss": 0.008919982239603996,
          "train_metric": 0.0013143322430551052,
          "validation_metric": 0.0008983879233710468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.098413522999181
        },
        {
          "epoch": 74,
          "updates": 4736,
          "samples_seen": 75776,
          "online_loss": 0.006015416821810504,
          "train_loss": 0.006485859863460064,
          "validation_loss": 0.005622646305710077,
          "train_metric": 0.000653232098557055,
          "validation_metric": 0.0005662923213094473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.283000882998749
        },
        {
          "epoch": 75,
          "updates": 4800,
          "samples_seen": 76800,
          "online_loss": 0.004642714549390803,
          "train_loss": 0.003063444048166275,
          "validation_loss": 0.0023032715544104576,
          "train_metric": 0.0003085389034822583,
          "validation_metric": 0.00023197707196231931,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.498774616999071
        },
        {
          "epoch": 76,
          "updates": 4864,
          "samples_seen": 77824,
          "online_loss": 0.0032353180770314793,
          "train_loss": 0.0024491206277161837,
          "validation_loss": 0.001662903348915279,
          "train_metric": 0.00024666645913384855,
          "validation_metric": 0.00016748151392675936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.725686798999959
        },
        {
          "epoch": 77,
          "updates": 4928,
          "samples_seen": 78848,
          "online_loss": 0.0035288714771013474,
          "train_loss": 0.004025372210890055,
          "validation_loss": 0.002818196779116988,
          "train_metric": 0.00040542075294069946,
          "validation_metric": 0.0002838384825736284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.926572238999142
        },
        {
          "epoch": 78,
          "updates": 4992,
          "samples_seen": 79872,
          "online_loss": 0.004764023032294062,
          "train_loss": 0.0030559422448277473,
          "validation_loss": 0.0019502791110426188,
          "train_metric": 0.00030778333893977106,
          "validation_metric": 0.00019642493862193078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.127512986999136
        },
        {
          "epoch": 79,
          "updates": 5056,
          "samples_seen": 80896,
          "online_loss": 0.007789909428083774,
          "train_loss": 0.004535281099379063,
          "validation_loss": 0.004028143361210823,
          "train_metric": 0.0004567769356071949,
          "validation_metric": 0.0004056998295709491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.365966624000066
        },
        {
          "epoch": 80,
          "updates": 5120,
          "samples_seen": 81920,
          "online_loss": 0.005262374678750348,
          "train_loss": 0.003964860923588276,
          "validation_loss": 0.0029106121510267258,
          "train_metric": 0.00039932626532390714,
          "validation_metric": 0.00029314623679965734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.552724175999174
        },
        {
          "epoch": 81,
          "updates": 5184,
          "samples_seen": 82944,
          "online_loss": 0.0034720643895980174,
          "train_loss": 0.003037024289369583,
          "validation_loss": 0.002528741490095854,
          "train_metric": 0.0003058779693674296,
          "validation_metric": 0.0002546855539549142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.74778469299963
        },
        {
          "epoch": 82,
          "updates": 5248,
          "samples_seen": 83968,
          "online_loss": 0.0055206258130056085,
          "train_loss": 0.004032639786601067,
          "validation_loss": 0.0024749296717345715,
          "train_metric": 0.00040615268517285585,
          "validation_metric": 0.0002492658677510917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.7857400339999
        },
        {
          "epoch": 83,
          "updates": 5312,
          "samples_seen": 84992,
          "online_loss": 0.0043859654447260255,
          "train_loss": 0.0029661557637155056,
          "validation_loss": 0.0019272128120064735,
          "train_metric": 0.00029874028405174613,
          "validation_metric": 0.00019410181266721338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.818023918998733
        },
        {
          "epoch": 84,
          "updates": 5376,
          "samples_seen": 86016,
          "online_loss": 0.00358266000284857,
          "train_loss": 0.0023910312447696924,
          "validation_loss": 0.0017106535378843546,
          "train_metric": 0.00024081594892777503,
          "validation_metric": 0.0001722907618386671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.849451871999918
        },
        {
          "epoch": 85,
          "updates": 5440,
          "samples_seen": 87040,
          "online_loss": 0.0033605572641590697,
          "train_loss": 0.005394404754042625,
          "validation_loss": 0.003675607033073902,
          "train_metric": 0.0005433047190308571,
          "validation_metric": 0.0003701936802826822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.880614428999252
        },
        {
          "epoch": 86,
          "updates": 5504,
          "samples_seen": 88064,
          "online_loss": 0.0046407893628384045,
          "train_loss": 0.00285455584526062,
          "validation_loss": 0.0017155753448605537,
          "train_metric": 0.0002875004429370165,
          "validation_metric": 0.00017278645827900618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.912016955999206
        },
        {
          "epoch": 87,
          "updates": 5568,
          "samples_seen": 89088,
          "online_loss": 0.003882064688241371,
          "train_loss": 0.0033572199754416943,
          "validation_loss": 0.0027819108217954636,
          "train_metric": 0.0003381269343663007,
          "validation_metric": 0.00028018385637551546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.943464071999188
        },
        {
          "epoch": 88,
          "updates": 5632,
          "samples_seen": 90112,
          "online_loss": 0.004051020854149101,
          "train_loss": 0.0031578191556036472,
          "validation_loss": 0.002636625897139311,
          "train_metric": 0.00031804401078261435,
          "validation_metric": 0.00026555132353678346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.975014536999879
        },
        {
          "epoch": 89,
          "updates": 5696,
          "samples_seen": 91136,
          "online_loss": 0.006655422046151216,
          "train_loss": 0.005760512314736843,
          "validation_loss": 0.004290754906833172,
          "train_metric": 0.0005801776424050331,
          "validation_metric": 0.00043214912875555456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.00789248699948
        },
        {
          "epoch": 90,
          "updates": 5760,
          "samples_seen": 92160,
          "online_loss": 0.004583923348036478,
          "train_loss": 0.004723873920738697,
          "validation_loss": 0.003866415238007903,
          "train_metric": 0.0004757712595164776,
          "validation_metric": 0.000389411230571568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.03930516299988
        },
        {
          "epoch": 91,
          "updates": 5824,
          "samples_seen": 93184,
          "online_loss": 0.0029618662265420426,
          "train_loss": 0.0018733730539679527,
          "validation_loss": 0.001382010756060481,
          "train_metric": 0.00018867927428800613,
          "validation_metric": 0.00013919107732363045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.071785533998991
        },
        {
          "epoch": 92,
          "updates": 5888,
          "samples_seen": 94208,
          "online_loss": 0.004818885888425939,
          "train_loss": 0.014789197593927383,
          "validation_loss": 0.010882163420319557,
          "train_metric": 0.001489513786509633,
          "validation_metric": 0.0010960117215290666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.10266563499863
        },
        {
          "epoch": 93,
          "updates": 5952,
          "samples_seen": 95232,
          "online_loss": 0.004985090478839993,
          "train_loss": 0.0025657706428319216,
          "validation_loss": 0.002222658833488822,
          "train_metric": 0.0002584150352049619,
          "validation_metric": 0.00022385804913938046,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.134499603000222
        },
        {
          "epoch": 94,
          "updates": 6016,
          "samples_seen": 96256,
          "online_loss": 0.006417918718625515,
          "train_loss": 0.00408197334036231,
          "validation_loss": 0.0029342484194785357,
          "train_metric": 0.00041112140752375126,
          "validation_metric": 0.0002955267555080354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.166175642998496
        },
        {
          "epoch": 95,
          "updates": 6080,
          "samples_seen": 97280,
          "online_loss": 0.0027732326404930063,
          "train_loss": 0.0022036663722246885,
          "validation_loss": 0.0013103829696774483,
          "train_metric": 0.00022194521443452686,
          "validation_metric": 0.00013197696534916759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.198154765999789
        },
        {
          "epoch": 96,
          "updates": 6144,
          "samples_seen": 98304,
          "online_loss": 0.00404850591576178,
          "train_loss": 0.002594135468825698,
          "validation_loss": 0.0015748029109090567,
          "train_metric": 0.0002612718672025949,
          "validation_metric": 0.00015860838175285608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.234221198999876
        },
        {
          "epoch": 97,
          "updates": 6208,
          "samples_seen": 99328,
          "online_loss": 0.003756221534786164,
          "train_loss": 0.0036464929580688477,
          "validation_loss": 0.003821144811809063,
          "train_metric": 0.00036726141115650535,
          "validation_metric": 0.0003848516789730638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.26726003099975
        },
        {
          "epoch": 98,
          "updates": 6272,
          "samples_seen": 100352,
          "online_loss": 0.004099623776482986,
          "train_loss": 0.002834114944562316,
          "validation_loss": 0.0019915886223316193,
          "train_metric": 0.0002854416670743376,
          "validation_metric": 0.00020058549125678837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.299032232000172
        },
        {
          "epoch": 99,
          "updates": 6336,
          "samples_seen": 101376,
          "online_loss": 0.002750217816355871,
          "train_loss": 0.0025713793002068996,
          "validation_loss": 0.0018483969615772367,
          "train_metric": 0.00025897991145029664,
          "validation_metric": 0.00018616377201396972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.330737102000057
        },
        {
          "epoch": 100,
          "updates": 6400,
          "samples_seen": 102400,
          "online_loss": 0.0024927342635692185,
          "train_loss": 0.0018196069868281484,
          "validation_loss": 0.0011499213287606835,
          "train_metric": 0.00018326415738556534,
          "validation_metric": 0.00011581585567910224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.364243492000242
        },
        {
          "epoch": 101,
          "updates": 6464,
          "samples_seen": 103424,
          "online_loss": 0.009023265430641914,
          "train_loss": 0.015969661995768547,
          "validation_loss": 0.013240575790405273,
          "train_metric": 0.0016084059607237577,
          "validation_metric": 0.001333542400971055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.397345094999764
        },
        {
          "epoch": 102,
          "updates": 6528,
          "samples_seen": 104448,
          "online_loss": 0.00564836786907108,
          "train_loss": 0.004069598391652107,
          "validation_loss": 0.0034624887630343437,
          "train_metric": 0.00040987509419210255,
          "validation_metric": 0.0003487291978672147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.43397123899922
        },
        {
          "epoch": 103,
          "updates": 6592,
          "samples_seen": 105472,
          "online_loss": 0.007236957029363111,
          "train_loss": 0.0045922077260911465,
          "validation_loss": 0.0031280340626835823,
          "train_metric": 0.00046251030289568007,
          "validation_metric": 0.0003150441334582865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.535397628998908
        },
        {
          "epoch": 104,
          "updates": 6656,
          "samples_seen": 106496,
          "online_loss": 0.003656768591326909,
          "train_loss": 0.0020081244874745607,
          "validation_loss": 0.0013998037902638316,
          "train_metric": 0.00020225094340275973,
          "validation_metric": 0.00014098308747634292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.58978970499993
        },
        {
          "epoch": 105,
          "updates": 6720,
          "samples_seen": 107520,
          "online_loss": 0.0031977894525425654,
          "train_loss": 0.002923674648627639,
          "validation_loss": 0.00220106216147542,
          "train_metric": 0.0002944618172477931,
          "validation_metric": 0.00022168291616253555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.640221918998577
        },
        {
          "epoch": 106,
          "updates": 6784,
          "samples_seen": 108544,
          "online_loss": 0.003579341725071572,
          "train_loss": 0.008362442255020142,
          "validation_loss": 0.005921755451709032,
          "train_metric": 0.0008422344690188766,
          "validation_metric": 0.0005964175797998905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.686880109999038
        },
        {
          "epoch": 107,
          "updates": 6848,
          "samples_seen": 109568,
          "online_loss": 0.0069666894369220245,
          "train_loss": 0.009133482351899147,
          "validation_loss": 0.006432458758354187,
          "train_metric": 0.0009198908810503781,
          "validation_metric": 0.0006478535942733288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.718138455998997
        },
        {
          "epoch": 108,
          "updates": 6912,
          "samples_seen": 110592,
          "online_loss": 0.005593484439941676,
          "train_loss": 0.007725676521658897,
          "validation_loss": 0.0072067417204380035,
          "train_metric": 0.0007781017920933664,
          "validation_metric": 0.0007258366094902158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.74956571399889
        },
        {
          "epoch": 109,
          "updates": 6976,
          "samples_seen": 111616,
          "online_loss": 0.0037276682655829063,
          "train_loss": 0.0024847157765179873,
          "validation_loss": 0.0017546082381159067,
          "train_metric": 0.00025025143986567855,
          "validation_metric": 0.0001767177163856104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.781406680000146
        },
        {
          "epoch": 110,
          "updates": 7040,
          "samples_seen": 112640,
          "online_loss": 0.0028266119466024975,
          "train_loss": 0.0075097163207829,
          "validation_loss": 0.0043962933123111725,
          "train_metric": 0.0007563511026091874,
          "validation_metric": 0.00044277854613028467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.812604613000076
        },
        {
          "epoch": 111,
          "updates": 7104,
          "samples_seen": 113664,
          "online_loss": 0.004872928667282395,
          "train_loss": 0.0016993175959214568,
          "validation_loss": 0.0012184997322037816,
          "train_metric": 0.0001711490476736799,
          "validation_metric": 0.0001227228349307552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.844500604000132
        },
        {
          "epoch": 112,
          "updates": 7168,
          "samples_seen": 114688,
          "online_loss": 0.0026508334452728377,
          "train_loss": 0.0018373558996245265,
          "validation_loss": 0.00124374870210886,
          "train_metric": 0.00018505177285987884,
          "validation_metric": 0.00012526579666882753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.876293823999731
        },
        {
          "epoch": 113,
          "updates": 7232,
          "samples_seen": 115712,
          "online_loss": 0.0027652755088638514,
          "train_loss": 0.0016986675327643752,
          "validation_loss": 0.0010867256205528975,
          "train_metric": 0.0001710835931589827,
          "validation_metric": 0.00010945102258119732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.908259381998505
        },
        {
          "epoch": 114,
          "updates": 7296,
          "samples_seen": 116736,
          "online_loss": 0.004180463824013714,
          "train_loss": 0.003471780801191926,
          "validation_loss": 0.0029710298404097557,
          "train_metric": 0.0003496650024317205,
          "validation_metric": 0.00029923126567155123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.939861248998568
        },
        {
          "epoch": 115,
          "updates": 7360,
          "samples_seen": 117760,
          "online_loss": 0.002663700509174305,
          "train_loss": 0.002227255841717124,
          "validation_loss": 0.002144285012036562,
          "train_metric": 0.00022432103287428617,
          "validation_metric": 0.00021596452279482037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.97171273299864
        },
        {
          "epoch": 116,
          "updates": 7424,
          "samples_seen": 118784,
          "online_loss": 0.003340711095233928,
          "train_loss": 0.0032345952931791544,
          "validation_loss": 0.0020003581885248423,
          "train_metric": 0.00032577660749666393,
          "validation_metric": 0.00020146874885540456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.003283448999355
        },
        {
          "epoch": 117,
          "updates": 7488,
          "samples_seen": 119808,
          "online_loss": 0.00495922394838999,
          "train_loss": 0.0018841223791241646,
          "validation_loss": 0.0013709280174225569,
          "train_metric": 0.00018976190767716616,
          "validation_metric": 0.00013807485811412334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.035955481998826
        },
        {
          "epoch": 118,
          "updates": 7552,
          "samples_seen": 120832,
          "online_loss": 0.0026092389068708144,
          "train_loss": 0.003442693268880248,
          "validation_loss": 0.002203419106081128,
          "train_metric": 0.0003467354108579457,
          "validation_metric": 0.00022192028700374067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.069073691000085
        },
        {
          "epoch": 119,
          "updates": 7616,
          "samples_seen": 121856,
          "online_loss": 0.004062540064296627,
          "train_loss": 0.0029082810506224632,
          "validation_loss": 0.003035532543435693,
          "train_metric": 0.0002929114270955324,
          "validation_metric": 0.0003057277062907815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.101465947998804
        },
        {
          "epoch": 120,
          "updates": 7680,
          "samples_seen": 122880,
          "online_loss": 0.003607565578477079,
          "train_loss": 0.0025224084965884686,
          "validation_loss": 0.0021505970507860184,
          "train_metric": 0.00025404777261428535,
          "validation_metric": 0.00021660026686731726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.133102195999527
        },
        {
          "epoch": 121,
          "updates": 7744,
          "samples_seen": 123904,
          "online_loss": 0.0050718171760308906,
          "train_loss": 0.0044335005804896355,
          "validation_loss": 0.0028078947216272354,
          "train_metric": 0.0004465259553398937,
          "validation_metric": 0.00028280087281018496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.16505506700014
        },
        {
          "epoch": 122,
          "updates": 7808,
          "samples_seen": 124928,
          "online_loss": 0.006047127636975347,
          "train_loss": 0.0025961282663047314,
          "validation_loss": 0.0018332140753045678,
          "train_metric": 0.00026147253811359406,
          "validation_metric": 0.00018463459855411202,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.198843706999469
        },
        {
          "epoch": 123,
          "updates": 7872,
          "samples_seen": 125952,
          "online_loss": 0.002754393559825985,
          "train_loss": 0.0020942320115864277,
          "validation_loss": 0.0013616011710837483,
          "train_metric": 0.00021092337556183338,
          "validation_metric": 0.0001371355028823018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.23031896700013
        },
        {
          "epoch": 124,
          "updates": 7936,
          "samples_seen": 126976,
          "online_loss": 0.004310360535328073,
          "train_loss": 0.001583161181770265,
          "validation_loss": 0.0012518602889031172,
          "train_metric": 0.00015945018094498664,
          "validation_metric": 0.00012608279939740896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.265052654000101
        },
        {
          "epoch": 125,
          "updates": 8000,
          "samples_seen": 128000,
          "online_loss": 0.006415617981929245,
          "train_loss": 0.0037046982906758785,
          "validation_loss": 0.0034112753346562386,
          "train_metric": 0.0003731236793100834,
          "validation_metric": 0.0003435711842030287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.297399651999513
        },
        {
          "epoch": 126,
          "updates": 8064,
          "samples_seen": 129024,
          "online_loss": 0.00694111169923417,
          "train_loss": 0.0023647926282137632,
          "validation_loss": 0.0015630663838237524,
          "train_metric": 0.0002381732629146427,
          "validation_metric": 0.00015742632967885584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.329857122000249
        },
        {
          "epoch": 127,
          "updates": 8128,
          "samples_seen": 130048,
          "online_loss": 0.004857647656535846,
          "train_loss": 0.0021193423308432102,
          "validation_loss": 0.00136666651815176,
          "train_metric": 0.00021345239656511694,
          "validation_metric": 0.00013764563482254744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.36296170499918
        },
        {
          "epoch": 128,
          "updates": 8192,
          "samples_seen": 131072,
          "online_loss": 0.003607344855709016,
          "train_loss": 0.004001163877546787,
          "validation_loss": 0.0022484289947897196,
          "train_metric": 0.00040298255044035614,
          "validation_metric": 0.00022645355784334242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.395199531998514
        },
        {
          "epoch": 129,
          "updates": 8256,
          "samples_seen": 132096,
          "online_loss": 0.0031861188904258597,
          "train_loss": 0.005948570556938648,
          "validation_loss": 0.004568220116198063,
          "train_metric": 0.0005991182988509536,
          "validation_metric": 0.0004600943939294666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.427260849999584
        },
        {
          "epoch": 130,
          "updates": 8320,
          "samples_seen": 133120,
          "online_loss": 0.005046530253821402,
          "train_loss": 0.009198085404932499,
          "validation_loss": 0.009069670923054218,
          "train_metric": 0.0009263973915949464,
          "validation_metric": 0.0009134639985859394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.46069762299885
        },
        {
          "epoch": 131,
          "updates": 8384,
          "samples_seen": 134144,
          "online_loss": 0.00347211183429863,
          "train_loss": 0.003217935562133789,
          "validation_loss": 0.002045743865892291,
          "train_metric": 0.0003240986552555114,
          "validation_metric": 0.00020603984012268484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.492598483999245
        },
        {
          "epoch": 132,
          "updates": 8448,
          "samples_seen": 135168,
          "online_loss": 0.0035932182501028365,
          "train_loss": 0.004441762808710337,
          "validation_loss": 0.003847437910735607,
          "train_metric": 0.00044735809206031263,
          "validation_metric": 0.000387499836506322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.525087502999668
        },
        {
          "epoch": 133,
          "updates": 8512,
          "samples_seen": 136192,
          "online_loss": 0.004275654991943156,
          "train_loss": 0.004963050130754709,
          "validation_loss": 0.0038479636423289776,
          "train_metric": 0.0004998602089472115,
          "validation_metric": 0.00038755283458158374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.556879413999923
        }
      ],
      "best": {
        "epoch": 113,
        "updates": 7232
      },
      "best_validation_metric": 0.00010945102258119732,
      "config": {
        "label": "Batch 16",
        "batch": 16,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010461884274890318
    },
    "5c044ad193b4_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7750606071203947,
          "train_loss": 0.38536715507507324,
          "validation_loss": 0.30186760425567627,
          "train_metric": 0.03881276398897171,
          "validation_metric": 0.030403001233935356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.012762815998939914
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.2060750015079975,
          "train_loss": 0.12688414752483368,
          "validation_loss": 0.1095275804400444,
          "train_metric": 0.01277930662035942,
          "validation_metric": 0.011031216941773891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02530561999992642
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.1240247250534594,
          "train_loss": 0.09904789179563522,
          "validation_loss": 0.08297248184680939,
          "train_metric": 0.009975740686058998,
          "validation_metric": 0.008356684818863869,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03777097599959234
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09276640950702131,
          "train_loss": 0.0849667638540268,
          "validation_loss": 0.06827860325574875,
          "train_metric": 0.00855754129588604,
          "validation_metric": 0.006876770872622728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05016154999975697
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07504639495164156,
          "train_loss": 0.06760113686323166,
          "validation_loss": 0.055644359439611435,
          "train_metric": 0.006808537989854813,
          "validation_metric": 0.005604295991361141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06299039499936043
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06591126136481762,
          "train_loss": 0.0652155801653862,
          "validation_loss": 0.052817028015851974,
          "train_metric": 0.0065682739950716496,
          "validation_metric": 0.005319537594914436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0756774729998142
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06305562076158822,
          "train_loss": 0.06514137238264084,
          "validation_loss": 0.05300118774175644,
          "train_metric": 0.00656080013141036,
          "validation_metric": 0.005338085815310478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08796611899924756
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06273406581021845,
          "train_loss": 0.05920339375734329,
          "validation_loss": 0.04777815192937851,
          "train_metric": 0.005962749011814594,
          "validation_metric": 0.0048120408318936825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10037986599854776
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05564656085334718,
          "train_loss": 0.052560701966285706,
          "validation_loss": 0.0434177927672863,
          "train_metric": 0.005293721333146095,
          "validation_metric": 0.004372881259769201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11258570599966333
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05131560959853232,
          "train_loss": 0.051526989787817,
          "validation_loss": 0.04277803748846054,
          "train_metric": 0.0051896097138524055,
          "validation_metric": 0.004308447707444429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12495047299853468
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04965876357164234,
          "train_loss": 0.0480007603764534,
          "validation_loss": 0.03882240504026413,
          "train_metric": 0.004834461025893688,
          "validation_metric": 0.003910050727427006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13718341299863823
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04525484796613455,
          "train_loss": 0.041656482964754105,
          "validation_loss": 0.034643612802028656,
          "train_metric": 0.004195488058030605,
          "validation_metric": 0.0034891776740550995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1495563439984835
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.0406425753608346,
          "train_loss": 0.03904619812965393,
          "validation_loss": 0.032580845057964325,
          "train_metric": 0.003932590130716562,
          "validation_metric": 0.003281423356384039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16227347899985034
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03730706369969994,
          "train_loss": 0.03512808308005333,
          "validation_loss": 0.029410352930426598,
          "train_metric": 0.003537971992045641,
          "validation_metric": 0.0029621028807014227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1746723709984508
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03614106320310384,
          "train_loss": 0.03985850140452385,
          "validation_loss": 0.033791474997997284,
          "train_metric": 0.004014401696622372,
          "validation_metric": 0.0034033532720059156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18633238099937444
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.034006054629571736,
          "train_loss": 0.03577851876616478,
          "validation_loss": 0.03025796450674534,
          "train_metric": 0.0036034814547747374,
          "validation_metric": 0.0030474714003503323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1978571039999224
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.030239578511100262,
          "train_loss": 0.02710607275366783,
          "validation_loss": 0.023179994896054268,
          "train_metric": 0.0027300245128571987,
          "validation_metric": 0.00233460427261889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20978566799931286
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029969822440762073,
          "train_loss": 0.03654274344444275,
          "validation_loss": 0.028627360239624977,
          "train_metric": 0.003680451540276408,
          "validation_metric": 0.0028832426760345697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22179265299928375
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.031163847947027534,
          "train_loss": 0.02538416162133217,
          "validation_loss": 0.019166603684425354,
          "train_metric": 0.002556599909439683,
          "validation_metric": 0.0019303900189697742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23401714200008428
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.023735167924314737,
          "train_loss": 0.02233070135116577,
          "validation_loss": 0.018863607197999954,
          "train_metric": 0.0022490667179226875,
          "validation_metric": 0.001899873255752027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24638067499836325
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023142108839238063,
          "train_loss": 0.01990034058690071,
          "validation_loss": 0.01686880923807621,
          "train_metric": 0.002004289533942938,
          "validation_metric": 0.0016989647410809994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2588599129994691
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02312045422149822,
          "train_loss": 0.020990870893001556,
          "validation_loss": 0.016307419165968895,
          "train_metric": 0.002114123897626996,
          "validation_metric": 0.0016424234490841627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2710824349996983
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.022139596985653043,
          "train_loss": 0.020363621413707733,
          "validation_loss": 0.01820952072739601,
          "train_metric": 0.0020509492605924606,
          "validation_metric": 0.0018339961534366012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28303431999847817
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.018391056975815445,
          "train_loss": 0.01689496450126171,
          "validation_loss": 0.01443193294107914,
          "train_metric": 0.001701599103398621,
          "validation_metric": 0.001453531440347433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29554312999971444
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019256686326116323,
          "train_loss": 0.019413551315665245,
          "validation_loss": 0.014614056795835495,
          "train_metric": 0.0019552616868168116,
          "validation_metric": 0.0014718743041157722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30737604399837437
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.018628374149557203,
          "train_loss": 0.021552782505750656,
          "validation_loss": 0.02029510959982872,
          "train_metric": 0.002170717343688011,
          "validation_metric": 0.0020440490916371346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3188643809990026
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.017534870916279033,
          "train_loss": 0.015243593603372574,
          "validation_loss": 0.013040601275861263,
          "train_metric": 0.001535278744995594,
          "validation_metric": 0.001313401386141777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33126219699988724
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015402383840410039,
          "train_loss": 0.014646788127720356,
          "validation_loss": 0.012762422673404217,
          "train_metric": 0.0014751707203686237,
          "validation_metric": 0.0012853844091296196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3435950209986913
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016872672975296155,
          "train_loss": 0.018696356564760208,
          "validation_loss": 0.015021326020359993,
          "train_metric": 0.0018830286571756005,
          "validation_metric": 0.001512893009930849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35554667899850756
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.016743629763368517,
          "train_loss": 0.01551545225083828,
          "validation_loss": 0.013348198495805264,
          "train_metric": 0.0015626593958586454,
          "validation_metric": 0.001344381715171039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3673225909988105
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.021813345549162477,
          "train_loss": 0.014776387251913548,
          "validation_loss": 0.011951805092394352,
          "train_metric": 0.001488223671913147,
          "validation_metric": 0.0012037419946864247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37941377899915096
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.018929659272544086,
          "train_loss": 0.014764338731765747,
          "validation_loss": 0.010697023011744022,
          "train_metric": 0.0014870099257677794,
          "validation_metric": 0.001077365013770759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3921295479995024
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.0153573855641298,
          "train_loss": 0.012026515789330006,
          "validation_loss": 0.009608716703951359,
          "train_metric": 0.0012112664990127087,
          "validation_metric": 0.0009677548077888787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4044959389993892
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013202551752328873,
          "train_loss": 0.01557985134422779,
          "validation_loss": 0.010827160440385342,
          "train_metric": 0.0015691454755142331,
          "validation_metric": 0.0010904718656092882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41617666199999803
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014358282889588736,
          "train_loss": 0.015606962144374847,
          "validation_loss": 0.014100464060902596,
          "train_metric": 0.0015718759968876839,
          "validation_metric": 0.001420147018507123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42780261499865446
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.016984775924356654,
          "train_loss": 0.02222522161900997,
          "validation_loss": 0.01624390110373497,
          "train_metric": 0.002238442888483405,
          "validation_metric": 0.001636026194319129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43937805099994875
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.012299630441702902,
          "train_loss": 0.011116322129964828,
          "validation_loss": 0.009583269245922565,
          "train_metric": 0.0011195952538400888,
          "validation_metric": 0.0009651916334405541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45123246099865355
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011669085710309446,
          "train_loss": 0.009766273200511932,
          "validation_loss": 0.007932213135063648,
          "train_metric": 0.0009836232056841254,
          "validation_metric": 0.0007989035220816731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46367447299962805
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01077001288649626,
          "train_loss": 0.009837810881435871,
          "validation_loss": 0.00763594638556242,
          "train_metric": 0.0009908282663673162,
          "validation_metric": 0.0007690644706599414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4764903269988281
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012661474363994785,
          "train_loss": 0.012773655354976654,
          "validation_loss": 0.009138710796833038,
          "train_metric": 0.0012865157332271338,
          "validation_metric": 0.0009204175439663231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4885734709987446
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010784950369270518,
          "train_loss": 0.009931527078151703,
          "validation_loss": 0.007522317115217447,
          "train_metric": 0.0010002671042457223,
          "validation_metric": 0.0007576202624477446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.500754913999117
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010854612351977266,
          "train_loss": 0.009563778527081013,
          "validation_loss": 0.007020589429885149,
          "train_metric": 0.0009632286964915693,
          "validation_metric": 0.0007070880965329707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5132092449985066
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011339242788380943,
          "train_loss": 0.0087605444714427,
          "validation_loss": 0.007057554554194212,
          "train_metric": 0.0008823300013318658,
          "validation_metric": 0.0007108110003173351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5254339719995187
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.00978471252892632,
          "train_loss": 0.00902597513049841,
          "validation_loss": 0.0073502580635249615,
          "train_metric": 0.0009090631501749158,
          "validation_metric": 0.0007402910850942135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.536997994999183
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013840032886946574,
          "train_loss": 0.010268403217196465,
          "validation_loss": 0.008520981296896935,
          "train_metric": 0.0010341958841308951,
          "validation_metric": 0.0008582019945606589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5490574929990544
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.012792111374437809,
          "train_loss": 0.010196718387305737,
          "validation_loss": 0.007601565681397915,
          "train_metric": 0.0010269759222865105,
          "validation_metric": 0.0007656019297428429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5610209819988086
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011586189008085057,
          "train_loss": 0.012726811692118645,
          "validation_loss": 0.011708073318004608,
          "train_metric": 0.0012817978858947754,
          "validation_metric": 0.00117919419426471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.573129546999553
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.009888891261653043,
          "train_loss": 0.014078420586884022,
          "validation_loss": 0.013561593368649483,
          "train_metric": 0.0014179269783198833,
          "validation_metric": 0.0013658740790560842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5847851109992916
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.01211502478690818,
          "train_loss": 0.010559041053056717,
          "validation_loss": 0.008403655141592026,
          "train_metric": 0.001063467818312347,
          "validation_metric": 0.0008463853155262768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5968669049998425
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.011117999012640212,
          "train_loss": 0.00816005002707243,
          "validation_loss": 0.006046057213097811,
          "train_metric": 0.0008218504954129457,
          "validation_metric": 0.0006089368253014982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6091163059991231
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.00928575212310534,
          "train_loss": 0.012129377573728561,
          "validation_loss": 0.01141818892210722,
          "train_metric": 0.0012216264149174094,
          "validation_metric": 0.0011499980464577675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6213109939999413
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008625807182397693,
          "train_loss": 0.00884421169757843,
          "validation_loss": 0.0063227578066289425,
          "train_metric": 0.0008907566079869866,
          "validation_metric": 0.0006368050235323608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6329813019983703
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008779473515460268,
          "train_loss": 0.010252740234136581,
          "validation_loss": 0.00750874076038599,
          "train_metric": 0.0010326184565201402,
          "validation_metric": 0.0007562528480775654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6445814719991176
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0091602051834343,
          "train_loss": 0.01068117842078209,
          "validation_loss": 0.010125876404345036,
          "train_metric": 0.0010757691925391555,
          "validation_metric": 0.0010198411764577031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6568648949996714
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.013468224395182915,
          "train_loss": 0.00999869592487812,
          "validation_loss": 0.007742003537714481,
          "train_metric": 0.0010070321150124073,
          "validation_metric": 0.0007797462749294937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6684954189986456
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.010853008774574846,
          "train_loss": 0.008161451667547226,
          "validation_loss": 0.0058862557634711266,
          "train_metric": 0.0008219915325753391,
          "validation_metric": 0.0005928422324359417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6807773539985646
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008787122336798348,
          "train_loss": 0.007462658453732729,
          "validation_loss": 0.006129884626716375,
          "train_metric": 0.0007516115438193083,
          "validation_metric": 0.0006173795554786921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6926497399999789
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007215693418402225,
          "train_loss": 0.006528432480990887,
          "validation_loss": 0.005115540232509375,
          "train_metric": 0.0006575197330676019,
          "validation_metric": 0.0005152184749022126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7051688269984879
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007845858577638865,
          "train_loss": 0.0069840820506215096,
          "validation_loss": 0.006143107544630766,
          "train_metric": 0.0007034111185930669,
          "validation_metric": 0.0006187112885527313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7170204359990748
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007135726671549492,
          "train_loss": 0.006284323520958424,
          "validation_loss": 0.005035600624978542,
          "train_metric": 0.0006329340394586325,
          "validation_metric": 0.0005071673076599836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.729623207998884
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.009110017359489575,
          "train_loss": 0.006696862168610096,
          "validation_loss": 0.005621404852718115,
          "train_metric": 0.0006744833663105965,
          "validation_metric": 0.0005661673494614661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7414981769998121
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007818591664545238,
          "train_loss": 0.006637724116444588,
          "validation_loss": 0.005256288219243288,
          "train_metric": 0.0006685271509923041,
          "validation_metric": 0.0005293941358104348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7529287799989106
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006970037764403969,
          "train_loss": 0.006174634676426649,
          "validation_loss": 0.004937969613820314,
          "train_metric": 0.0006218865746632218,
          "validation_metric": 0.0004973342875018716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.765209192999464
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007135511885280721,
          "train_loss": 0.009442456066608429,
          "validation_loss": 0.008755910210311413,
          "train_metric": 0.0009510095696896315,
          "validation_metric": 0.0008818632340990007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.776969108999765
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008260824164608493,
          "train_loss": 0.008309222757816315,
          "validation_loss": 0.0070204706862568855,
          "train_metric": 0.0008368745329789817,
          "validation_metric": 0.0007070761639624834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7889315359989268
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0069822732475586236,
          "train_loss": 0.0067450315691530704,
          "validation_loss": 0.004807147663086653,
          "train_metric": 0.000679334974847734,
          "validation_metric": 0.0004841582849621773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8012147829995229
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008227432139392477,
          "train_loss": 0.00827042106539011,
          "validation_loss": 0.005731015466153622,
          "train_metric": 0.0008329665288329124,
          "validation_metric": 0.0005772068398073316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8129991529986
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00791827542707324,
          "train_loss": 0.006597149185836315,
          "validation_loss": 0.00453611183911562,
          "train_metric": 0.000664440740365535,
          "validation_metric": 0.00045686058001592755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8252837
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0060867796710226685,
          "train_loss": 0.006203494966030121,
          "validation_loss": 0.005195828154683113,
          "train_metric": 0.0006247933488339186,
          "validation_metric": 0.0005233047413639724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8368050999997649
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006417378521291539,
          "train_loss": 0.005572378169745207,
          "validation_loss": 0.0043088579550385475,
          "train_metric": 0.0005612294771708548,
          "validation_metric": 0.000433972425526008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8491546589993959
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.007080948373186402,
          "train_loss": 0.006128726992756128,
          "validation_loss": 0.00446541141718626,
          "train_metric": 0.0006172629073262215,
          "validation_metric": 0.0004497398913372308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8613863889986533
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006899416650412604,
          "train_loss": 0.005498890299350023,
          "validation_loss": 0.0047007291577756405,
          "train_metric": 0.0005538280820474029,
          "validation_metric": 0.00047344021731987596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8735639909991733
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.010936821228824556,
          "train_loss": 0.01833980344235897,
          "validation_loss": 0.014932270161807537,
          "train_metric": 0.0018471179064363241,
          "validation_metric": 0.0015039234422147274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8851871599999868
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.010608561802655458,
          "train_loss": 0.006225930526852608,
          "validation_loss": 0.005035202018916607,
          "train_metric": 0.0006270530284382403,
          "validation_metric": 0.0005071270861662924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8976141159982944
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006480617972556502,
          "train_loss": 0.006028239615261555,
          "validation_loss": 0.004369248636066914,
          "train_metric": 0.0006071422249078751,
          "validation_metric": 0.0004400547768454999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9120192059999681
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.009216629012371413,
          "train_loss": 0.006511844228953123,
          "validation_loss": 0.005704014096409082,
          "train_metric": 0.0006558491731993854,
          "validation_metric": 0.0005744873778894544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9246584829998028
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007850948211853392,
          "train_loss": 0.009311886504292488,
          "validation_loss": 0.007155830971896648,
          "train_metric": 0.0009378591785207391,
          "validation_metric": 0.0007207090966403484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9381915829999343
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.008934477751608938,
          "train_loss": 0.012272932566702366,
          "validation_loss": 0.01282534096390009,
          "train_metric": 0.0012360847322270274,
          "validation_metric": 0.001291721360757947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9502145419992303
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00737155377282761,
          "train_loss": 0.005852990318089724,
          "validation_loss": 0.004597283899784088,
          "train_metric": 0.0005894917994737625,
          "validation_metric": 0.0004630216280929744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.962366909998309
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.006133655791927595,
          "train_loss": 0.005515925120562315,
          "validation_loss": 0.004237538203597069,
          "train_metric": 0.0005555438692681491,
          "validation_metric": 0.0004267893673386425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0577063919990906
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006042832348612137,
          "train_loss": 0.0049070012755692005,
          "validation_loss": 0.004042955581098795,
          "train_metric": 0.0004942152881994843,
          "validation_metric": 0.00040719169192016125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0709459259996947
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.005204449022130575,
          "train_loss": 0.005303549580276012,
          "validation_loss": 0.0037937252782285213,
          "train_metric": 0.0005341540672816336,
          "validation_metric": 0.0003820901329163462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.083118229998945
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006180046577355824,
          "train_loss": 0.006082515232264996,
          "validation_loss": 0.006012081168591976,
          "train_metric": 0.0006126086227595806,
          "validation_metric": 0.0006055147969163954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.095398617999308
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.007441859052050859,
          "train_loss": 0.008779780007898808,
          "validation_loss": 0.006213271990418434,
          "train_metric": 0.0008842672104947269,
          "validation_metric": 0.0006257780478335917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1072723339984805
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.006668586407613475,
          "train_loss": 0.005109636578708887,
          "validation_loss": 0.004553762264549732,
          "train_metric": 0.0005146238836459816,
          "validation_metric": 0.0004586382710840553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.11856317999991
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.009025622712215409,
          "train_loss": 0.006044340785592794,
          "validation_loss": 0.004663453903049231,
          "train_metric": 0.0006087638903409243,
          "validation_metric": 0.00046968599781394005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1300283139989915
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005890286578505766,
          "train_loss": 0.004747289698570967,
          "validation_loss": 0.003448423696681857,
          "train_metric": 0.00047812960110604763,
          "validation_metric": 0.00034731256891973317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1417282249985874
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.005285567880491726,
          "train_loss": 0.005177311599254608,
          "validation_loss": 0.003495278302580118,
          "train_metric": 0.0005214398261159658,
          "validation_metric": 0.00035203166771680117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1529005550000875
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0053627890010830015,
          "train_loss": 0.004558279644697905,
          "validation_loss": 0.0040615759789943695,
          "train_metric": 0.0004590932803694159,
          "validation_metric": 0.00040906708454713225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1643528519998654
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.005066284713393543,
          "train_loss": 0.004958731587976217,
          "validation_loss": 0.0037932542618364096,
          "train_metric": 0.000499425339512527,
          "validation_metric": 0.0003820426354650408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1755964449985186
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004962477920344099,
          "train_loss": 0.004719328135251999,
          "validation_loss": 0.0037969392724335194,
          "train_metric": 0.00047531345626339316,
          "validation_metric": 0.000382413825718686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1867256149998866
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004873314610449597,
          "train_loss": 0.004178574774414301,
          "validation_loss": 0.003482517320662737,
          "train_metric": 0.0004208507889416069,
          "validation_metric": 0.00035074641346000135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1984951539998292
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0052191740469424985,
          "train_loss": 0.004869947675615549,
          "validation_loss": 0.0042724465020000935,
          "train_metric": 0.0004904833040200174,
          "validation_metric": 0.00043030513916164637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2110727399995085
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004895404661510838,
          "train_loss": 0.004149910062551498,
          "validation_loss": 0.003166336566209793,
          "train_metric": 0.0004179638053756207,
          "validation_metric": 0.0003189018170814961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.223447568998381
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.005768019502284005,
          "train_loss": 0.00506983045488596,
          "validation_loss": 0.004650712013244629,
          "train_metric": 0.0005106147145852447,
          "validation_metric": 0.0004684026644099504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2351106999994954
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.005765228270320222,
          "train_loss": 0.004359415732324123,
          "validation_loss": 0.0034883166663348675,
          "train_metric": 0.00043906443170271814,
          "validation_metric": 0.0003513304691296071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.246699646999332
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005973441817332059,
          "train_loss": 0.004715489223599434,
          "validation_loss": 0.004207362420856953,
          "train_metric": 0.00047492681187577546,
          "validation_metric": 0.0004237501125317067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2587828359992272
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.005162642170034815,
          "train_loss": 0.004261848516762257,
          "validation_loss": 0.0031990520656108856,
          "train_metric": 0.0004292377270758152,
          "validation_metric": 0.0003221968072466552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2702911389988003
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.004465270045329817,
          "train_loss": 0.004232842940837145,
          "validation_loss": 0.003191899275407195,
          "train_metric": 0.00042631648830138147,
          "validation_metric": 0.00032147637102752924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.281693467999503
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004748770137666725,
          "train_loss": 0.006879557855427265,
          "validation_loss": 0.004877622704952955,
          "train_metric": 0.00069288385566324,
          "validation_metric": 0.0004912563017569482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2934092979994603
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005494646236911649,
          "train_loss": 0.004157326649874449,
          "validation_loss": 0.003637386718764901,
          "train_metric": 0.0004187106969766319,
          "validation_metric": 0.00036634429125115275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3051845989994035
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0061596137238666415,
          "train_loss": 0.01087939739227295,
          "validation_loss": 0.010176428593695164,
          "train_metric": 0.0010957331396639347,
          "validation_metric": 0.001024932600557804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3168906579994655
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0053262883448041975,
          "train_loss": 0.005102952476590872,
          "validation_loss": 0.003219985868781805,
          "train_metric": 0.0005139507120475173,
          "validation_metric": 0.00032430514693260193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3289986179988773
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004766024787386414,
          "train_loss": 0.003594722831621766,
          "validation_loss": 0.0028451078105717897,
          "train_metric": 0.0003620473144110292,
          "validation_metric": 0.00028654883499257267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3411963929993362
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004065916029503569,
          "train_loss": 0.006581991910934448,
          "validation_loss": 0.0065535674802958965,
          "train_metric": 0.0006629141862504184,
          "validation_metric": 0.0006600513006560504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3530582129988034
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004911221309157554,
          "train_loss": 0.00416746037080884,
          "validation_loss": 0.003098068991675973,
          "train_metric": 0.0004197313392069191,
          "validation_metric": 0.0003120261535514146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.364949748998697
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004501591378357261,
          "train_loss": 0.0038424625527113676,
          "validation_loss": 0.003060749266296625,
          "train_metric": 0.00038699875585734844,
          "validation_metric": 0.0003082674811594188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3766447989983135
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.004057249549077824,
          "train_loss": 0.004752623848617077,
          "validation_loss": 0.004476326052099466,
          "train_metric": 0.0004786668869201094,
          "validation_metric": 0.00045083920122124255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3884470439988945
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0054652090548188426,
          "train_loss": 0.0038708809297531843,
          "validation_loss": 0.002997540170326829,
          "train_metric": 0.000389860913855955,
          "validation_metric": 0.00030190128018148243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.399907164999604
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.007330891356104985,
          "train_loss": 0.005862218793481588,
          "validation_loss": 0.005248975940048695,
          "train_metric": 0.0005904212011955678,
          "validation_metric": 0.0005286575760692358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4116299789984623
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.013736682129092515,
          "train_loss": 0.015439323149621487,
          "validation_loss": 0.012253369204699993,
          "train_metric": 0.0015549920499324799,
          "validation_metric": 0.0012341145193204284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4234420059983677
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.011613299779128283,
          "train_loss": 0.006985759828239679,
          "validation_loss": 0.00433517899364233,
          "train_metric": 0.0007035801536403596,
          "validation_metric": 0.00043662331881932914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.434871055998883
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.00577308090578299,
          "train_loss": 0.004430924542248249,
          "validation_loss": 0.003574095666408539,
          "train_metric": 0.0004462665237952024,
          "validation_metric": 0.00035996982478536665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4469099149991962
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004578897365718149,
          "train_loss": 0.004321723245084286,
          "validation_loss": 0.003344193333759904,
          "train_metric": 0.00043526815716177225,
          "validation_metric": 0.0003368149045854807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4584998519985675
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.00459577061701566,
          "train_loss": 0.005147032905369997,
          "validation_loss": 0.004615332931280136,
          "train_metric": 0.0005183902685530484,
          "validation_metric": 0.000464839453343302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4698495399989042
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.005683752548065968,
          "train_loss": 0.007339784875512123,
          "validation_loss": 0.00704315397888422,
          "train_metric": 0.0007392362458631396,
          "validation_metric": 0.0007093606400303543,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5710658969983342
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.005010478154872544,
          "train_loss": 0.003816662821918726,
          "validation_loss": 0.003023763420060277,
          "train_metric": 0.0003844003367703408,
          "validation_metric": 0.0003045423945877701,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.584394389999943
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003978270528023131,
          "train_loss": 0.003530483925715089,
          "validation_loss": 0.0027423917781561613,
          "train_metric": 0.0003555773582775146,
          "validation_metric": 0.000276203645626083,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5968709789995046
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.004352009404101409,
          "train_loss": 0.005612063221633434,
          "validation_loss": 0.003644774667918682,
          "train_metric": 0.0005652264226227999,
          "validation_metric": 0.00036708833067677915,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.609195634999196
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.004016492894152179,
          "train_loss": 0.0033286181278526783,
          "validation_loss": 0.0025449807289987803,
          "train_metric": 0.0003352461790200323,
          "validation_metric": 0.0002563211601227522,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6215535659994202
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0035443547094473615,
          "train_loss": 0.003258566837757826,
          "validation_loss": 0.002547127427533269,
          "train_metric": 0.00032819091575220227,
          "validation_metric": 0.0002565373433753848,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6336616199987475
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.003934620981453918,
          "train_loss": 0.0033144131302833557,
          "validation_loss": 0.002838343381881714,
          "train_metric": 0.00033381552202627063,
          "validation_metric": 0.00028586757252924144,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6458141610000894
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0034742190400720574,
          "train_loss": 0.003855963470414281,
          "validation_loss": 0.0027003863360732794,
          "train_metric": 0.0003883585159201175,
          "validation_metric": 0.0002719729964155704,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6582298719986284
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.00366973718337249,
          "train_loss": 0.003487525973469019,
          "validation_loss": 0.002650123555213213,
          "train_metric": 0.000351250811945647,
          "validation_metric": 0.00026691073435358703,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6695389569995314
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0035939534136559814,
          "train_loss": 0.005163641646504402,
          "validation_loss": 0.004899652674794197,
          "train_metric": 0.0005200630403123796,
          "validation_metric": 0.0004934750613756478,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6826999759996397
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.005035716079873964,
          "train_loss": 0.0038094106130301952,
          "validation_loss": 0.003419130574911833,
          "train_metric": 0.00038366985972970724,
          "validation_metric": 0.0003443622845225036,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.694457595998756
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0034688602609094232,
          "train_loss": 0.003159724874421954,
          "validation_loss": 0.002616121666505933,
          "train_metric": 0.0003182359505444765,
          "validation_metric": 0.000263486203039065,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7064622189991496
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.003359427602845244,
          "train_loss": 0.0032016620971262455,
          "validation_loss": 0.002495418069884181,
          "train_metric": 0.00032245967304334044,
          "validation_metric": 0.0002513293584343046,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7187273079998704
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.003339350630994886,
          "train_loss": 0.0031924089416861534,
          "validation_loss": 0.002435301663354039,
          "train_metric": 0.0003215277392882854,
          "validation_metric": 0.0002452746848575771,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7310282009984803
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.00373233002028428,
          "train_loss": 0.0035193695221096277,
          "validation_loss": 0.0025345648173242807,
          "train_metric": 0.0003544579667504877,
          "validation_metric": 0.0002552721125539392,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7431799649984896
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0033833367633633316,
          "train_loss": 0.004622151143848896,
          "validation_loss": 0.0030483321752399206,
          "train_metric": 0.00046552610001526773,
          "validation_metric": 0.0003070168604608625,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7572647069991945
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0042204486016999,
          "train_loss": 0.0038546163123100996,
          "validation_loss": 0.0027219706680625677,
          "train_metric": 0.0003882228338625282,
          "validation_metric": 0.00027414687792770565,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7715797199998633
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.003577223062166013,
          "train_loss": 0.0031604866962879896,
          "validation_loss": 0.0027624310459941626,
          "train_metric": 0.00031831266824156046,
          "validation_metric": 0.00027822190895676613,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7831818939994264
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0031502308338531293,
          "train_loss": 0.003365479176864028,
          "validation_loss": 0.002499296795576811,
          "train_metric": 0.0003389587509445846,
          "validation_metric": 0.00025172001915052533,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7946768859983422
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0037181002262514085,
          "train_loss": 0.004074526019394398,
          "validation_loss": 0.002954816212877631,
          "train_metric": 0.0004103713436052203,
          "validation_metric": 0.0002975982497446239,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8065444079984445
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0037235970521578565,
          "train_loss": 0.003560285549610853,
          "validation_loss": 0.0032159090042114258,
          "train_metric": 0.00035857895272783935,
          "validation_metric": 0.0003238945791963488,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8179955189989414
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0038016927464923356,
          "train_loss": 0.004056481644511223,
          "validation_loss": 0.00400276854634285,
          "train_metric": 0.0004085540131200105,
          "validation_metric": 0.0004031441640108824,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8295627719999175
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.003610362320614513,
          "train_loss": 0.0045798541978001595,
          "validation_loss": 0.004470268730074167,
          "train_metric": 0.00046126614324748516,
          "validation_metric": 0.00045022909762337804,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8411845819991868
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.00413737721828511,
          "train_loss": 0.00682070292532444,
          "validation_loss": 0.00643257237970829,
          "train_metric": 0.000686956278514117,
          "validation_metric": 0.0006478651193901896,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.852633866999895
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.003552504953404423,
          "train_loss": 0.0033443965949118137,
          "validation_loss": 0.002379385521635413,
          "train_metric": 0.00033683539368212223,
          "validation_metric": 0.00023964299180079252,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8648460929998691
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0035576228256104514,
          "train_loss": 0.003919692710042,
          "validation_loss": 0.002742104697972536,
          "train_metric": 0.000394777103792876,
          "validation_metric": 0.00027617477462626994,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.876567803999933
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0033106888804468326,
          "train_loss": 0.0033046985045075417,
          "validation_loss": 0.0030471195932477713,
          "train_metric": 0.0003328371385578066,
          "validation_metric": 0.0003068947698920965,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8881628609997279
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.003343187505379319,
          "train_loss": 0.003721953136846423,
          "validation_loss": 0.003453755285590887,
          "train_metric": 0.0003748614981304854,
          "validation_metric": 0.00034784956369549036,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8998416569993424
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0036308242997620255,
          "train_loss": 0.0034510265104472637,
          "validation_loss": 0.003162798937410116,
          "train_metric": 0.0003475747653283179,
          "validation_metric": 0.0003185455861967057,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9112783419986954
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.00311415885516908,
          "train_loss": 0.0031701333355158567,
          "validation_loss": 0.0028390525840222836,
          "train_metric": 0.0003192841832060367,
          "validation_metric": 0.0002859390224330127,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9231551259999833
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0047577103396179155,
          "train_loss": 0.003085935488343239,
          "validation_loss": 0.0026589143089950085,
          "train_metric": 0.0003108041128143668,
          "validation_metric": 0.00026779610197991133,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.934447106999869
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0043068755767308176,
          "train_loss": 0.0033874784130603075,
          "validation_loss": 0.002709385473281145,
          "train_metric": 0.00034117442555725574,
          "validation_metric": 0.000272879347903654,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.946188092999364
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.004246916647389298,
          "train_loss": 0.007905171252787113,
          "validation_loss": 0.005464683752506971,
          "train_metric": 0.0007961799274198711,
          "validation_metric": 0.0005503828870132565,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9578617489987664
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.004712146706879139,
          "train_loss": 0.0047450014390051365,
          "validation_loss": 0.003453448647633195,
          "train_metric": 0.00047789918608032167,
          "validation_metric": 0.00034781868453137577,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9692150119990401
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.004263356313458644,
          "train_loss": 0.004524735268205404,
          "validation_loss": 0.0030907834880053997,
          "train_metric": 0.00045571476221084595,
          "validation_metric": 0.00031129238777793944,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9819181469993055
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.004550763373117661,
          "train_loss": 0.003458936233073473,
          "validation_loss": 0.0025599903892725706,
          "train_metric": 0.00034837136627174914,
          "validation_metric": 0.0002578329003881663,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9955273830000806
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0030753744795219973,
          "train_loss": 0.002841443056240678,
          "validation_loss": 0.002359999343752861,
          "train_metric": 0.0002861797693185508,
          "validation_metric": 0.0002376904885750264,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0961392409990367
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0031994156270229723,
          "train_loss": 0.003393355291336775,
          "validation_loss": 0.0031773706432431936,
          "train_metric": 0.00034176631015725434,
          "validation_metric": 0.00032001311774365604,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1097042779983894
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.003293241730716545,
          "train_loss": 0.0033503773156553507,
          "validation_loss": 0.003103978931903839,
          "train_metric": 0.00033743769745342433,
          "validation_metric": 0.0003126213850919157,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1224899559983896
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0033750834336387925,
          "train_loss": 0.003209258895367384,
          "validation_loss": 0.0029141854029148817,
          "train_metric": 0.000323224812746048,
          "validation_metric": 0.00029350610566325486,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.134062567000001
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.003108756987785455,
          "train_loss": 0.0029217698611319065,
          "validation_loss": 0.0026341991033405066,
          "train_metric": 0.00029426993569359183,
          "validation_metric": 0.00026530688046477735,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.146685844998501
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0031775455208844505,
          "train_loss": 0.0033516401890665293,
          "validation_loss": 0.003178255632519722,
          "train_metric": 0.00033756496850401163,
          "validation_metric": 0.000320102262776345,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.159257090999745
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0030768151555093937,
          "train_loss": 0.0027629102114588022,
          "validation_loss": 0.0023047944996505976,
          "train_metric": 0.00027827019221149385,
          "validation_metric": 0.0002321304491488263,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1724017019987514
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.00280075286354986,
          "train_loss": 0.002755584893748164,
          "validation_loss": 0.002239038236439228,
          "train_metric": 0.0002775323810055852,
          "validation_metric": 0.00022550772700924426,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.184608796998873
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.002828561217029346,
          "train_loss": 0.002739500720053911,
          "validation_loss": 0.0022287429310381413,
          "train_metric": 0.00027591249090619385,
          "validation_metric": 0.00022447083028964698,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.19772734199978
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0028092607390135527,
          "train_loss": 0.002745160134509206,
          "validation_loss": 0.0023163261357694864,
          "train_metric": 0.00027648243121802807,
          "validation_metric": 0.00023329186660703272,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2105645859992364
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0029548874736065045,
          "train_loss": 0.002793640596792102,
          "validation_loss": 0.002205412834882736,
          "train_metric": 0.0002813652390614152,
          "validation_metric": 0.0002221210888819769,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2235277959989617
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.00296430380694801,
          "train_loss": 0.002823741640895605,
          "validation_loss": 0.002206485951319337,
          "train_metric": 0.0002843968977686018,
          "validation_metric": 0.00022222916595637798,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.235618529999556
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.002903971075284062,
          "train_loss": 0.0028342194855213165,
          "validation_loss": 0.0022194322664290667,
          "train_metric": 0.00028545220266096294,
          "validation_metric": 0.0002235330903204158,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.247717449999982
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.002912571952037979,
          "train_loss": 0.00294770416803658,
          "validation_loss": 0.00264930073171854,
          "train_metric": 0.0002968819753732532,
          "validation_metric": 0.0002668278757482767,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2605997059999936
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0029336094594327733,
          "train_loss": 0.0028556224424391985,
          "validation_loss": 0.0025443285703659058,
          "train_metric": 0.00028760783607140183,
          "validation_metric": 0.0002562554436735809,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.27421296600005
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.002859988009731751,
          "train_loss": 0.002771772211417556,
          "validation_loss": 0.002171830739825964,
          "train_metric": 0.0002791627775877714,
          "validation_metric": 0.00021873883088119328,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2878456899998127
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0028795815669582225,
          "train_loss": 0.00312988692894578,
          "validation_loss": 0.0028650204185396433,
          "train_metric": 0.00031523071811534464,
          "validation_metric": 0.00028855435084551573,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3001588139995874
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0027516804584593046,
          "train_loss": 0.0029557200614362955,
          "validation_loss": 0.0021830257028341293,
          "train_metric": 0.000297689315630123,
          "validation_metric": 0.00021986635692883283,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3129258009994373
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0029718453042733017,
          "train_loss": 0.002697867574170232,
          "validation_loss": 0.0021389867179095745,
          "train_metric": 0.00027171935653313994,
          "validation_metric": 0.0002154309331672266,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3259226539994415
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.002831816309480928,
          "train_loss": 0.002722811419516802,
          "validation_loss": 0.0022500487975776196,
          "train_metric": 0.00027423157007433474,
          "validation_metric": 0.00022661668481305242,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3389069349996134
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.002783941454254091,
          "train_loss": 0.0027452935464680195,
          "validation_loss": 0.0024016438983380795,
          "train_metric": 0.00027649593539536,
          "validation_metric": 0.0002418848016532138,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3507897599993157
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0029187558102421463,
          "train_loss": 0.002714521950110793,
          "validation_loss": 0.002167937345802784,
          "train_metric": 0.0002733967558015138,
          "validation_metric": 0.00021834672952536494,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.364363147999029
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0028104015091230394,
          "train_loss": 0.0030482872389256954,
          "validation_loss": 0.0028842382598668337,
          "train_metric": 0.0003070123784709722,
          "validation_metric": 0.0002904899010900408,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3777542269999685
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.002962520382425282,
          "train_loss": 0.0026691637467592955,
          "validation_loss": 0.0022671872284263372,
          "train_metric": 0.0002688283857423812,
          "validation_metric": 0.0002283428329974413,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3910675889983395
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.002868089486582903,
          "train_loss": 0.0033135004341602325,
          "validation_loss": 0.0024141406174749136,
          "train_metric": 0.0003337236412335187,
          "validation_metric": 0.00024314338224940002,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4039338679995126
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0030205203220248222,
          "train_loss": 0.002858797088265419,
          "validation_loss": 0.002517556771636009,
          "train_metric": 0.0002879275707527995,
          "validation_metric": 0.00025355909019708633,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4164004169997497
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.002969360866700299,
          "train_loss": 0.0027719377540051937,
          "validation_loss": 0.0024612308479845524,
          "train_metric": 0.0002791794540826231,
          "validation_metric": 0.0002478861133567989,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.429223468998316
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0028620106750167906,
          "train_loss": 0.002782144583761692,
          "validation_loss": 0.002109371591359377,
          "train_metric": 0.0002802073722705245,
          "validation_metric": 0.0002124481979990378,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4426384439993853
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0027860066838911735,
          "train_loss": 0.0026162557769566774,
          "validation_loss": 0.00208803778514266,
          "train_metric": 0.00026349970721639693,
          "validation_metric": 0.00021029952040407807,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.456516702999579
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0027973560863756575,
          "train_loss": 0.002608503680676222,
          "validation_loss": 0.002195283304899931,
          "train_metric": 0.0002627188805490732,
          "validation_metric": 0.00022110091231297702,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4689495819984586
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0027880889756488614,
          "train_loss": 0.002608825685456395,
          "validation_loss": 0.002134128240868449,
          "train_metric": 0.00026275136042386293,
          "validation_metric": 0.00021494159591384232,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4815671050000674
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.002794772073684726,
          "train_loss": 0.002643426414579153,
          "validation_loss": 0.002268358366563916,
          "train_metric": 0.00026623622397892177,
          "validation_metric": 0.00022846074716653675,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5107423379995453
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.002682893795281416,
          "train_loss": 0.0025970488786697388,
          "validation_loss": 0.002241747220978141,
          "train_metric": 0.00026156523381359875,
          "validation_metric": 0.0002257806045236066,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5526059589992656
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0027198361203772947,
          "train_loss": 0.0027812623884528875,
          "validation_loss": 0.0021181146148592234,
          "train_metric": 0.00028011854737997055,
          "validation_metric": 0.00021332877804525197,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5701974969997536
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0027523263943294296,
          "train_loss": 0.0026868765708059072,
          "validation_loss": 0.0020905681885778904,
          "train_metric": 0.0002706123632378876,
          "validation_metric": 0.00021055436809547246,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5872097899991786
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0029178711265558377,
          "train_loss": 0.0027324380353093147,
          "validation_loss": 0.0023420192301273346,
          "train_metric": 0.00027520113508217037,
          "validation_metric": 0.00023587961914017797,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.604488083999968
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0026653312415874097,
          "train_loss": 0.002569593722000718,
          "validation_loss": 0.0020864680409431458,
          "train_metric": 0.0002588000788819045,
          "validation_metric": 0.0002101414283970371,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.7228367019997677
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0028244993518455885,
          "train_loss": 0.003633235115557909,
          "validation_loss": 0.0034746697638183832,
          "train_metric": 0.0003659261274151504,
          "validation_metric": 0.00034995604073628783,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.740148163999038
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0034887204128608573,
          "train_loss": 0.004312189761549234,
          "validation_loss": 0.004061693791300058,
          "train_metric": 0.00043430793448351324,
          "validation_metric": 0.0004090789589099586,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.756675959999484
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.003344781682244502,
          "train_loss": 0.00274540763348341,
          "validation_loss": 0.0025173125322908163,
          "train_metric": 0.00027650740230455995,
          "validation_metric": 0.0002535344974603504,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.773303699999815
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002622687265102286,
          "train_loss": 0.0025465942453593016,
          "validation_loss": 0.002170603722333908,
          "train_metric": 0.00025648364680819213,
          "validation_metric": 0.00021861527056898922,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.790153346999432
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0026150169142056257,
          "train_loss": 0.0025433371774852276,
          "validation_loss": 0.002081310609355569,
          "train_metric": 0.0002561556175351143,
          "validation_metric": 0.00020962199778296053,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.8077719199991407
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0028154743304185104,
          "train_loss": 0.003307884559035301,
          "validation_loss": 0.0031400471925735474,
          "train_metric": 0.0003331579500809312,
          "validation_metric": 0.00031625403789803386,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.824502653998934
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0028242958505870774,
          "train_loss": 0.0027884419541805983,
          "validation_loss": 0.0020695249550044537,
          "train_metric": 0.00028084166115149856,
          "validation_metric": 0.00020843495440203696,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.842041889000029
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.002919525228207931,
          "train_loss": 0.00462714396417141,
          "validation_loss": 0.004522036295384169,
          "train_metric": 0.00046602898510172963,
          "validation_metric": 0.00045544293243438005,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.85923842199918
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0033547593047842383,
          "train_loss": 0.0034673218615353107,
          "validation_loss": 0.0032330129761248827,
          "train_metric": 0.000349215988535434,
          "validation_metric": 0.00032561723492108285,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.876437081999029
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.003225386160920607,
          "train_loss": 0.0026206497568637133,
          "validation_loss": 0.0020513536874204874,
          "train_metric": 0.000263942260062322,
          "validation_metric": 0.00020660481823142618,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.893942500999401
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.002671302354428917,
          "train_loss": 0.0025585880503058434,
          "validation_loss": 0.0020574182271957397,
          "train_metric": 0.00025769160129129887,
          "validation_metric": 0.00020721562032122165,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.910994520998429
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0026472109129827004,
          "train_loss": 0.0026932971086353064,
          "validation_loss": 0.002477973699569702,
          "train_metric": 0.0002712590212468058,
          "validation_metric": 0.0002495724183972925,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.927504910998323
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0027181112200196367,
          "train_loss": 0.003329554805532098,
          "validation_loss": 0.0032571847550570965,
          "train_metric": 0.00033534059184603393,
          "validation_metric": 0.0003280517121311277,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.9445682049990864
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0026472714998817537,
          "train_loss": 0.003315748181194067,
          "validation_loss": 0.0023980820551514626,
          "train_metric": 0.0003339500108268112,
          "validation_metric": 0.00024152605328708887,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.961251471999276
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0030090924046817236,
          "train_loss": 0.002812316408380866,
          "validation_loss": 0.0021719264332205057,
          "train_metric": 0.000283246161416173,
          "validation_metric": 0.0002187484788009897,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.978632876998745
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.003010188440384809,
          "train_loss": 0.0025419481098651886,
          "validation_loss": 0.0021190596744418144,
          "train_metric": 0.00025601571542210877,
          "validation_metric": 0.00021342393301893026,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.9953438969987474
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0026300889912818093,
          "train_loss": 0.0025493092834949493,
          "validation_loss": 0.0021874948870390654,
          "train_metric": 0.0002567571063991636,
          "validation_metric": 0.00022031647677067667,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.012188064998554
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0026160589877690654,
          "train_loss": 0.002600632607936859,
          "validation_loss": 0.002376858377829194,
          "train_metric": 0.00026192617951892316,
          "validation_metric": 0.00023938846425153315,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.0288505149983393
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0025893636520777363,
          "train_loss": 0.0024161692708730698,
          "validation_loss": 0.002008316805586219,
          "train_metric": 0.00024334770569112152,
          "validation_metric": 0.00020227029745001346,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.0465010049992998
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.002507945326215122,
          "train_loss": 0.002466978505253792,
          "validation_loss": 0.00207957299426198,
          "train_metric": 0.00024846504675224423,
          "validation_metric": 0.0002094469964504242,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.063244754999687
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.002902829430240672,
          "train_loss": 0.0034691374748945236,
          "validation_loss": 0.0025082582142204046,
          "train_metric": 0.0003493988770060241,
          "validation_metric": 0.00025262258714064956,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.079999177998616
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.002779967471724376,
          "train_loss": 0.0024827320594340563,
          "validation_loss": 0.0020955556537956,
          "train_metric": 0.00025005167117342353,
          "validation_metric": 0.00021105668565724045,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.0968926229998033
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0024997494292620104,
          "train_loss": 0.002536236075684428,
          "validation_loss": 0.001978684216737747,
          "train_metric": 0.0002554404200054705,
          "validation_metric": 0.00019928583060391247,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.1150161669993395
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0025524643970129546,
          "train_loss": 0.0024134761188179255,
          "validation_loss": 0.0020584161393344402,
          "train_metric": 0.00024307650164701045,
          "validation_metric": 0.00020731613039970398,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.1318904869986
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0025758287229109555,
          "train_loss": 0.002411142224445939,
          "validation_loss": 0.0019517159089446068,
          "train_metric": 0.00024284141545649618,
          "validation_metric": 0.00019656965741887689,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.149590208999143
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.002740628577157622,
          "train_loss": 0.002584166591987014,
          "validation_loss": 0.0022684617433696985,
          "train_metric": 0.000260267814155668,
          "validation_metric": 0.0002284711808897555,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.180459734998294
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.002658192770468304,
          "train_loss": 0.003012641565874219,
          "validation_loss": 0.0022179519291967154,
          "train_metric": 0.00030342224636115134,
          "validation_metric": 0.00022338404960464686,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.2228416509988165
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0025649940653238446,
          "train_loss": 0.0024090022780001163,
          "validation_loss": 0.002076969016343355,
          "train_metric": 0.00024262590159196407,
          "validation_metric": 0.00020918472728226334,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.266313526999511
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.002531085308874026,
          "train_loss": 0.0028814871329814196,
          "validation_loss": 0.0022042100317776203,
          "train_metric": 0.0002902128326240927,
          "validation_metric": 0.00022199998784344643,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.312567606999437
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0028287192981224507,
          "train_loss": 0.0023729903623461723,
          "validation_loss": 0.002014672849327326,
          "train_metric": 0.00023899890948086977,
          "validation_metric": 0.00020291046530473977,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.3546900329984055
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0026753967540571466,
          "train_loss": 0.0023880673106759787,
          "validation_loss": 0.0020208752248436213,
          "train_metric": 0.00024051740183494985,
          "validation_metric": 0.0002035351499216631,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.4003987299984146
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0027670009985740762,
          "train_loss": 0.003152624936774373,
          "validation_loss": 0.0023784253280609846,
          "train_metric": 0.00031752081122249365,
          "validation_metric": 0.00023954630887601525,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.4444090409997443
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0028100516865379177,
          "train_loss": 0.002753671258687973,
          "validation_loss": 0.0021072484087198973,
          "train_metric": 0.00027733968454413116,
          "validation_metric": 0.00021223434305284172,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.4875255609986198
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0031335284875240177,
          "train_loss": 0.0037623299285769463,
          "validation_loss": 0.002569067757576704,
          "train_metric": 0.00037892808904871345,
          "validation_metric": 0.00025874710991047323,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.5337935909992666
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.0030469831253867596,
          "train_loss": 0.0031403088942170143,
          "validation_loss": 0.002332167001441121,
          "train_metric": 0.00031628040596842766,
          "validation_metric": 0.00023488732404075563,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.5758759129985265
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0028556293254951015,
          "train_loss": 0.0025542767252773046,
          "validation_loss": 0.0020347232930362225,
          "train_metric": 0.00025725740124471486,
          "validation_metric": 0.0002049298636848107,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.780260441999417
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.002535286301281303,
          "train_loss": 0.002320523839443922,
          "validation_loss": 0.0019217767985537648,
          "train_metric": 0.00023371467250399292,
          "validation_metric": 0.00019355432596057653,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.808389242998601
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0025023443668032996,
          "train_loss": 0.002337847836315632,
          "validation_loss": 0.001955759944394231,
          "train_metric": 0.00023545947624370456,
          "validation_metric": 0.00019697699462994933,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.8207976519988733
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0024340683012269437,
          "train_loss": 0.0023722711484879255,
          "validation_loss": 0.0019217970548197627,
          "train_metric": 0.00023892645549494773,
          "validation_metric": 0.0001935563632287085,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.8328948210000817
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0025979642850870732,
          "train_loss": 0.0024957433342933655,
          "validation_loss": 0.0022487700916826725,
          "train_metric": 0.00025136215845122933,
          "validation_metric": 0.0002264879149151966,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.845154176999131
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.0024280159777845256,
          "train_loss": 0.002397232921794057,
          "validation_loss": 0.0018944763578474522,
          "train_metric": 0.00024144053168129176,
          "validation_metric": 0.00019080471247434616,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.858017935999669
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.002416699153400259,
          "train_loss": 0.0023478998336941004,
          "validation_loss": 0.001915342640131712,
          "train_metric": 0.00023647189664188772,
          "validation_metric": 0.0001929063000716269,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.8698798669993266
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.0025372985619469546,
          "train_loss": 0.0023780460469424725,
          "validation_loss": 0.0021064893808215857,
          "train_metric": 0.00023950809554662555,
          "validation_metric": 0.00021215793094597757,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.882197796998298
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0025351029908051714,
          "train_loss": 0.002408508211374283,
          "validation_loss": 0.002168921520933509,
          "train_metric": 0.00024257614859379828,
          "validation_metric": 0.00021844582806807011,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.8942244319987367
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.0025234560962417163,
          "train_loss": 0.002323683351278305,
          "validation_loss": 0.00203173840418458,
          "train_metric": 0.00023403290833812207,
          "validation_metric": 0.00020462922111619264,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.906152961999396
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.002424548307317309,
          "train_loss": 0.00232362630777061,
          "validation_loss": 0.0019006406655535102,
          "train_metric": 0.00023402711667586118,
          "validation_metric": 0.0001914255553856492,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.918478479999976
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.002323940945643699,
          "train_loss": 0.0023026326671242714,
          "validation_loss": 0.0019882963970303535,
          "train_metric": 0.00023191270884126425,
          "validation_metric": 0.00020025394042022526,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.930672776999927
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.0023763628378219437,
          "train_loss": 0.002310818061232567,
          "validation_loss": 0.0018582214834168553,
          "train_metric": 0.0002327371039427817,
          "validation_metric": 0.00018715325859375298,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.943537325998477
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.0023440508921339642,
          "train_loss": 0.002299927407875657,
          "validation_loss": 0.0019388798391446471,
          "train_metric": 0.00023164025333244354,
          "validation_metric": 0.000195276879821904,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.9559082619998662
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.002312598364369478,
          "train_loss": 0.00230198516510427,
          "validation_loss": 0.0018640891648828983,
          "train_metric": 0.00023184750170912594,
          "validation_metric": 0.00018774424097500741,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.9674168709989317
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.002379172743530944,
          "train_loss": 0.0023172704968601465,
          "validation_loss": 0.002036160323768854,
          "train_metric": 0.00023338699247688055,
          "validation_metric": 0.00020507461158558726,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.979804302998673
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.0024643399883643724,
          "train_loss": 0.0023316910956054926,
          "validation_loss": 0.0021063925232738256,
          "train_metric": 0.00023483937548007816,
          "validation_metric": 0.0002121481520589441,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.9917212609998387
        }
      ],
      "best": {
        "epoch": 236,
        "updates": 3776
      },
      "best_validation_metric": 0.00018715325859375298,
      "config": {
        "label": "Validation plateau",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.013680396872669776
    },
    "1298f0867b9f_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7251859698444605,
          "train_loss": 0.2715291380882263,
          "validation_loss": 0.2244172841310501,
          "train_metric": 0.027347417548298836,
          "validation_metric": 0.022602485492825508,
          "lr_first": 0.003,
          "lr_last": 0.002999887429249521,
          "seconds": 0.030910441999367322
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.1925148726440966,
          "train_loss": 0.13974425196647644,
          "validation_loss": 0.11790812760591507,
          "train_metric": 0.014074530452489853,
          "validation_metric": 0.0118752745911479,
          "lr_first": 0.0029998719197231404,
          "lr_last": 0.0029995192175544397,
          "seconds": 0.06823344000076759
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12273702304810286,
          "train_loss": 0.1003313660621643,
          "validation_loss": 0.08399085700511932,
          "train_metric": 0.010105007328093052,
          "validation_metric": 0.008459251374006271,
          "lr_first": 0.0029994877008384588,
          "lr_last": 0.002998894927685237,
          "seconds": 0.08762678099992627
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09153524460271001,
          "train_loss": 0.08049804717302322,
          "validation_loss": 0.06612914055585861,
          "train_metric": 0.008107468485832214,
          "validation_metric": 0.00666028494015336,
          "lr_first": 0.002998847409179882,
          "lr_last": 0.002998014666610766,
          "seconds": 0.1003486179997708
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07135118520818651,
          "train_loss": 0.06517963111400604,
          "validation_loss": 0.052674103528261185,
          "train_metric": 0.006564653478562832,
          "validation_metric": 0.005305142141878605,
          "lr_first": 0.0029979511544580887,
          "lr_last": 0.0029968785851592315,
          "seconds": 0.11308139799984929
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.060715499450452626,
          "train_loss": 0.057129450142383575,
          "validation_loss": 0.04709867760539055,
          "train_metric": 0.005753868725150824,
          "validation_metric": 0.0047436063177883625,
          "lr_first": 0.002996799090241712,
          "lr_last": 0.002995486877992343,
          "seconds": 0.12526557499950286
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05504207592457533,
          "train_loss": 0.05026744678616524,
          "validation_loss": 0.04124567285180092,
          "train_metric": 0.005062754265964031,
          "validation_metric": 0.004154113586992025,
          "lr_first": 0.0029953914139310266,
          "lr_last": 0.0029938397835719568,
          "seconds": 0.1377815260002535
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04906744183972478,
          "train_loss": 0.04576471447944641,
          "validation_loss": 0.03794929385185242,
          "train_metric": 0.004609254188835621,
          "validation_metric": 0.0038221138529479504,
          "lr_first": 0.002993728366724119,
          "lr_last": 0.002991937584119225,
          "seconds": 0.15050348300064798
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0452122229617089,
          "train_loss": 0.05027538910508156,
          "validation_loss": 0.041954390704631805,
          "train_metric": 0.005063552875071764,
          "validation_metric": 0.004225492477416992,
          "lr_first": 0.0029918102335755687,
          "lr_last": 0.0029897806055662304,
          "seconds": 0.16310171199984325
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.04940412542782724,
          "train_loss": 0.04624531790614128,
          "validation_loss": 0.03932812064886093,
          "train_metric": 0.004657658748328686,
          "validation_metric": 0.003960984293371439,
          "lr_first": 0.0029896373431476165,
          "lr_last": 0.0029873692175001417,
          "seconds": 0.17572889700022643
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.040977864409796894,
          "train_loss": 0.03849419578909874,
          "validation_loss": 0.03277565911412239,
          "train_metric": 0.0038769945967942476,
          "validation_metric": 0.003301044227555394,
          "lr_first": 0.002987210067753847,
          "lr_last": 0.0029847038330998916,
          "seconds": 0.18790192199958256
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.036936569260433316,
          "train_loss": 0.03291647136211395,
          "validation_loss": 0.028423961251974106,
          "train_metric": 0.003315226873382926,
          "validation_metric": 0.002862757770344615,
          "lr_first": 0.0029845288232954064,
          "lr_last": 0.0029817849090653776,
          "seconds": 0.2025914459991327
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03453731327317655,
          "train_loss": 0.03703802078962326,
          "validation_loss": 0.029439831152558327,
          "train_metric": 0.0037303336430341005,
          "validation_metric": 0.002965072402730584,
          "lr_first": 0.0029815940691897305,
          "lr_last": 0.002978612945539204,
          "seconds": 0.22770252699956472
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03215481748338789,
          "train_loss": 0.029102806001901627,
          "validation_loss": 0.02457115426659584,
          "train_metric": 0.00293112825602293,
          "validation_metric": 0.002474716631695628,
          "lr_first": 0.002978406308291826,
          "lr_last": 0.0029751884860209923,
          "seconds": 0.24332758499986085
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.030516590108163655,
          "train_loss": 0.027691422030329704,
          "validation_loss": 0.023406490683555603,
          "train_metric": 0.0027889786288142204,
          "validation_metric": 0.0023574160877615213,
          "lr_first": 0.0029749660868081114,
          "lr_last": 0.0029715121172742523,
          "seconds": 0.25853120399915497
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.026533346681389958,
          "train_loss": 0.027976153418421745,
          "validation_loss": 0.023318512365221977,
          "train_metric": 0.0028176556807011366,
          "validation_metric": 0.0023485547862946987,
          "lr_first": 0.0029712739942028285,
          "lr_last": 0.0029675844692258427,
          "seconds": 0.28897187300026417
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.024512063362635672,
          "train_loss": 0.024151409044861794,
          "validation_loss": 0.02060062065720558,
          "train_metric": 0.002432441571727395,
          "validation_metric": 0.0020748190581798553,
          "lr_first": 0.002967330663097039,
          "lr_last": 0.002963406214858036,
          "seconds": 0.3089508390003175
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.023496865818742663,
          "train_loss": 0.020987579599022865,
          "validation_loss": 0.01757306233048439,
          "train_metric": 0.002113792346790433,
          "validation_metric": 0.001769894384779036,
          "lr_first": 0.0029631367691602254,
          "lr_last": 0.00295897807009321,
          "seconds": 0.32126652099941566
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.021496260887943208,
          "train_loss": 0.022907618433237076,
          "validation_loss": 0.01876150071620941,
          "train_metric": 0.0023071717005223036,
          "validation_metric": 0.001889589475467801,
          "lr_first": 0.002958693030994526,
          "lr_last": 0.002954300793671172,
          "seconds": 0.3331662300006428
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02543235692428425,
          "train_loss": 0.022229831665754318,
          "validation_loss": 0.018077053129673004,
          "train_metric": 0.0022389069199562073,
          "validation_metric": 0.0018206544918939471,
          "lr_first": 0.002954000210011597,
          "lr_last": 0.00294937518701916,
          "seconds": 0.34512510700005805
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023038860119413584,
          "train_loss": 0.02083730697631836,
          "validation_loss": 0.017441947013139725,
          "train_metric": 0.002098657190799713,
          "validation_metric": 0.0017566886963322759,
          "lr_first": 0.002949059110302156,
          "lr_last": 0.0029442020941145156,
          "seconds": 0.35757910199936305
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.026678126130718738,
          "train_loss": 0.025941405445337296,
          "validation_loss": 0.02111535333096981,
          "train_metric": 0.0026127230376005173,
          "validation_metric": 0.0021266613621264696,
          "lr_first": 0.0029438705784982015,
          "lr_last": 0.002938782401340078,
          "seconds": 0.3692786609990435
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01959718661964871,
          "train_loss": 0.016222894191741943,
          "validation_loss": 0.014173045754432678,
          "train_metric": 0.001633910578675568,
          "validation_metric": 0.0014274573186412454,
          "lr_first": 0.0029384355036279485,
          "lr_last": 0.0029331170373323025,
          "seconds": 0.38161071399918
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.017307784437434748,
          "train_loss": 0.014933668076992035,
          "validation_loss": 0.012548163533210754,
          "train_metric": 0.0015040644211694598,
          "validation_metric": 0.0012638050829991698,
          "lr_first": 0.002932754816963494,
          "lr_last": 0.002927206972822146,
          "seconds": 0.39399905299978855
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019855680293403566,
          "train_loss": 0.014427846297621727,
          "validation_loss": 0.012330617755651474,
          "train_metric": 0.0014531196793541312,
          "validation_metric": 0.0012418944388628006,
          "lr_first": 0.002926829491861254,
          "lr_last": 0.002921053220468738,
          "seconds": 0.4059767210001155
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.017444699275074527,
          "train_loss": 0.014268944039940834,
          "validation_loss": 0.012481297366321087,
          "train_metric": 0.0014371158322319388,
          "validation_metric": 0.0012570704566314816,
          "lr_first": 0.0029206605435951826,
          "lr_last": 0.0029146568346858654,
          "seconds": 0.41822825700000976
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015758515481138602,
          "train_loss": 0.013558336533606052,
          "validation_loss": 0.01140686497092247,
          "train_metric": 0.0013655460206791759,
          "validation_metric": 0.0011488576419651508,
          "lr_first": 0.002914249029182806,
          "lr_last": 0.002908018911461302,
          "seconds": 0.430695115999697
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.013744735100772232,
          "train_loss": 0.012914834544062614,
          "validation_loss": 0.011324552819132805,
          "train_metric": 0.0013007349334657192,
          "validation_metric": 0.001140567590482533,
          "lr_first": 0.002907596047204113,
          "lr_last": 0.0029011405881690204,
          "seconds": 0.4434295970004314
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016192262788536027,
          "train_loss": 0.025270475074648857,
          "validation_loss": 0.020363643765449524,
          "train_metric": 0.0025451499968767166,
          "validation_metric": 0.002050951588898897,
          "lr_first": 0.002900702737613315,
          "lr_last": 0.0028940230433743067,
          "seconds": 0.4552027200006705
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.015590675990097225,
          "train_loss": 0.014106245711445808,
          "validation_loss": 0.011063610203564167,
          "train_metric": 0.0014207292115315795,
          "validation_metric": 0.0011142862495034933,
          "lr_first": 0.002893570281543524,
          "lr_last": 0.002886667496631822,
          "seconds": 0.46766919900073844
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.015503121627261862,
          "train_loss": 0.019660038873553276,
          "validation_loss": 0.016860032454133034,
          "train_metric": 0.0019800872541964054,
          "validation_metric": 0.0016980806831270456,
          "lr_first": 0.0028861999011043707,
          "lr_last": 0.002879075208276634,
          "seconds": 0.4806006399994658
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.019719137257197872,
          "train_loss": 0.01544372271746397,
          "validation_loss": 0.011891700327396393,
          "train_metric": 0.0015554351266473532,
          "validation_metric": 0.001197688514366746,
          "lr_first": 0.0028785928591726004,
          "lr_last": 0.0028712474792082655,
          "seconds": 0.4939993699990737
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01538277353392914,
          "train_loss": 0.011829396709799767,
          "validation_loss": 0.00971394032239914,
          "train_metric": 0.0011914132628589869,
          "validation_metric": 0.000978352501988411,
          "lr_first": 0.0028707504591756854,
          "lr_last": 0.0028631856506677973,
          "seconds": 0.5078051609998511
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013724651289521717,
          "train_loss": 0.02285131812095642,
          "validation_loss": 0.01923787221312523,
          "train_metric": 0.00230150087736547,
          "validation_metric": 0.001937568187713623,
          "lr_first": 0.0028626740448684956,
          "lr_last": 0.0028548911040080457,
          "seconds": 0.5207030660003511
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.015488424556679092,
          "train_loss": 0.0113980732858181,
          "validation_loss": 0.010152469389140606,
          "train_metric": 0.0011479721870273352,
          "validation_metric": 0.0010225195437669754,
          "lr_first": 0.002854365000103041,
          "lr_last": 0.002846365260456879,
          "seconds": 0.5344158659991081
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.012582450493937358,
          "train_loss": 0.011517004109919071,
          "validation_loss": 0.009377049282193184,
          "train_metric": 0.001159950508736074,
          "validation_metric": 0.0009444220922887325,
          "lr_first": 0.0028458247485913675,
          "lr_last": 0.002837609580873701,
          "seconds": 0.549150130000271
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013435143016977236,
          "train_loss": 0.013571500778198242,
          "validation_loss": 0.010471123270690441,
          "train_metric": 0.0013668719911947846,
          "validation_metric": 0.0010546131525188684,
          "lr_first": 0.0028370547536616087,
          "lr_last": 0.002828625565499134,
          "seconds": 0.5618628470001568
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.01124393637292087,
          "train_loss": 0.01161445677280426,
          "validation_loss": 0.010729078203439713,
          "train_metric": 0.0011697656009346247,
          "validation_metric": 0.0010805934434756637,
          "lr_first": 0.002828056518007247,
          "lr_last": 0.00281941475369796,
          "seconds": 0.5767070729998522
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01167955485288985,
          "train_loss": 0.013665623962879181,
          "validation_loss": 0.00972699373960495,
          "train_metric": 0.0013763514580205083,
          "validation_metric": 0.0009796672966331244,
          "lr_first": 0.0028188315834296366,
          "lr_last": 0.0028099787236953665,
          "seconds": 0.5915235060001578
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.011979028291534632,
          "train_loss": 0.011836156249046326,
          "validation_loss": 0.010595330968499184,
          "train_metric": 0.0011920944089069963,
          "validation_metric": 0.0010671227937564254,
          "lr_first": 0.00280938153057383,
          "lr_last": 0.0028003190923065156,
          "seconds": 0.6088978930001758
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011515582300489768,
          "train_loss": 0.009551207534968853,
          "validation_loss": 0.007671106141060591,
          "train_metric": 0.0009619627380743623,
          "validation_metric": 0.0007726057665422559,
          "lr_first": 0.0027997079786577336,
          "lr_last": 0.00279043751465952,
          "seconds": 0.62734368800011
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.009912974477629177,
          "train_loss": 0.008801949210464954,
          "validation_loss": 0.007740339729934931,
          "train_metric": 0.0008865000563673675,
          "validation_metric": 0.0007795786950737238,
          "lr_first": 0.00278981258519467,
          "lr_last": 0.0027803356839118383,
          "seconds": 0.6441876560002129
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009496939048403874,
          "train_loss": 0.012293331325054169,
          "validation_loss": 0.009023615159094334,
          "train_metric": 0.001238139346241951,
          "validation_metric": 0.0009088256629183888,
          "lr_first": 0.0027796970457093662,
          "lr_last": 0.002770015330960163,
          "seconds": 0.6612565439991158
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.012326479278272018,
          "train_loss": 0.010183742269873619,
          "validation_loss": 0.007562595419585705,
          "train_metric": 0.0010256691602990031,
          "validation_metric": 0.0007616769871674478,
          "lr_first": 0.0027693630934474377,
          "lr_last": 0.0027594782241438446,
          "seconds": 0.6790690329999052
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.01129301248874981,
          "train_loss": 0.010686910711228848,
          "validation_loss": 0.009900334291160107,
          "train_metric": 0.0010763463797047734,
          "validation_metric": 0.0009971254039555788,
          "lr_first": 0.0027588124990784056,
          "lr_last": 0.0027487261689418912,
          "seconds": 0.6962307180001517
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.01134641989483498,
          "train_loss": 0.010665815323591232,
          "validation_loss": 0.009331231005489826,
          "train_metric": 0.0010742219164967537,
          "validation_metric": 0.0009398075053468347,
          "lr_first": 0.0027480470703923022,
          "lr_last": 0.002737761007663611,
          "seconds": 0.7127714369999012
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011402511765481904,
          "train_loss": 0.013552173972129822,
          "validation_loss": 0.010280298069119453,
          "train_metric": 0.0013649254105985165,
          "validation_metric": 0.001035394030623138,
          "lr_first": 0.0027370686519899118,
          "lr_last": 0.0027265846191329374,
          "seconds": 0.7296041499994317
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010607103555230424,
          "train_loss": 0.009023490361869335,
          "validation_loss": 0.0073521020822227,
          "train_metric": 0.0009088129154406488,
          "validation_metric": 0.0007404768257401884,
          "lr_first": 0.0027258791249667085,
          "lr_last": 0.0027151989183665085,
          "seconds": 0.7469830640002328
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.008982313505839556,
          "train_loss": 0.008136033080518246,
          "validation_loss": 0.00740336999297142,
          "train_metric": 0.000819431443233043,
          "validation_metric": 0.0007456402527168393,
          "lr_first": 0.002714480406590544,
          "lr_last": 0.0027036058562455397,
          "seconds": 0.7643442579992552
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.008628644354757853,
          "train_loss": 0.008209314197301865,
          "validation_loss": 0.006842899136245251,
          "train_metric": 0.0008268120582215488,
          "validation_metric": 0.0006891916855238378,
          "lr_first": 0.0027028744499731326,
          "lr_last": 0.002691807419181544,
          "seconds": 0.7819830270000239
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009830878319917247,
          "train_loss": 0.009498635306954384,
          "validation_loss": 0.007439521607011557,
          "train_metric": 0.000956667703576386,
          "validation_metric": 0.0007492813747376204,
          "lr_first": 0.002691063243735394,
          "lr_last": 0.0026798056287759726,
          "seconds": 0.8182597470004112
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.01223428595403675,
          "train_loss": 0.009745948016643524,
          "validation_loss": 0.008411222137510777,
          "train_metric": 0.0009815761586651206,
          "validation_metric": 0.0008471474866382778,
          "lr_first": 0.0026790488116667116,
          "lr_last": 0.002667602541473826,
          "seconds": 0.8360350399998424
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.009216838210704736,
          "train_loss": 0.008078389801084995,
          "validation_loss": 0.0070915562100708485,
          "train_metric": 0.0008136258111335337,
          "validation_metric": 0.0007142355316318572,
          "lr_first": 0.0026668332123781697,
          "lr_last": 0.0026552002482112917,
          "seconds": 0.853043021999838
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.008003936047316529,
          "train_loss": 0.007596672512590885,
          "validation_loss": 0.006668164860457182,
          "train_metric": 0.0007651090854778886,
          "validation_metric": 0.0006715930649079382,
          "lr_first": 0.00265441853894982,
          "lr_last": 0.0026426008740574737,
          "seconds": 0.870706572999552
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008144992127199657,
          "train_loss": 0.006774345878511667,
          "validation_loss": 0.0058698528446257114,
          "train_metric": 0.000682287325616926,
          "validation_metric": 0.0005911901826038957,
          "lr_first": 0.0026418069185720444,
          "lr_last": 0.0026298065778502713,
          "seconds": 0.8883609250005975
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.01094335324887652,
          "train_loss": 0.006918420549482107,
          "validation_loss": 0.0055067772045731544,
          "train_metric": 0.000696798088029027,
          "validation_metric": 0.000554622441995889,
          "lr_first": 0.002629000512181069,
          "lr_last": 0.0026168195518264732,
          "seconds": 0.9070284139997966
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008089434588328004,
          "train_loss": 0.006745518185198307,
          "validation_loss": 0.0056754969991743565,
          "train_metric": 0.0006793838110752404,
          "validation_metric": 0.0005716152954846621,
          "lr_first": 0.002616001514088701,
          "lr_last": 0.0026036420212461322,
          "seconds": 0.9243845349992625
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.00804146326117916,
          "train_loss": 0.006869228556752205,
          "validation_loss": 0.006068085320293903,
          "train_metric": 0.0006918435683473945,
          "validation_metric": 0.0006111552356742322,
          "lr_first": 0.0026028121516063455,
          "lr_last": 0.0025902762440112768,
          "seconds": 0.9410279919993627
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007033574540400878,
          "train_loss": 0.006862299982458353,
          "validation_loss": 0.00559230474755168,
          "train_metric": 0.000691145658493042,
          "validation_metric": 0.0005632364773191512,
          "lr_first": 0.0025894346846633653,
          "lr_last": 0.002576724510279026,
          "seconds": 0.9583307550001336
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006901695473061409,
          "train_loss": 0.006613331381231546,
          "validation_loss": 0.00608318205922842,
          "train_metric": 0.0006660704966634512,
          "validation_metric": 0.0006126757944002748,
          "lr_first": 0.0025758714054198526,
          "lr_last": 0.0025629891420691935,
          "seconds": 0.9778754559993104
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.006993660790612921,
          "train_loss": 0.006119656376540661,
          "validation_loss": 0.005049709230661392,
          "train_metric": 0.0006163494545035064,
          "validation_metric": 0.0005085881566628814,
          "lr_first": 0.0025621246378738837,
          "lr_last": 0.0025490724928664103,
          "seconds": 1.0234423319998314
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0067405021109152585,
          "train_loss": 0.005868026055395603,
          "validation_loss": 0.004781602881848812,
          "train_metric": 0.0005910061299800873,
          "validation_metric": 0.000481585506349802,
          "lr_first": 0.0025481967374633083,
          "lr_last": 0.002534976947216873,
          "seconds": 1.0428026970002975
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.00838101701810956,
          "train_loss": 0.007894291542470455,
          "validation_loss": 0.005632361397147179,
          "train_metric": 0.0007950841682031751,
          "validation_metric": 0.0005672708502970636,
          "lr_first": 0.002534090090662157,
          "lr_last": 0.002520704920319766,
          "seconds": 1.0601234890000342
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.00808023194258567,
          "train_loss": 0.0058925277553498745,
          "validation_loss": 0.00501546123996377,
          "train_metric": 0.0005934738437645137,
          "validation_metric": 0.000505138945300132,
          "lr_first": 0.0025198071145717425,
          "lr_last": 0.0025062588576134272,
          "seconds": 1.0766663169997628
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0065352300080121495,
          "train_loss": 0.007373854052275419,
          "validation_loss": 0.0064874496310949326,
          "train_metric": 0.0007426675874739885,
          "validation_metric": 0.0006533921114169061,
          "lr_first": 0.0025053502565064887,
          "lr_last": 0.00249164123435633,
          "seconds": 1.0936139679997723
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0066015841730404645,
          "train_loss": 0.00590624799951911,
          "validation_loss": 0.004865074530243874,
          "train_metric": 0.0005948556354269385,
          "validation_metric": 0.000489992497023195,
          "lr_first": 0.0024907219935745985,
          "lr_last": 0.0024768545552029684,
          "seconds": 1.110784041999068
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007689833288168302,
          "train_loss": 0.006481770891696215,
          "validation_loss": 0.006060460582375526,
          "train_metric": 0.0006528202793560922,
          "validation_metric": 0.0006103874184191227,
          "lr_first": 0.0024759248322536198,
          "lr_last": 0.002461901353774695,
          "seconds": 1.1282458469995618
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0067074968974338844,
          "train_loss": 0.006767417769879103,
          "validation_loss": 0.005279066041111946,
          "train_metric": 0.0006815894739702344,
          "validation_metric": 0.0005316881579346955,
          "lr_first": 0.0024609613079609708,
          "lr_last": 0.0024467841922255927,
          "seconds": 1.1461724159998994
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.007924699675641023,
          "train_loss": 0.005496966186910868,
          "validation_loss": 0.00455567566677928,
          "train_metric": 0.0005536343669518828,
          "validation_metric": 0.00045883096754550934,
          "lr_first": 0.0024458339846195024,
          "lr_last": 0.0024315056608034706,
          "seconds": 1.1650973289997637
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.005936291163379792,
          "train_loss": 0.00560044264420867,
          "validation_loss": 0.005091919098049402,
          "train_metric": 0.0005640560411848128,
          "validation_metric": 0.0005128394695930183,
          "lr_first": 0.002430545454218197,
          "lr_last": 0.0024160683774060343,
          "seconds": 1.186424576000718
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.005315494316164404,
          "train_loss": 0.005288402084261179,
          "validation_loss": 0.004098151810467243,
          "train_metric": 0.0005326284444890916,
          "validation_metric": 0.0004127509309910238,
          "lr_first": 0.0024150983363680325,
          "lr_last": 0.0024004749871323217,
          "seconds": 1.3095338490002177
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00542874480743194,
          "train_loss": 0.005552208516746759,
          "validation_loss": 0.004337090067565441,
          "train_metric": 0.0005591981462202966,
          "validation_metric": 0.00043681581155397,
          "lr_first": 0.0023994952778531296,
          "lr_last": 0.0023847281618294826,
          "seconds": 1.3278813590004575
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.005502161511685699,
          "train_loss": 0.006161343306303024,
          "validation_loss": 0.005645361728966236,
          "train_metric": 0.0006205479730851948,
          "validation_metric": 0.0005685801734216511,
          "lr_first": 0.002383738952177242,
          "lr_last": 0.002368830599634967,
          "seconds": 1.3448301419994095
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.006020491040544584,
          "train_loss": 0.005249392706900835,
          "validation_loss": 0.004110285546630621,
          "train_metric": 0.0005286996602080762,
          "validation_metric": 0.00041397291352041066,
          "lr_first": 0.002367832059105659,
          "lr_last": 0.002352785024514216,
          "seconds": 1.3625338389992976
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006541691240272485,
          "train_loss": 0.0046327379532158375,
          "validation_loss": 0.0040024686604738235,
          "train_metric": 0.00046659246436320245,
          "validation_metric": 0.00040311398333869874,
          "lr_first": 0.00235177732420262,
          "lr_last": 0.0023365941857939245,
          "seconds": 1.3804494910000358
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0059012248857470695,
          "train_loss": 0.005000920034945011,
          "validation_loss": 0.003981641959398985,
          "train_metric": 0.0005036743823438883,
          "validation_metric": 0.0004010164411738515,
          "lr_first": 0.002335577498364301,
          "lr_last": 0.0023202608576909545,
          "seconds": 1.3986716060007893
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.006988127453951165,
          "train_loss": 0.004718727432191372,
          "validation_loss": 0.004291585646569729,
          "train_metric": 0.00047525297850370407,
          "validation_metric": 0.0004322327731642872,
          "lr_first": 0.002319235357347461,
          "lr_last": 0.002303787838836996,
          "seconds": 1.4160229059998528
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.0059924984379904345,
          "train_loss": 0.005437495186924934,
          "validation_loss": 0.004692526999861002,
          "train_metric": 0.0005476445658132434,
          "validation_metric": 0.0004726140759885311,
          "lr_first": 0.002302753701293836,
          "lr_last": 0.002287177951799022,
          "seconds": 1.434008834999986
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006752290195436217,
          "train_loss": 0.0048941876739263535,
          "validation_loss": 0.00366609962657094,
          "train_metric": 0.0004929245915263891,
          "validation_metric": 0.00036923616426065564,
          "lr_first": 0.002286135354250341,
          "lr_last": 0.002270434042595671,
          "seconds": 1.452611335000256
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004887861417955719,
          "train_loss": 0.00427401065826416,
          "validation_loss": 0.003682337701320648,
          "train_metric": 0.00043046276550740004,
          "validation_metric": 0.0003708715667016804,
          "lr_first": 0.0022693831636851907,
          "lr_last": 0.002253558980209585,
          "seconds": 1.4701021189994208
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0059921069332631305,
          "train_loss": 0.004884126130491495,
          "validation_loss": 0.004313332960009575,
          "train_metric": 0.0004919113125652075,
          "validation_metric": 0.000434423127444461,
          "lr_first": 0.0022524999999999976,
          "lr_last": 0.0022365556560958325,
          "seconds": 1.4867890519999492
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.00538314989535138,
          "train_loss": 0.004749474115669727,
          "validation_loss": 0.0041917166672647,
          "train_metric": 0.0004783495969604701,
          "validation_metric": 0.0004221743147354573,
          "lr_first": 0.0022354887560379452,
          "lr_last": 0.002219426983686468,
          "seconds": 1.5034032149997074
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0055290802047238685,
          "train_loss": 0.004229568876326084,
          "validation_loss": 0.0034761354327201843,
          "train_metric": 0.0004259867127984762,
          "validation_metric": 0.00035010362626053393,
          "lr_first": 0.002218352346588113,
          "lr_last": 0.0022021758978913284,
          "seconds": 1.5211697390004701
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0047168705859803595,
          "train_loss": 0.005088552366942167,
          "validation_loss": 0.0036988428328186274,
          "train_metric": 0.0005125003517605364,
          "validation_metric": 0.00037253391928970814,
          "lr_first": 0.0022010937078860405,
          "lr_last": 0.002184805354595156,
          "seconds": 1.5389645289997134
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.006909428586368449,
          "train_loss": 0.017710203304886818,
          "validation_loss": 0.014367511495947838,
          "train_metric": 0.0017837068298831582,
          "validation_metric": 0.0014470430323854089,
          "lr_first": 0.0021837157971106207,
          "lr_last": 0.0021673183301511214,
          "seconds": 1.5562338370000361
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.010496564835193567,
          "train_loss": 0.0212322399020195,
          "validation_loss": 0.01379497442394495,
          "train_metric": 0.002138433512300253,
          "validation_metric": 0.001389379147440195,
          "lr_first": 0.002166221591877406,
          "lr_last": 0.0021497178208708407,
          "seconds": 1.5741557159999502
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.014673833444248885,
          "train_loss": 0.009992914274334908,
          "validation_loss": 0.007832761853933334,
          "train_metric": 0.001006449805572629,
          "validation_metric": 0.0007888870895840228,
          "lr_first": 0.0021486140897284006,
          "lr_last": 0.002132006842510969,
          "seconds": 1.5916246010001487
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.007168109186750371,
          "train_loss": 0.004521162714809179,
          "validation_loss": 0.004167466890066862,
          "train_metric": 0.0004553549224510789,
          "validation_metric": 0.0004197319794911891,
          "lr_first": 0.002130896307618455,
          "lr_last": 0.0021141884297564742,
          "seconds": 1.6091374629995698
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004966280757798813,
          "train_loss": 0.00435306690633297,
          "validation_loss": 0.0036179868038743734,
          "train_metric": 0.00043842490413226187,
          "validation_metric": 0.000364390347385779,
          "lr_first": 0.0021130712813983224,
          "lr_last": 0.0020962656357006496,
          "seconds": 1.6267102679994423
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004534911189693958,
          "train_loss": 0.004397424403578043,
          "validation_loss": 0.003413344267755747,
          "train_metric": 0.00044289251673035324,
          "validation_metric": 0.000343779509421438,
          "lr_first": 0.00209514206529448,
          "lr_last": 0.002078241531321991,
          "seconds": 1.6459673970002768
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.005380436799896415,
          "train_loss": 0.005194388329982758,
          "validation_loss": 0.0045433794148266315,
          "train_metric": 0.000523159746080637,
          "validation_metric": 0.00045759257045574486,
          "lr_first": 0.0020771117313858087,
          "lr_last": 0.0020601192049579968,
          "seconds": 1.6639289630002168
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004173387307673693,
          "train_loss": 0.006594696082174778,
          "validation_loss": 0.004887993447482586,
          "train_metric": 0.0006641935906372964,
          "validation_metric": 0.0004923008382320404,
          "lr_first": 0.0020589833690772045,
          "lr_last": 0.0020419017617759943,
          "seconds": 1.6816326360003586
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.005957176901574712,
          "train_loss": 0.011280108243227005,
          "validation_loss": 0.007655049674212933,
          "train_metric": 0.0011360910721123219,
          "validation_metric": 0.0007709884666837752,
          "lr_first": 0.0020407600845702236,
          "lr_last": 0.0020235923232410895,
          "seconds": 1.698434083000393
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.006135721450846177,
          "train_loss": 0.006941481493413448,
          "validation_loss": 0.005835406947880983,
          "train_metric": 0.0006991205736994743,
          "validation_metric": 0.0005877208895981312,
          "lr_first": 0.0020224450003308516,
          "lr_last": 0.002005194026581318,
          "seconds": 1.7152788719995442
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.005795098310045432,
          "train_loss": 0.009224599227309227,
          "validation_loss": 0.008569610305130482,
          "train_metric": 0.0009290678426623344,
          "validation_metric": 0.0008630998781882226,
          "lr_first": 0.0020040412545544867,
          "lr_last": 0.001986710024250099,
          "seconds": 1.7328079590006382
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.006119333658716641,
          "train_loss": 0.004988968372344971,
          "validation_loss": 0.0037223678082227707,
          "train_metric": 0.000502470531500876,
          "validation_metric": 0.00037490323302336037,
          "lr_first": 0.0019855520006282273,
          "lr_last": 0.0019681434833860745,
          "seconds": 1.7507479899995815
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.004443795063707512,
          "train_loss": 0.003811020404100418,
          "validation_loss": 0.0034251585602760315,
          "train_metric": 0.0003838320553768426,
          "validation_metric": 0.00034496941952966154,
          "lr_first": 0.0019669804065905482,
          "lr_last": 0.0019494975852704384,
          "seconds": 1.7683931280007528
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004663072642870247,
          "train_loss": 0.005586594808846712,
          "validation_loss": 0.0037754822988063097,
          "train_metric": 0.000562661443836987,
          "validation_metric": 0.0003802528080996126,
          "lr_first": 0.0019483296545884795,
          "lr_last": 0.0019307755247818476,
          "seconds": 1.7864414769992436
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.005070583087217528,
          "train_loss": 0.005787883419543505,
          "validation_loss": 0.0048891701735556126,
          "train_metric": 0.0005829343572258949,
          "validation_metric": 0.0004924192908219993,
          "lr_first": 0.0019296029403323655,
          "lr_last": 0.0019119805098489838,
          "seconds": 1.8038167089998751
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.007682548253796995,
          "train_loss": 0.006514808628708124,
          "validation_loss": 0.004743662197142839,
          "train_metric": 0.0006561477202922106,
          "validation_metric": 0.0004777642898261547,
          "lr_first": 0.001910803472548287,
          "lr_last": 0.0018931157609008988,
          "seconds": 1.820667598
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.006093055839301087,
          "train_loss": 0.0064359693787992,
          "validation_loss": 0.00619365694001317,
          "train_metric": 0.0006482072058133781,
          "validation_metric": 0.0006238024216145277,
          "lr_first": 0.0018919344724282693,
          "lr_last": 0.0018741845103152086,
          "seconds": 1.8380561560006754
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.004456404276425019,
          "train_loss": 0.0041289664804935455,
          "validation_loss": 0.0029522611293941736,
          "train_metric": 0.00041585430153645575,
          "validation_metric": 0.0002973409427795559,
          "lr_first": 0.0018729991730783457,
          "lr_last": 0.0018551900018642434,
          "seconds": 1.8561947979997058
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.004705163031758275,
          "train_loss": 0.0074348668567836285,
          "validation_loss": 0.006543965078890324,
          "train_metric": 0.0007488125702366233,
          "validation_metric": 0.0006590841803699732,
          "lr_first": 0.00185400081896458,
          "lr_last": 0.0018361354901592413,
          "seconds": 1.8736864519996743
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004130573928705417,
          "train_loss": 0.003651684382930398,
          "validation_loss": 0.0027447547763586044,
          "train_metric": 0.000367784290574491,
          "validation_metric": 0.0002764416567515582,
          "lr_first": 0.0018349426653571461,
          "lr_last": 0.001817024240092691,
          "seconds": 1.8923766949992569
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003771734787733294,
          "train_loss": 0.003523301100358367,
          "validation_loss": 0.0031950126867741346,
          "train_metric": 0.00035485398257151246,
          "validation_metric": 0.00032179002300836146,
          "lr_first": 0.0018158279777725514,
          "lr_last": 0.0017978595262789054,
          "seconds": 1.9093828389995906
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.00392830854980275,
          "train_loss": 0.0058105154894292355,
          "validation_loss": 0.0038438322953879833,
          "train_metric": 0.0005852138274349272,
          "validation_metric": 0.0003871366789098829,
          "lr_first": 0.0017966600314141131,
          "lr_last": 0.001778644632492937,
          "seconds": 1.9260808039998665
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004424143895448651,
          "train_loss": 0.009961973875761032,
          "validation_loss": 0.008678271435201168,
          "train_metric": 0.0010033336002379656,
          "validation_metric": 0.0008740436751395464,
          "lr_first": 0.0017774421106107614,
          "lr_last": 0.0017593828511079182,
          "seconds": 2.041298697999082
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.005590688590018544,
          "train_loss": 0.003375086933374405,
          "validation_loss": 0.0026494250632822514,
          "train_metric": 0.0003399263950996101,
          "validation_metric": 0.0002668403903953731,
          "lr_first": 0.0017581775082542933,
          "lr_last": 0.0017400774825309276,
          "seconds": 2.0601104119996307
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.003521689744957257,
          "train_loss": 0.0032802908681333065,
          "validation_loss": 0.00279081123881042,
          "train_metric": 0.0003303788835182786,
          "validation_metric": 0.00028108031256124377,
          "lr_first": 0.0017388695252351471,
          "lr_last": 0.001720731834637488,
          "seconds": 2.0774069929993857
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.004158441653999034,
          "train_loss": 0.005047429818660021,
          "validation_loss": 0.003503147978335619,
          "train_metric": 0.0005083586438558996,
          "validation_metric": 0.0003528242523316294,
          "lr_first": 0.001719521469876815,
          "lr_last": 0.0017013492222047732,
          "seconds": 2.095286067999041
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004058905396959744,
          "train_loss": 0.004154746886342764,
          "validation_loss": 0.004102368373423815,
          "train_metric": 0.00041845085797831416,
          "validation_metric": 0.00041317552677355707,
          "lr_first": 0.0017001366573689788,
          "lr_last": 0.0016819329663436428,
          "seconds": 2.112538940000377
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.0036807692085858434,
          "train_loss": 0.0034017199650406837,
          "validation_loss": 0.003098095767199993,
          "train_metric": 0.00034260874963365495,
          "validation_metric": 0.0003120288602076471,
          "lr_first": 0.0016807184091994691,
          "lr_last": 0.0016624863939295827,
          "seconds": 2.1306466949990863
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0034619671205291525,
          "train_loss": 0.0038415270391851664,
          "validation_loss": 0.003471375210210681,
          "train_metric": 0.00038690454675816,
          "validation_metric": 0.00034962414065375924,
          "lr_first": 0.001661270052585144,
          "lr_last": 0.0016430128370326644,
          "seconds": 2.14813735899952
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.00389003715099534,
          "train_loss": 0.003688469296321273,
          "validation_loss": 0.002730254316702485,
          "train_metric": 0.0003714891499839723,
          "validation_metric": 0.0002749812265392393,
          "lr_first": 0.0016417949199017884,
          "lr_last": 0.0016235156323466118,
          "seconds": 2.1659546390001196
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004175188747467473,
          "train_loss": 0.003578161122277379,
          "validation_loss": 0.003222572384402156,
          "train_metric": 0.00036037928657606244,
          "validation_metric": 0.0003245656844228506,
          "lr_first": 0.001622296348113129,
          "lr_last": 0.0016039981206170785,
          "seconds": 2.1828131889997167
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0037590605752484407,
          "train_loss": 0.0035156894009560347,
          "validation_loss": 0.0030441435519605875,
          "train_metric": 0.0003540873876772821,
          "validation_metric": 0.00030659500043839216,
          "lr_first": 0.0016027776781990647,
          "lr_last": 0.0015844636460692202,
          "seconds": 2.200214987999061
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003264231578214094,
          "train_loss": 0.004011242650449276,
          "validation_loss": 0.002853727899491787,
          "train_metric": 0.0004039976338390261,
          "validation_metric": 0.00028741700225509703,
          "lr_first": 0.0015832422545832014,
          "lr_last": 0.0015649155558346884,
          "seconds": 2.216935526999805
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0037720314794569276,
          "train_loss": 0.0030000112019479275,
          "validation_loss": 0.0023646829649806023,
          "train_metric": 0.0003021501761395484,
          "validation_metric": 0.0002381621889071539,
          "lr_first": 0.0015636934245598084,
          "lr_last": 0.0015453571993781102,
          "seconds": 2.234615997000219
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0033819125092122704,
          "train_loss": 0.0029604427982121706,
          "validation_loss": 0.002605428220704198,
          "train_metric": 0.00029816493042744696,
          "validation_metric": 0.0002624091866891831,
          "lr_first": 0.0015441345377202708,
          "lr_last": 0.0015257919279231778,
          "seconds": 2.2519101510006294
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.002960735509986989,
          "train_loss": 0.0026892749592661858,
          "validation_loss": 0.002305225469172001,
          "train_metric": 0.0002708539250306785,
          "validation_metric": 0.0002321738429600373,
          "lr_first": 0.0015245689453791606,
          "lr_last": 0.0015062230938784333,
          "seconds": 2.270145173999481
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004802550516615156,
          "train_loss": 0.009682489559054375,
          "validation_loss": 0.006195166613906622,
          "train_metric": 0.0009751848410815,
          "validation_metric": 0.0006239545182324946,
          "lr_first": 0.001505000000000001,
          "lr_last": 0.0014866540502628492,
          "seconds": 2.2869949720006844
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.007125715925212717,
          "train_loss": 0.003451906843110919,
          "validation_loss": 0.0029023096431046724,
          "train_metric": 0.0003476634155958891,
          "validation_metric": 0.000292309996439144,
          "lr_first": 0.0014854310546208404,
          "lr_last": 0.0014670881501313076,
          "seconds": 2.3042627569993783
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.005241642837063409,
          "train_loss": 0.0038496116176247597,
          "validation_loss": 0.003631044877693057,
          "train_metric": 0.0003877187555190176,
          "validation_metric": 0.0003657055785879493,
          "lr_first": 0.0014658654622797308,
          "lr_last": 0.0014475287460000708,
          "seconds": 2.321801907999543
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.003858392119582277,
          "train_loss": 0.003502295585349202,
          "validation_loss": 0.002822090405970812,
          "train_metric": 0.0003527383378241211,
          "validation_metric": 0.000284230598481372,
          "lr_first": 0.0014463065754401938,
          "lr_last": 0.001427979189272343,
          "seconds": 2.3383712520007975
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0037518762874242384,
          "train_loss": 0.002955059055238962,
          "validation_loss": 0.0024677892215549946,
          "train_metric": 0.000297622726066038,
          "validation_metric": 0.00024854668299667537,
          "lr_first": 0.0014267577454168008,
          "lr_last": 0.0014084428296640297,
          "seconds": 2.354844792000222
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.003264395138103282,
          "train_loss": 0.0029270777013152838,
          "validation_loss": 0.0022846702486276627,
          "train_metric": 0.0002948045439552516,
          "validation_metric": 0.0002301036292919889,
          "lr_first": 0.001407222321800938,
          "lr_last": 0.0013889230146297755,
          "seconds": 2.372433562999504
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0035303977274452336,
          "train_loss": 0.003151364391669631,
          "validation_loss": 0.0023622652515769005,
          "train_metric": 0.0003173938603140414,
          "validation_metric": 0.0002379186771577224,
          "lr_first": 0.0013877036518868736,
          "lr_last": 0.0013694230887893984,
          "seconds": 2.389794427999732
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0037527243039221503,
          "train_loss": 0.003293094225227833,
          "validation_loss": 0.0029193481896072626,
          "train_metric": 0.0003316683869343251,
          "validation_metric": 0.0002940260455943644,
          "lr_first": 0.001368205080098215,
          "lr_last": 0.0013499463933548063,
          "seconds": 2.4075715279996075
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0033558494833414443,
          "train_loss": 0.002587852068245411,
          "validation_loss": 0.0021636318415403366,
          "train_metric": 0.0002606390044093132,
          "validation_metric": 0.00021791309700347483,
          "lr_first": 0.00134872994741486,
          "lr_last": 0.0013304962655574931,
          "seconds": 2.448243494000053
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0030792401084909216,
          "train_loss": 0.002686889609321952,
          "validation_loss": 0.002148510655388236,
          "train_metric": 0.00027061370201408863,
          "validation_metric": 0.00021639010810758919,
          "lr_first": 0.0013292815908005342,
          "lr_last": 0.0013110760380767297,
          "seconds": 2.4658127349994174
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.002903830936702434,
          "train_loss": 0.0028690695762634277,
          "validation_loss": 0.002691016299650073,
          "train_metric": 0.00028896218282170594,
          "validation_metric": 0.0002710293047130108,
          "lr_first": 0.0013098633426310243,
          "lr_last": 0.0012916890384685182,
          "seconds": 2.4826728949992685
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0036908760885125957,
          "train_loss": 0.002929008100181818,
          "validation_loss": 0.0020899649243801832,
          "train_metric": 0.000294998986646533,
          "validation_metric": 0.00021049361384939402,
          "lr_first": 0.0012904785301231876,
          "lr_last": 0.0012723385885954383,
          "seconds": 2.500818795999294
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0031118146507651545,
          "train_loss": 0.002596708945930004,
          "validation_loss": 0.001980316359549761,
          "train_metric": 0.0002615310368128121,
          "validation_metric": 0.0001994502090383321,
          "lr_first": 0.001271130474764856,
          "lr_last": 0.0012530280040574575,
          "seconds": 2.5183324629997514
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.002899355757108424,
          "train_loss": 0.0027116364799439907,
          "validation_loss": 0.002112343907356262,
          "train_metric": 0.0002731060958467424,
          "validation_metric": 0.00021274750179145485,
          "lr_first": 0.0012518224917457093,
          "lr_last": 0.001233760593623827,
          "seconds": 2.5353660069995385
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0030625395411334466,
          "train_loss": 0.0028333389200270176,
          "validation_loss": 0.0026317450683563948,
          "train_metric": 0.00028536352328956127,
          "validation_metric": 0.0002650597016327083,
          "lr_first": 0.001232557889389241,
          "lr_last": 0.0012145396586661337,
          "seconds": 2.551725997000176
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004016841241536895,
          "train_loss": 0.003974650055170059,
          "validation_loss": 0.0036809933371841908,
          "train_metric": 0.00040031224489212036,
          "validation_metric": 0.0003707361756823957,
          "lr_first": 0.0012133399685858898,
          "lr_last": 0.0011953684925926351,
          "seconds": 2.5683339520001027
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003961379516113084,
          "train_loss": 0.0034665081184357405,
          "validation_loss": 0.0025009880773723125,
          "train_metric": 0.00034913403214886785,
          "validation_metric": 0.0002518903638701886,
          "lr_first": 0.0011941720222274513,
          "lr_last": 0.0011762503802839421,
          "seconds": 2.584718445999897
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0041010062595887575,
          "train_loss": 0.0023443468380719423,
          "validation_loss": 0.00197978806681931,
          "train_metric": 0.00023611399228684604,
          "validation_metric": 0.00019939699268434197,
          "lr_first": 0.001175057334642857,
          "lr_last": 0.001157188597530181,
          "seconds": 2.602137997999307
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.003085446918703383,
          "train_loss": 0.002449149265885353,
          "validation_loss": 0.0021404370199888945,
          "train_metric": 0.0002466693695168942,
          "validation_metric": 0.00021557699074037373,
          "lr_first": 0.0011559991810354228,
          "lr_last": 0.0011381864104696943,
          "seconds": 2.6187301219997607
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.003253339178627357,
          "train_loss": 0.0035648939665406942,
          "validation_loss": 0.003294103778898716,
          "train_metric": 0.0003590430715121329,
          "validation_metric": 0.00033177004661411047,
          "lr_first": 0.0011370008269216568,
          "lr_last": 0.0011192470750294122,
          "seconds": 2.6357096009996894
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.003729358606506139,
          "train_loss": 0.0026112052146345377,
          "validation_loss": 0.002005320508033037,
          "train_metric": 0.0002629910595715046,
          "validation_metric": 0.00020196854893583804,
          "lr_first": 0.0011180655275717327,
          "lr_last": 0.0011003738363669622,
          "seconds": 2.652315104000081
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.002579557112767361,
          "train_loss": 0.002310127718374133,
          "validation_loss": 0.0020448947325348854,
          "train_metric": 0.00023266757489182055,
          "validation_metric": 0.0002059543039649725,
          "lr_first": 0.001099196527451715,
          "lr_last": 0.0010815699283146308,
          "seconds": 2.6686788009992597
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0024097221685224213,
          "train_loss": 0.002583614317700267,
          "validation_loss": 0.002407585736364126,
          "train_metric": 0.0002602121967356652,
          "validation_metric": 0.00024248322006314993,
          "lr_first": 0.0010803970596676372,
          "lr_last": 0.0010628385728252573,
          "seconds": 2.7856324889999087
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0026452244874235475,
          "train_loss": 0.0022593445610255003,
          "validation_loss": 0.0019526699325069785,
          "train_metric": 0.00022755295503884554,
          "validation_metric": 0.00019666575826704502,
          "lr_first": 0.0010616703454115225,
          "lr_last": 0.0010441829794201776,
          "seconds": 2.8036895299992466
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0025317960753454827,
          "train_loss": 0.003003488527610898,
          "validation_loss": 0.002030471106991172,
          "train_metric": 0.0003025003825314343,
          "validation_metric": 0.00020450164447538555,
          "lr_first": 0.001043019593409455,
          "lr_last": 0.0010256063446392812,
          "seconds": 2.820490879999852
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.002649924164870754,
          "train_loss": 0.003346616867929697,
          "validation_loss": 0.0021903591696172953,
          "train_metric": 0.00033705896930769086,
          "validation_metric": 0.0002206049393862486,
          "lr_first": 0.0010244479993717752,
          "lr_last": 0.001007111851493306,
          "seconds": 2.8369622619993606
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0026152136451855768,
          "train_loss": 0.0022293408401310444,
          "validation_loss": 0.0018448764458298683,
          "train_metric": 0.00022453100245911628,
          "validation_metric": 0.0001858092291513458,
          "lr_first": 0.0010059587454455136,
          "lr_last": 0.0009887026689184423,
          "seconds": 2.854419125000277
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0025685949221951887,
          "train_loss": 0.0022953893058001995,
          "validation_loss": 0.0020819446071982384,
          "train_metric": 0.00023118322133086622,
          "validation_metric": 0.00020968583703506738,
          "lr_first": 0.0009875549996691494,
          "lr_last": 0.0009703819512333538,
          "seconds": 2.870947157000046
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.002312010583409574,
          "train_loss": 0.002137363888323307,
          "validation_loss": 0.0017982295248657465,
          "train_metric": 0.00021526745695155114,
          "validation_metric": 0.00018111109966412187,
          "lr_first": 0.0009692399154297775,
          "lr_last": 0.0009521528375986995,
          "seconds": 2.888520291999157
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.002273390353366267,
          "train_loss": 0.0025140256620943546,
          "validation_loss": 0.0018870464991778135,
          "train_metric": 0.0002532034122850746,
          "validation_metric": 0.00019005636568181217,
          "lr_first": 0.0009510166309227964,
          "lr_last": 0.0009340184514792582,
          "seconds": 2.905863325999235
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.002307065213244641,
          "train_loss": 0.0027239080518484116,
          "validation_loss": 0.0019195199711248279,
          "train_metric": 0.00027434201911091805,
          "validation_metric": 0.00019332701049279422,
          "lr_first": 0.0009328882686141923,
          "lr_last": 0.0009159819001087345,
          "seconds": 2.922937006000211
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0032943594378593843,
          "train_loss": 0.006299200467765331,
          "validation_loss": 0.004232880659401417,
          "train_metric": 0.000634432421065867,
          "validation_metric": 0.0004263203009031713,
          "lr_first": 0.0009148579347055221,
          "lr_last": 0.0008980462739573539,
          "seconds": 2.9401036130002467
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.006937366735655814,
          "train_loss": 0.006319642998278141,
          "validation_loss": 0.003634357126429677,
          "train_metric": 0.0006364912260323763,
          "validation_metric": 0.0003660391375888139,
          "lr_first": 0.0008969287186016802,
          "lr_last": 0.000880214646202324,
          "seconds": 2.9564317040003516
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0029666348345926963,
          "train_loss": 0.0023308859672397375,
          "validation_loss": 0.001848874380812049,
          "train_metric": 0.0002347582922084257,
          "validation_metric": 0.00018621185154188424,
          "lr_first": 0.0008791036923815459,
          "lr_last": 0.0008624900722012644,
          "seconds": 2.9727240880001773
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.002728487645072164,
          "train_loss": 0.0022519647609442472,
          "validation_loss": 0.001955418847501278,
          "train_metric": 0.00022680967231281102,
          "validation_metric": 0.00019694262300617993,
          "lr_first": 0.0008613859102716003,
          "lr_last": 0.0008448755889686828,
          "seconds": 2.989545268999791
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.00217493293894222,
          "train_loss": 0.00241066818125546,
          "validation_loss": 0.0022305031307041645,
          "train_metric": 0.0002427936706226319,
          "validation_metric": 0.00022464808716904372,
          "lr_first": 0.0008437784081225947,
          "lr_last": 0.000827374214655601,
          "seconds": 3.007543960999101
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.002568972744484199,
          "train_loss": 0.002449438674375415,
          "validation_loss": 0.002226038370281458,
          "train_metric": 0.00024669847334735096,
          "validation_metric": 0.0002241984475404024,
          "lr_first": 0.0008262842028893788,
          "lr_last": 0.0008099889480324116,
          "seconds": 3.025165126999127
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.002307692582689924,
          "train_loss": 0.00281748129054904,
          "validation_loss": 0.002628678921610117,
          "train_metric": 0.00028376636328175664,
          "validation_metric": 0.0002647509390953928,
          "lr_first": 0.0008089062921139598,
          "lr_last": 0.0007927227679750494,
          "seconds": 3.042764358999193
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.002572721343312878,
          "train_loss": 0.002545326016843319,
          "validation_loss": 0.002263410948216915,
          "train_metric": 0.00025635591009631753,
          "validation_metric": 0.000227962460485287,
          "lr_first": 0.0007916476534118866,
          "lr_last": 0.0007755786329545818,
          "seconds": 3.0598835159999
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0022898796960362233,
          "train_loss": 0.0021364318672567606,
          "validation_loss": 0.0017564193112775683,
          "train_metric": 0.00021517358254641294,
          "validation_metric": 0.00017690012464299798,
          "lr_first": 0.0007745112439620533,
          "lr_last": 0.0007585594805302867,
          "seconds": 3.077694228000837
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0021316471029422246,
          "train_loss": 0.002158189658075571,
          "validation_loss": 0.001762334257364273,
          "train_metric": 0.00021736497001256794,
          "validation_metric": 0.00017749588005244732,
          "lr_first": 0.0007575000000000017,
          "lr_last": 0.0007416682268463151,
          "seconds": 3.095510295999702
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0023351027266471647,
          "train_loss": 0.0025381469167768955,
          "validation_loss": 0.0017691563116386533,
          "train_metric": 0.0002556328836362809,
          "validation_metric": 0.0001781829196261242,
          "lr_first": 0.0007406168363148086,
          "lr_last": 0.000724907766132032,
          "seconds": 3.112617834000048
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0023777357419021428,
          "train_loss": 0.0024680302012711763,
          "validation_loss": 0.0017938928212970495,
          "train_metric": 0.00024857098469510674,
          "validation_metric": 0.0001806743093766272,
          "lr_first": 0.0007238646457496579,
          "lr_last": 0.0007082809702060954,
          "seconds": 3.1293796509999083
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.002526291646063328,
          "train_loss": 0.0020850994624197483,
          "validation_loss": 0.0017593179363757372,
          "train_metric": 0.00021000357810407877,
          "validation_metric": 0.00017719205061439425,
          "lr_first": 0.000707246298706162,
          "lr_last": 0.0006917906879843936,
          "seconds": 3.1467986299994664
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.002481993033143226,
          "train_loss": 0.0026356850285083055,
          "validation_loss": 0.0018704135436564684,
          "train_metric": 0.00026545656146481633,
          "validation_metric": 0.0001883812074083835,
          "lr_first": 0.0006907646426525363,
          "lr_last": 0.0006754397449918927,
          "seconds": 3.1634915720005665
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.003075890836043982,
          "train_loss": 0.002323625609278679,
          "validation_loss": 0.0020580922719091177,
          "train_metric": 0.00023402705846820027,
          "validation_metric": 0.00020728353410959244,
          "lr_first": 0.0006744225016356963,
          "lr_last": 0.0006592309428784994,
          "seconds": 3.180694004000543
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.002977398798975628,
          "train_loss": 0.0025822960305958986,
          "validation_loss": 0.0024260058999061584,
          "train_metric": 0.0002600794250611216,
          "validation_metric": 0.0002443384437356144,
          "lr_first": 0.0006582226757973757,
          "lr_last": 0.0006431670589390128,
          "seconds": 3.1978459670008306
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0024743740432313643,
          "train_loss": 0.001945280353538692,
          "validation_loss": 0.001703765825368464,
          "train_metric": 0.00019592151511460543,
          "validation_metric": 0.00017159705748781562,
          "lr_first": 0.0006421679408943365,
          "lr_last": 0.0006272508456372499,
          "seconds": 3.2157864670007257
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0020498063895502128,
          "train_loss": 0.0019388191867619753,
          "validation_loss": 0.0016389975789934397,
          "train_metric": 0.0001952707680175081,
          "validation_metric": 0.00016507385589648038,
          "lr_first": 0.0006262610478227541,
          "lr_last": 0.0006114850301344212,
          "seconds": 3.23357131100056
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0020262611578800716,
          "train_loss": 0.0020765671506524086,
          "validation_loss": 0.001886236248537898,
          "train_metric": 0.00020914425840601325,
          "validation_metric": 0.0001899748167488724,
          "lr_first": 0.0006105047221468661,
          "lr_last": 0.0005958723138218511,
          "seconds": 3.250296594000247
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.002073356525215786,
          "train_loss": 0.0019363347673788667,
          "validation_loss": 0.001606847858056426,
          "train_metric": 0.00019502056238707155,
          "validation_metric": 0.00016183583647944033,
          "lr_first": 0.0005949016636319636,
          "lr_last": 0.0005804153718581059,
          "seconds": 3.267143618000773
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.002044210625172127,
          "train_loss": 0.0019721484277397394,
          "validation_loss": 0.0015450094360858202,
          "train_metric": 0.00019862756016664207,
          "validation_metric": 0.00015560768952127546,
          "lr_first": 0.0005794545457817997,
          "lr_last": 0.0005651168527106182,
          "seconds": 3.284947439999087
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.00215798719364102,
          "train_loss": 0.0019043623469769955,
          "validation_loss": 0.0016514547169208527,
          "train_metric": 0.00019180038361810148,
          "validation_metric": 0.00016632849292363971,
          "lr_first": 0.0005641660153804942,
          "lr_last": 0.0005499793777018868,
          "seconds": 3.3014217859999917
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0018669443215912906,
          "train_loss": 0.002553333528339863,
          "validation_loss": 0.0017905571730807424,
          "train_metric": 0.0002571624063421041,
          "validation_metric": 0.00018033836386166513,
          "lr_first": 0.000549038692039027,
          "lr_last": 0.0005350055405603265,
          "seconds": 3.3189635510007065
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.002890486819524085,
          "train_loss": 0.002695917384698987,
          "validation_loss": 0.0018946971977129579,
          "train_metric": 0.0002715229638852179,
          "validation_metric": 0.00019082697690464556,
          "lr_first": 0.0005340751677463768,
          "lr_last": 0.0005201979069758457,
          "seconds": 3.3361028719991737
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.003583319361496251,
          "train_loss": 0.0020202931482344866,
          "validation_loss": 0.0017550443299114704,
          "train_metric": 0.00020347654935903847,
          "validation_metric": 0.0001767616340657696,
          "lr_first": 0.0005192780064253987,
          "lr_last": 0.0005055590141602302,
          "seconds": 3.352870393999183
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0028599829020095058,
          "train_loss": 0.002039683051407337,
          "validation_loss": 0.0016666343435645103,
          "train_metric": 0.0002054294163826853,
          "validation_metric": 0.00016785731713753194,
          "lr_first": 0.0005046497434935084,
          "lr_last": 0.0004910913704124016,
          "seconds": 3.36903193299986
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0021255863539408892,
          "train_loss": 0.0018984234193339944,
          "validation_loss": 0.0016350183868780732,
          "train_metric": 0.00019120225624646991,
          "validation_metric": 0.00016467308159917593,
          "lr_first": 0.0004901928854282528,
          "lr_last": 0.0004767974546886371,
          "seconds": 3.385670489000404
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.001998800307774218,
          "train_loss": 0.0019094679737463593,
          "validation_loss": 0.0015076580457389355,
          "train_metric": 0.00019231460464652628,
          "validation_metric": 0.0001518458011560142,
          "lr_first": 0.0004759099093378383,
          "lr_last": 0.00046267971617781114,
          "seconds": 3.5046830890005367
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0022969020865275525,
          "train_loss": 0.0025769060011953115,
          "validation_loss": 0.0017612017691135406,
          "train_metric": 0.00025953655131161213,
          "validation_metric": 0.00017738179303705692,
          "lr_first": 0.0004618032625366892,
          "lr_last": 0.00044874057388173796,
          "seconds": 3.523269203000382
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002372177623328753,
          "train_loss": 0.002435145201161504,
          "validation_loss": 0.0017385244136676192,
          "train_metric": 0.0002452589396853,
          "validation_metric": 0.00017509782628621906,
          "lr_first": 0.0004478753621261122,
          "lr_last": 0.00043498241620069174,
          "seconds": 3.5405734550004127
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.002384761977737071,
          "train_loss": 0.0018931108061224222,
          "validation_loss": 0.001637766370549798,
          "train_metric": 0.00019066716777160764,
          "validation_metric": 0.000164949829922989,
          "lr_first": 0.00043412859458014243,
          "lr_last": 0.0004214076005241641,
          "seconds": 3.5582234619996598
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0022188766342878807,
          "train_loss": 0.0023871506564319134,
          "validation_loss": 0.0021532659884542227,
          "train_metric": 0.00024042506993282586,
          "validation_metric": 0.00021686906984541565,
          "lr_first": 0.0004205653153366304,
          "lr_last": 0.00040801845282693807,
          "seconds": 3.575383077999504
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0019723605837498326,
          "train_loss": 0.0019219480454921722,
          "validation_loss": 0.0015394806396216154,
          "train_metric": 0.00019357155542820692,
          "validation_metric": 0.000155050860485062,
          "lr_first": 0.0004071878483936508,
          "lr_last": 0.0003948172672705435,
          "seconds": 3.5933872690002318
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0019186427016393282,
          "train_loss": 0.0018788283923640847,
          "validation_loss": 0.0014654338592663407,
          "train_metric": 0.00018922869639936835,
          "validation_metric": 0.00014759312034584582,
          "lr_first": 0.00039399848591129634,
          "lr_last": 0.0003818063058101653,
          "seconds": 3.6115538769990962
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0018919489157269709,
          "train_loss": 0.0018009424675256014,
          "validation_loss": 0.0015308299334719777,
          "train_metric": 0.00018138432642444968,
          "validation_metric": 0.00015417956456076354,
          "lr_first": 0.0003809994878189294,
          "lr_last": 0.00036898779780706985,
          "seconds": 3.628712354999152
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.001841742670876556,
          "train_loss": 0.001803146442398429,
          "validation_loss": 0.001499534584581852,
          "train_metric": 0.00018160631589125842,
          "validation_metric": 0.0001510276342742145,
          "lr_first": 0.00036819308142795417,
          "lr_last": 0.000356363939646614,
          "seconds": 3.6464904810000007
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.002140743516065413,
          "train_loss": 0.0020356676541268826,
          "validation_loss": 0.0018206194508820772,
          "train_metric": 0.00020502498955465853,
          "validation_metric": 0.0001833661226555705,
          "lr_first": 0.0003555814610501785,
          "lr_last": 0.0003439368943619046,
          "seconds": 3.6636598570003116
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.002176133159082383,
          "train_loss": 0.0018124831840395927,
          "validation_loss": 0.001579855801537633,
          "train_metric": 0.0001825466752052307,
          "validation_metric": 0.00015911729133222252,
          "lr_first": 0.0003431667876218289,
          "lr_last": 0.0003317087912631771,
          "seconds": 3.6811286739994102
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0018789709283737466,
          "train_loss": 0.001959177665412426,
          "validation_loss": 0.0014908561715856194,
          "train_metric": 0.00019732120563276112,
          "validation_metric": 0.0001501536025898531,
          "lr_first": 0.00033095118833328664,
          "lr_last": 0.00031968172557294704,
          "seconds": 3.6989258310004516
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.002023745146289002,
          "train_loss": 0.0018676096806302667,
          "validation_loss": 0.001428284216672182,
          "train_metric": 0.0001880988129414618,
          "validation_metric": 0.0001438515610061586,
          "lr_first": 0.00031893675626460375,
          "lr_last": 0.00030785775806700536,
          "seconds": 3.717136556000696
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.0018183473584940657,
          "train_loss": 0.0017691772663965821,
          "validation_loss": 0.0014757192693650723,
          "train_metric": 0.00017818504420574754,
          "validation_metric": 0.00014862902753520757,
          "lr_first": 0.000307125550026865,
          "lr_last": 0.00029623891472131333,
          "seconds": 3.734594924999328
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0018914655502157984,
          "train_loss": 0.0019149467116221786,
          "validation_loss": 0.0014758111210539937,
          "train_metric": 0.00019286641327198595,
          "validation_metric": 0.00014863826800137758,
          "lr_first": 0.0002955195934094543,
          "lr_last": 0.00028482718636486487,
          "seconds": 3.7522062600000936
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.001851631393947173,
          "train_loss": 0.0017949688481166959,
          "validation_loss": 0.0015465071192011237,
          "train_metric": 0.00018078269204124808,
          "validation_metric": 0.0001557585346745327,
          "lr_first": 0.0002841208750332906,
          "lr_last": 0.0002736245283385659,
          "seconds": 3.770035043999087
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0018082212445733603,
          "train_loss": 0.0017544702859595418,
          "validation_loss": 0.001495448173955083,
          "train_metric": 0.00017670380475465208,
          "validation_metric": 0.00015061604790389538,
          "lr_first": 0.00027293134801008756,
          "lr_last": 0.00026263286016019817,
          "seconds": 3.78695387299922
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.0017914954078150913,
          "train_loss": 0.0020556109957396984,
          "validation_loss": 0.0015096935676410794,
          "train_metric": 0.00020703359041363,
          "validation_metric": 0.00015205082308966666,
          "lr_first": 0.0002619529296076959,
          "lr_last": 0.0002518540651955178,
          "seconds": 3.8038661280006636
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0020132281533733476,
          "train_loss": 0.001785198925063014,
          "validation_loss": 0.0015593321295455098,
          "train_metric": 0.00017979869153350592,
          "validation_metric": 0.00015705020632594824,
          "lr_first": 0.00025118750092159196,
          "lr_last": 0.0002412899903355538,
          "seconds": 3.821307569000055
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0018096509520546533,
          "train_loss": 0.0017514406936243176,
          "validation_loss": 0.0014179275603964925,
          "train_metric": 0.0001763986801961437,
          "validation_metric": 0.00014280847972258925,
          "lr_first": 0.00024063690655256104,
          "lr_last": 0.00023094244568014875,
          "seconds": 3.8403421879993402
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0017741380179359112,
          "train_loss": 0.0017427927814424038,
          "validation_loss": 0.0014849697472527623,
          "train_metric": 0.00017552771896589547,
          "validation_metric": 0.00014956069935578853,
          "lr_first": 0.0002303029542906331,
          "lr_last": 0.00022081320422781038,
          "seconds": 3.8580019590008305
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0017504577608633554,
          "train_loss": 0.001737695885822177,
          "validation_loss": 0.001428878866136074,
          "train_metric": 0.00017501437105238438,
          "validation_metric": 0.00014391147124115378,
          "lr_first": 0.00022018741480532905,
          "lr_last": 0.00021090400157191562,
          "seconds": 3.876058906000253
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0017597238984308206,
          "train_loss": 0.0017287284135818481,
          "validation_loss": 0.0014584545278921723,
          "train_metric": 0.00017411119188182056,
          "validation_metric": 0.00014689019008073956,
          "lr_first": 0.00021029202134226492,
          "lr_last": 0.00020121653560332717,
          "seconds": 3.8981381569992664
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0017909517719090218,
          "train_loss": 0.0017761485651135445,
          "validation_loss": 0.001551697263494134,
          "train_metric": 0.00017888717411551625,
          "validation_metric": 0.00015628125402145088,
          "lr_first": 0.00020061846942616898,
          "lr_last": 0.00019175246621946768,
          "seconds": 3.91777172699949
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.001891089865239337,
          "train_loss": 0.0017485471908003092,
          "validation_loss": 0.00151586695574224,
          "train_metric": 0.00017610727809369564,
          "validation_metric": 0.00015267255366779864,
          "lr_first": 0.0001911684165703626,
          "lr_last": 0.00018251341503990455,
          "seconds": 3.934841349000635
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.00179568279418163,
          "train_loss": 0.0017194231040775776,
          "validation_loss": 0.0014413910685107112,
          "train_metric": 0.00017317399033345282,
          "validation_metric": 0.00014517162344418466,
          "lr_first": 0.00018194348199275317,
          "lr_last": 0.00017350096512849466,
          "seconds": 3.9528443460003473
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0017567186623637099,
          "train_loss": 0.0017181806033477187,
          "validation_loss": 0.001442562323063612,
          "train_metric": 0.00017304884386248887,
          "validation_metric": 0.0001452895812690258,
          "lr_first": 0.00017294524633839077,
          "lr_last": 0.00016471666072213429,
          "seconds": 3.9707362549997924
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0017750615807017311,
          "train_loss": 0.0017288140952587128,
          "validation_loss": 0.0014030812308192253,
          "train_metric": 0.00017411982116755098,
          "validation_metric": 0.00014131319767329842,
          "lr_first": 0.00016417525140863167,
          "lr_last": 0.00015616200696616197,
          "seconds": 3.9895082149996597
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0017802590336941648,
          "train_loss": 0.001711469260044396,
          "validation_loss": 0.0014308929676190019,
          "train_metric": 0.00017237290740013123,
          "validation_metric": 0.00014411431038752198,
          "lr_first": 0.00015563499989695927,
          "lr_last": 0.00014783846965645865,
          "seconds": 4.007634612000402
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.001748504819261143,
          "train_loss": 0.0017109343316406012,
          "validation_loss": 0.0014222253812476993,
          "train_metric": 0.00017231905076187104,
          "validation_metric": 0.00014324134099297225,
          "lr_first": 0.00014732595513150429,
          "lr_last": 0.0001397474749882931,
          "seconds": 4.025261175000196
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0018259296648466261,
          "train_loss": 0.0017133672954514623,
          "validation_loss": 0.0014177034609019756,
          "train_metric": 0.00017256406135857105,
          "validation_metric": 0.0001427858805982396,
          "lr_first": 0.0001392495408243121,
          "lr_last": 0.00013189040931194934,
          "seconds": 4.042855911000515
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0017588107803021558,
          "train_loss": 0.001715411781333387,
          "validation_loss": 0.0014023425756022334,
          "train_metric": 0.00017277001461479813,
          "validation_metric": 0.00014123879373073578,
          "lr_first": 0.00013140714082739743,
          "lr_last": 0.00012426861889518266,
          "seconds": 4.060898432000613
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0017620093021832872,
          "train_loss": 0.0017073806375265121,
          "validation_loss": 0.0014442536048591137,
          "train_metric": 0.00017196113185491413,
          "validation_metric": 0.00014545991143677384,
          "lr_first": 0.00012380009889562699,
          "lr_last": 0.00011688340969254351,
          "seconds": 4.078147790000003
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0017391563233104534,
          "train_loss": 0.001702042412944138,
          "validation_loss": 0.0014153230004012585,
          "train_metric": 0.00017142348224297166,
          "validation_metric": 0.00014254615234676749,
          "lr_first": 0.0001164297184564739,
          "lr_last": 0.00010973604712160967,
          "seconds": 4.0964135760004865
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0017210777914442588,
          "train_loss": 0.0017038426594808698,
          "validation_loss": 0.0014483584091067314,
          "train_metric": 0.0001716047991067171,
          "validation_metric": 0.0001458733604522422,
          "lr_first": 0.00010929726238668399,
          "lr_last": 0.00010282775584616308,
          "seconds": 4.118653506000555
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.001735910260322271,
          "train_loss": 0.0017003265675157309,
          "validation_loss": 0.0014060076791793108,
          "train_metric": 0.00017125066369771957,
          "validation_metric": 0.000141607946716249,
          "lr_first": 0.00010240395279588673,
          "lr_last": 9.615971956635015e-05,
          "seconds": 4.144117179999739
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0017273115554417018,
          "train_loss": 0.0016962122172117233,
          "validation_loss": 0.0014173132367432117,
          "train_metric": 0.00017083626880776137,
          "validation_metric": 0.00014274660497903824,
          "lr_first": 9.575097081719392e-05,
          "lr_last": 8.973308081585969e-05,
          "seconds": 4.303178976000709
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0017302994056080934,
          "train_loss": 0.0016945230308920145,
          "validation_loss": 0.001419585314579308,
          "train_metric": 0.00017066614236682653,
          "validation_metric": 0.00014297544839791954,
          "lr_first": 8.933945640481766e-05,
          "lr_last": 8.354894076615733e-05,
          "seconds": 4.324971944999561
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0017173524556710618,
          "train_loss": 0.0016932226717472076,
          "validation_loss": 0.0014155051903799176,
          "train_metric": 0.00017053518968168646,
          "validation_metric": 0.00014256450231187046,
          "lr_first": 8.317050813874592e-05,
          "lr_last": 7.760835903780457e-05,
          "seconds": 4.347822911999174
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.001752232277794974,
          "train_loss": 0.0017011265736073256,
          "validation_loss": 0.0013876663288101554,
          "train_metric": 0.0001713312231004238,
          "validation_metric": 0.0001397606683894992,
          "lr_first": 7.724518303650655e-05,
          "lr_last": 7.191235351889875e-05,
          "seconds": 4.375568763000047
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0017264605830860091,
          "train_loss": 0.0017198225250467658,
          "validation_loss": 0.0014872044557705522,
          "train_metric": 0.0001732141972752288,
          "validation_metric": 0.0001497857883805409,
          "lr_first": 7.156449637205192e-05,
          "lr_last": 6.646190019066193e-05,
          "seconds": 4.400092994999795
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.001726244965539081,
          "train_loss": 0.0017054794589057565,
          "validation_loss": 0.001386648160405457,
          "train_metric": 0.00017176964320242405,
          "validation_metric": 0.00013965812104288489,
          "lr_first": 6.61294215017977e-05,
          "lr_last": 6.125793296021323e-05,
          "seconds": 4.426101041999573
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.001725583329971414,
          "train_loss": 0.0016894637374207377,
          "validation_loss": 0.0014227009378373623,
          "train_metric": 0.00017015659250319004,
          "validation_metric": 0.00014328921679407358,
          "lr_first": 6.094088969784328e-05,
          "lr_last": 5.630134350054723e-05,
          "seconds": 4.450855786999455
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0017050061505869962,
          "train_loss": 0.001693856786005199,
          "validation_loss": 0.0013906573876738548,
          "train_metric": 0.00017059905803762376,
          "validation_metric": 0.0001400619075866416,
          "lr_first": 5.599978998840269e-05,
          "lr_last": 5.1592981097751505e-05,
          "seconds": 4.467645016000461
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.0017244382906937972,
          "train_loss": 0.0016881163464859128,
          "validation_loss": 0.0014229011721909046,
          "train_metric": 0.0001700208813417703,
          "validation_metric": 0.00014330940030049533,
          "lr_first": 5.130696900547388e-05,
          "lr_last": 4.7133652505484666e-05,
          "seconds": 4.4841192759995465
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.001698677777312696,
          "train_loss": 0.0016873283311724663,
          "validation_loss": 0.0014182263985276222,
          "train_metric": 0.00016994152974803,
          "validation_metric": 0.00014283857308328152,
          "lr_first": 4.686323083977387e-05,
          "lr_last": 4.292412180674455e-05,
          "seconds": 4.500468401000035
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0017130220257968176,
          "train_loss": 0.0016864616191014647,
          "validation_loss": 0.0014024532865732908,
          "train_metric": 0.00016985424736049026,
          "validation_metric": 0.00014124995504971594,
          "lr_first": 4.266933690296076e-05,
          "lr_last": 3.896511028294589e-05,
          "seconds": 4.517492155999207
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.001694643699011067,
          "train_loss": 0.0016850957181304693,
          "validation_loss": 0.0014144243905320764,
          "train_metric": 0.00016971665900200605,
          "validation_metric": 0.00014245565398596227,
          "lr_first": 3.872600579717075e-05,
          "lr_last": 3.52572962903319e-05,
          "seconds": 4.534402490999128
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0017144511730293743,
          "train_loss": 0.0016846212092787027,
          "validation_loss": 0.0014063591370359063,
          "train_metric": 0.0001696688705123961,
          "validation_metric": 0.00014164333697408438,
          "lr_first": 3.503391319188826e-05,
          "lr_last": 3.18013151437425e-05,
          "seconds": 4.550917164000566
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.001717766323054093,
          "train_loss": 0.0016886790981516242,
          "validation_loss": 0.0014354977756738663,
          "train_metric": 0.00017007760470733047,
          "validation_metric": 0.00014457807992585003,
          "lr_first": 3.159369170817401e-05,
          "lr_last": 2.8597759007755454e-05,
          "seconds": 4.567877171999498
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.0016862289139680797,
          "train_loss": 0.001684109796769917,
          "validation_loss": 0.001399948145262897,
          "train_metric": 0.0001696173712844029,
          "validation_metric": 0.0001409976539434865,
          "lr_first": 2.8405930810269434e-05,
          "lr_last": 2.5647176795223176e-05,
          "seconds": 4.584533888000806
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0016943014816206414,
          "train_loss": 0.0016855801222845912,
          "validation_loss": 0.0013946460094302893,
          "train_metric": 0.000169765466125682,
          "validation_metric": 0.0001404636277584359,
          "lr_first": 2.5471176704593375e-05,
          "lr_last": 2.295007407321785e-05,
          "seconds": 4.601850217999527
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.001685774075667723,
          "train_loss": 0.0016825860366225243,
          "validation_loss": 0.0014053300255909562,
          "train_metric": 0.00016946389223448932,
          "validation_metric": 0.00014153969823382795,
          "lr_first": 2.2789932246153667e-05,
          "lr_last": 2.0506912976406118e-05,
          "seconds": 4.618750829000419
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0016880971743375994,
          "train_loss": 0.001682372996583581,
          "validation_loss": 0.0014099112013354897,
          "train_metric": 0.00016944242815952748,
          "validation_metric": 0.000142001110361889,
          "lr_first": 2.036265685238534e-05,
          "lr_last": 1.831811212786458e-05,
          "seconds": 4.635838494999916
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.001688984233624069,
          "train_loss": 0.001682990463450551,
          "validation_loss": 0.001418429659679532,
          "train_metric": 0.00016950462304521352,
          "validation_metric": 0.0001428590330760926,
          "lr_first": 1.818976642443138e-05,
          "lr_last": 1.6384046567350722e-05,
          "seconds": 4.652664833000017
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.0016902648385439534,
          "train_loss": 0.0016816845163702965,
          "validation_loss": 0.0014074768405407667,
          "train_metric": 0.00016937310283537954,
          "validation_metric": 0.00014175592514220625,
          "lr_first": 1.6271633275880907e-05,
          "lr_last": 1.4705047687041804e-05,
          "seconds": 4.669799658999182
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.0016837418352224631,
          "train_loss": 0.0016818827716633677,
          "validation_loss": 0.0013991653686389327,
          "train_metric": 0.0001693930826149881,
          "validation_metric": 0.00014091882621869445,
          "lr_first": 1.4608586068973758e-05,
          "lr_last": 1.3281403174752126e-05,
          "seconds": 4.68657051000082
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.0016957414882199373,
          "train_loss": 0.0016814791597425938,
          "validation_loss": 0.001411357196047902,
          "train_metric": 0.00016935242456384003,
          "validation_metric": 0.00014214673137757927,
          "lr_first": 1.3200909758287839e-05,
          "lr_last": 1.2113356964640166e-05,
          "seconds": 4.703327310999157
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.0016829219475766877,
          "train_loss": 0.0016809891676530242,
          "validation_loss": 0.0014064990682527423,
          "train_metric": 0.00016930307901930064,
          "validation_metric": 0.0001416574523318559,
          "lr_first": 1.2048845541912142e-05,
          "lr_last": 1.120110919541105e-05,
          "seconds": 4.720882255000106
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.0016836058257467812,
          "train_loss": 0.0016807870706543326,
          "validation_loss": 0.0014066994190216064,
          "train_metric": 0.00016928272088989615,
          "validation_metric": 0.0001416776212863624,
          "lr_first": 1.1152590820119275e-05,
          "lr_last": 1.0544816176023438e-05,
          "seconds": 4.738028518000647
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.001683117741777096,
          "train_loss": 0.0016806942876428366,
          "validation_loss": 0.001402638852596283,
          "train_metric": 0.00016927336400840431,
          "validation_metric": 0.0001412686542607844,
          "lr_first": 1.0512299161541902e-05,
          "lr_last": 1.0144590358907933e-05,
          "seconds": 4.7551966959999845
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.0016887263918761164,
          "train_loss": 0.0016805214108899236,
          "validation_loss": 0.0014021301176398993,
          "train_metric": 0.0001692559599177912,
          "validation_metric": 0.00014121740241535008,
          "lr_first": 1.0128080276859537e-05,
          "lr_last": 1.0000500320697657e-05,
          "seconds": 4.772298493000562
        }
      ],
      "best": {
        "epoch": 220,
        "updates": 3520
      },
      "best_validation_metric": 0.00013965812104288489,
      "config": {
        "label": "Cosine decay",
        "batch": 64,
        "lr": 0.003,
        "schedule": "cosine",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.011817703712772836
    },
    "1298f0867b9f_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.578583518974483,
          "train_loss": 0.24066148698329926,
          "validation_loss": 0.15968552231788635,
          "train_metric": 0.02423853985965252,
          "validation_metric": 0.016082942485809326,
          "lr_first": 0.003,
          "lr_last": 0.002999887429249521,
          "seconds": 0.029688462998819887
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.20282045425847173,
          "train_loss": 0.14825640618801117,
          "validation_loss": 0.12120170891284943,
          "train_metric": 0.014931841753423214,
          "validation_metric": 0.012206990271806717,
          "lr_first": 0.0029998719197231404,
          "lr_last": 0.0029995192175544397,
          "seconds": 0.04501414699916495
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13219130085781217,
          "train_loss": 0.11560314148664474,
          "validation_loss": 0.09988448768854141,
          "train_metric": 0.011643127538263798,
          "validation_metric": 0.010060000233352184,
          "lr_first": 0.0029994877008384588,
          "lr_last": 0.002998894927685237,
          "seconds": 0.058394669998961035
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10902860760688782,
          "train_loss": 0.09609463065862656,
          "validation_loss": 0.08106571435928345,
          "train_metric": 0.00967829953879118,
          "validation_metric": 0.008164641447365284,
          "lr_first": 0.002998847409179882,
          "lr_last": 0.002998014666610766,
          "seconds": 0.07205752799927723
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08854833571240306,
          "train_loss": 0.07872656732797623,
          "validation_loss": 0.06392683833837509,
          "train_metric": 0.007929051294922829,
          "validation_metric": 0.006438478361815214,
          "lr_first": 0.0029979511544580887,
          "lr_last": 0.0029968785851592315,
          "seconds": 0.08434489399951417
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.07389985385816544,
          "train_loss": 0.06745704263448715,
          "validation_loss": 0.055806588381528854,
          "train_metric": 0.006794026121497154,
          "validation_metric": 0.005620635114610195,
          "lr_first": 0.002996799090241712,
          "lr_last": 0.002995486877992343,
          "seconds": 0.09645522099890513
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06628730800002813,
          "train_loss": 0.06211603432893753,
          "validation_loss": 0.051012832671403885,
          "train_metric": 0.0062560997903347015,
          "validation_metric": 0.005137825850397348,
          "lr_first": 0.0029953914139310266,
          "lr_last": 0.0029938397835719568,
          "seconds": 0.1087630989986792
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.0606337949866429,
          "train_loss": 0.058541543781757355,
          "validation_loss": 0.04750123620033264,
          "train_metric": 0.005896090064197779,
          "validation_metric": 0.004784150514751673,
          "lr_first": 0.002993728366724119,
          "lr_last": 0.002991937584119225,
          "seconds": 0.12118398599886859
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05869300907943398,
          "train_loss": 0.05522774159908295,
          "validation_loss": 0.0470036156475544,
          "train_metric": 0.0055623361840844154,
          "validation_metric": 0.004734032321721315,
          "lr_first": 0.0029918102335755687,
          "lr_last": 0.0029897806055662304,
          "seconds": 0.13365441799942346
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05454411043319851,
          "train_loss": 0.050176169723272324,
          "validation_loss": 0.04220312088727951,
          "train_metric": 0.00505355978384614,
          "validation_metric": 0.004250544123351574,
          "lr_first": 0.0029896373431476165,
          "lr_last": 0.0029873692175001417,
          "seconds": 0.14592130899836775
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.049018904101103544,
          "train_loss": 0.04562188312411308,
          "validation_loss": 0.03885018080472946,
          "train_metric": 0.00459486898034811,
          "validation_metric": 0.003912847954779863,
          "lr_first": 0.002987210067753847,
          "lr_last": 0.0029847038330998916,
          "seconds": 0.15843853499973193
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.045386731857433915,
          "train_loss": 0.04298870638012886,
          "validation_loss": 0.036047253757715225,
          "train_metric": 0.004329665098339319,
          "validation_metric": 0.0036305473186075687,
          "lr_first": 0.0029845288232954064,
          "lr_last": 0.0029817849090653776,
          "seconds": 0.170835493998311
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.041057373862713575,
          "train_loss": 0.03856996074318886,
          "validation_loss": 0.032930053770542145,
          "train_metric": 0.003884625155478716,
          "validation_metric": 0.003316594287753105,
          "lr_first": 0.0029815940691897305,
          "lr_last": 0.002978612945539204,
          "seconds": 0.18345199199939088
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03568117297254503,
          "train_loss": 0.03248097747564316,
          "validation_loss": 0.027904314920306206,
          "train_metric": 0.0032713646069169044,
          "validation_metric": 0.0028104204684495926,
          "lr_first": 0.002978406308291826,
          "lr_last": 0.0029751884860209923,
          "seconds": 0.1958531829986896
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.034274683508556336,
          "train_loss": 0.030090607702732086,
          "validation_loss": 0.024579912424087524,
          "train_metric": 0.0030306153930723667,
          "validation_metric": 0.0024755988270044327,
          "lr_first": 0.0029749660868081114,
          "lr_last": 0.0029715121172742523,
          "seconds": 0.2085402829998202
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.029022879898548126,
          "train_loss": 0.028205307200551033,
          "validation_loss": 0.0248072799295187,
          "train_metric": 0.0028407350182533264,
          "validation_metric": 0.002498498186469078,
          "lr_first": 0.0029712739942028285,
          "lr_last": 0.0029675844692258427,
          "seconds": 0.2203908319988841
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.02795686520403251,
          "train_loss": 0.02664227783679962,
          "validation_loss": 0.022779712453484535,
          "train_metric": 0.0026833126321434975,
          "validation_metric": 0.0022942894138395786,
          "lr_first": 0.002967330663097039,
          "lr_last": 0.002963406214858036,
          "seconds": 0.2329154919989378
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.02989691775292158,
          "train_loss": 0.02862733043730259,
          "validation_loss": 0.02330332249403,
          "train_metric": 0.002883239882066846,
          "validation_metric": 0.0023470253217965364,
          "lr_first": 0.0029631367691602254,
          "lr_last": 0.00295897807009321,
          "seconds": 0.2451141559995449
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.025351357588078827,
          "train_loss": 0.02274473011493683,
          "validation_loss": 0.01956823654472828,
          "train_metric": 0.002290765754878521,
          "validation_metric": 0.001970841083675623,
          "lr_first": 0.002958693030994526,
          "lr_last": 0.002954300793671172,
          "seconds": 0.2581743169994297
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02053322858409956,
          "train_loss": 0.019641339778900146,
          "validation_loss": 0.016889983788132668,
          "train_metric": 0.0019782038871198893,
          "validation_metric": 0.001701097353361547,
          "lr_first": 0.002954000210011597,
          "lr_last": 0.00294937518701916,
          "seconds": 0.2702974469993933
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.021511508501134813,
          "train_loss": 0.019828028976917267,
          "validation_loss": 0.017100099474191666,
          "train_metric": 0.001997006591409445,
          "validation_metric": 0.001722259446978569,
          "lr_first": 0.002949059110302156,
          "lr_last": 0.0029442020941145156,
          "seconds": 0.2856471569994028
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.019893265853170305,
          "train_loss": 0.018307020887732506,
          "validation_loss": 0.015154242515563965,
          "train_metric": 0.0018438161350786686,
          "validation_metric": 0.0015262797242030501,
          "lr_first": 0.0029438705784982015,
          "lr_last": 0.002938782401340078,
          "seconds": 0.404207111998403
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.018516995303798467,
          "train_loss": 0.01883808895945549,
          "validation_loss": 0.016379857435822487,
          "train_metric": 0.001897303038276732,
          "validation_metric": 0.0016497190808877349,
          "lr_first": 0.0029384355036279485,
          "lr_last": 0.0029331170373323025,
          "seconds": 0.41655364999860467
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019479445705655962,
          "train_loss": 0.018992234021425247,
          "validation_loss": 0.015719939023256302,
          "train_metric": 0.0019128283020108938,
          "validation_metric": 0.001583254779689014,
          "lr_first": 0.002932754816963494,
          "lr_last": 0.002927206972822146,
          "seconds": 0.428403679999974
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.02608981484081596,
          "train_loss": 0.021976452320814133,
          "validation_loss": 0.017568442970514297,
          "train_metric": 0.002213387517258525,
          "validation_metric": 0.0017694291891530156,
          "lr_first": 0.002926829491861254,
          "lr_last": 0.002921053220468738,
          "seconds": 0.44042049599920574
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.020194696786347777,
          "train_loss": 0.020052773877978325,
          "validation_loss": 0.01591920293867588,
          "train_metric": 0.0020196419209241867,
          "validation_metric": 0.0016033239662647247,
          "lr_first": 0.0029206605435951826,
          "lr_last": 0.0029146568346858654,
          "seconds": 0.45472889999837207
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01837464710115455,
          "train_loss": 0.01548013836145401,
          "validation_loss": 0.013297850266098976,
          "train_metric": 0.001559102674946189,
          "validation_metric": 0.0013393107801675797,
          "lr_first": 0.002914249029182806,
          "lr_last": 0.002908018911461302,
          "seconds": 0.46728615699976217
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01576888223644346,
          "train_loss": 0.014512110501527786,
          "validation_loss": 0.01254633255302906,
          "train_metric": 0.0014616067055612803,
          "validation_metric": 0.0012636205647140741,
          "lr_first": 0.002907596047204113,
          "lr_last": 0.0029011405881690204,
          "seconds": 0.47974648799936404
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.018055916269076988,
          "train_loss": 0.015955207869410515,
          "validation_loss": 0.013088359497487545,
          "train_metric": 0.00160695007070899,
          "validation_metric": 0.0013182115508243442,
          "lr_first": 0.002900702737613315,
          "lr_last": 0.0028940230433743067,
          "seconds": 0.49349622699992324
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01718138565775007,
          "train_loss": 0.025777781382203102,
          "validation_loss": 0.022368567064404488,
          "train_metric": 0.0025962437503039837,
          "validation_metric": 0.0022528800182044506,
          "lr_first": 0.002893570281543524,
          "lr_last": 0.002886667496631822,
          "seconds": 0.505396872998972
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.018736071011517197,
          "train_loss": 0.014844364486634731,
          "validation_loss": 0.012379889376461506,
          "train_metric": 0.0014950702898204327,
          "validation_metric": 0.0012468572240322828,
          "lr_first": 0.0028861999011043707,
          "lr_last": 0.002879075208276634,
          "seconds": 0.5180554909984494
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.015360942314146087,
          "train_loss": 0.013344631530344486,
          "validation_loss": 0.011334178037941456,
          "train_metric": 0.0013440223410725594,
          "validation_metric": 0.001141536864452064,
          "lr_first": 0.0028785928591726004,
          "lr_last": 0.0028712474792082655,
          "seconds": 0.5305837939995399
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.017086882464354858,
          "train_loss": 0.01850997470319271,
          "validation_loss": 0.016079531982541084,
          "train_metric": 0.0018642565701156855,
          "validation_metric": 0.0016194714698940516,
          "lr_first": 0.0028707504591756854,
          "lr_last": 0.0028631856506677973,
          "seconds": 0.5427749699993001
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01807147398358211,
          "train_loss": 0.015497952699661255,
          "validation_loss": 0.013539548963308334,
          "train_metric": 0.0015608968678861856,
          "validation_metric": 0.0013636538060382009,
          "lr_first": 0.0028626740448684956,
          "lr_last": 0.0028548911040080457,
          "seconds": 0.5549006659985025
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014992505603004247,
          "train_loss": 0.015684911981225014,
          "validation_loss": 0.012609146535396576,
          "train_metric": 0.0015797269297763705,
          "validation_metric": 0.001269947038963437,
          "lr_first": 0.002854365000103041,
          "lr_last": 0.002846365260456879,
          "seconds": 0.5669501479987957
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.013836863086908124,
          "train_loss": 0.01204605307430029,
          "validation_loss": 0.010162314400076866,
          "train_metric": 0.0012132342671975493,
          "validation_metric": 0.0010235110530629754,
          "lr_first": 0.0028458247485913675,
          "lr_last": 0.002837609580873701,
          "seconds": 0.5791141519985104
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013854811782948673,
          "train_loss": 0.013904976658523083,
          "validation_loss": 0.012152324430644512,
          "train_metric": 0.0014004583936184645,
          "validation_metric": 0.0012239376083016396,
          "lr_first": 0.0028370547536616087,
          "lr_last": 0.002828625565499134,
          "seconds": 0.5918377419984608
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.013593538576969877,
          "train_loss": 0.012505619786679745,
          "validation_loss": 0.010740327648818493,
          "train_metric": 0.0012595203006640077,
          "validation_metric": 0.0010817265138030052,
          "lr_first": 0.002828056518007247,
          "lr_last": 0.00281941475369796,
          "seconds": 0.6036625689994253
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.014365337876370177,
          "train_loss": 0.012418260797858238,
          "validation_loss": 0.010178769007325172,
          "train_metric": 0.0012507217470556498,
          "validation_metric": 0.0010251684579998255,
          "lr_first": 0.0028188315834296366,
          "lr_last": 0.0028099787236953665,
          "seconds": 0.6157962129982479
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.01172087315353565,
          "train_loss": 0.011007699184119701,
          "validation_loss": 0.009332867339253426,
          "train_metric": 0.0011086552403867245,
          "validation_metric": 0.0009399721166118979,
          "lr_first": 0.00280938153057383,
          "lr_last": 0.0028003190923065156,
          "seconds": 0.6280387349997909
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.01173132541589439,
          "train_loss": 0.011560788378119469,
          "validation_loss": 0.009075101464986801,
          "train_metric": 0.0011643602047115564,
          "validation_metric": 0.0009140109177678823,
          "lr_first": 0.0027997079786577336,
          "lr_last": 0.00279043751465952,
          "seconds": 0.6406542400000035
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.0130742161127273,
          "train_loss": 0.011134405620396137,
          "validation_loss": 0.008632044307887554,
          "train_metric": 0.0011214164551347494,
          "validation_metric": 0.0008693879935890436,
          "lr_first": 0.00278981258519467,
          "lr_last": 0.0027803356839118383,
          "seconds": 0.6556824209983461
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011922438730834983,
          "train_loss": 0.017909348011016846,
          "validation_loss": 0.015437968075275421,
          "train_metric": 0.0018037636764347553,
          "validation_metric": 0.001554855378344655,
          "lr_first": 0.0027796970457093662,
          "lr_last": 0.002770015330960163,
          "seconds": 0.6677493309998681
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014054103477974422,
          "train_loss": 0.010764545761048794,
          "validation_loss": 0.009049308486282825,
          "train_metric": 0.0010841656476259232,
          "validation_metric": 0.000911413284484297,
          "lr_first": 0.0027693630934474377,
          "lr_last": 0.0027594782241438446,
          "seconds": 0.679418623998572
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.011824608445749618,
          "train_loss": 0.010339589789509773,
          "validation_loss": 0.00873603206127882,
          "train_metric": 0.0010413656709715724,
          "validation_metric": 0.0008798611233942211,
          "lr_first": 0.0027588124990784056,
          "lr_last": 0.0027487261689418912,
          "seconds": 0.6918858559984074
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.01065640544402413,
          "train_loss": 0.009946569800376892,
          "validation_loss": 0.008324718102812767,
          "train_metric": 0.001001782133243978,
          "validation_metric": 0.0008384351385757327,
          "lr_first": 0.0027480470703923022,
          "lr_last": 0.002737761007663611,
          "seconds": 0.7040580099983345
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.010274254746036604,
          "train_loss": 0.009527480229735374,
          "validation_loss": 0.008359488099813461,
          "train_metric": 0.000959572964347899,
          "validation_metric": 0.0008419370278716087,
          "lr_first": 0.0027370686519899118,
          "lr_last": 0.0027265846191329374,
          "seconds": 0.7162982069985446
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010652008699253201,
          "train_loss": 0.011080658994615078,
          "validation_loss": 0.009136163629591465,
          "train_metric": 0.0011160033755004406,
          "validation_metric": 0.0009201609645970166,
          "lr_first": 0.0027258791249667085,
          "lr_last": 0.0027151989183665085,
          "seconds": 0.7279986769990501
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.011638111449428834,
          "train_loss": 0.009088186547160149,
          "validation_loss": 0.007848307490348816,
          "train_metric": 0.0009153288556262851,
          "validation_metric": 0.0007904528174549341,
          "lr_first": 0.002714480406590544,
          "lr_last": 0.0027036058562455397,
          "seconds": 0.7403734979998262
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009375215304316953,
          "train_loss": 0.00907275639474392,
          "validation_loss": 0.007699868641793728,
          "train_metric": 0.0009137748274952173,
          "validation_metric": 0.0007755024125799537,
          "lr_first": 0.0027028744499731326,
          "lr_last": 0.002691807419181544,
          "seconds": 0.7529581809994852
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.00890913448529318,
          "train_loss": 0.009123887866735458,
          "validation_loss": 0.008331665769219398,
          "train_metric": 0.0009189246338792145,
          "validation_metric": 0.0008391349110752344,
          "lr_first": 0.002691063243735394,
          "lr_last": 0.0026798056287759726,
          "seconds": 0.7652670339994074
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.009579642879543826,
          "train_loss": 0.008897438645362854,
          "validation_loss": 0.007818013429641724,
          "train_metric": 0.0008961174171417952,
          "validation_metric": 0.0007874016300775111,
          "lr_first": 0.0026790488116667116,
          "lr_last": 0.002667602541473826,
          "seconds": 0.7770808519999264
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.009568192253937013,
          "train_loss": 0.008747031912207603,
          "validation_loss": 0.007833478972315788,
          "train_metric": 0.000880969048012048,
          "validation_metric": 0.0007889593252912164,
          "lr_first": 0.0026668332123781697,
          "lr_last": 0.0026552002482112917,
          "seconds": 0.7891971669996565
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009333474954473786,
          "train_loss": 0.00865683052688837,
          "validation_loss": 0.007670341059565544,
          "train_metric": 0.0008718842873349786,
          "validation_metric": 0.0007725286413915455,
          "lr_first": 0.00265441853894982,
          "lr_last": 0.0026426008740574737,
          "seconds": 0.8015530479988229
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008793021785095334,
          "train_loss": 0.00882020965218544,
          "validation_loss": 0.007026957347989082,
          "train_metric": 0.0008883391274139285,
          "validation_metric": 0.0007077294867485762,
          "lr_first": 0.0026418069185720444,
          "lr_last": 0.0026298065778502713,
          "seconds": 0.8141570129992033
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.008371356321731582,
          "train_loss": 0.00817162822932005,
          "validation_loss": 0.006710435263812542,
          "train_metric": 0.0008230165694840252,
          "validation_metric": 0.000675850547850132,
          "lr_first": 0.002629000512181069,
          "lr_last": 0.0026168195518264732,
          "seconds": 0.8265913789982733
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.0082578812725842,
          "train_loss": 0.007806184235960245,
          "validation_loss": 0.006388547830283642,
          "train_metric": 0.0007862102356739342,
          "validation_metric": 0.0006434312090277672,
          "lr_first": 0.002616001514088701,
          "lr_last": 0.0026036420212461322,
          "seconds": 0.8390280589992472
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007955854816827923,
          "train_loss": 0.007695130072534084,
          "validation_loss": 0.006465144921094179,
          "train_metric": 0.0007750253425911069,
          "validation_metric": 0.0006511457613669336,
          "lr_first": 0.0026028121516063455,
          "lr_last": 0.0025902762440112768,
          "seconds": 0.9375129299987748
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.008732302972930484,
          "train_loss": 0.008196339011192322,
          "validation_loss": 0.00741692166775465,
          "train_metric": 0.0008255051798187196,
          "validation_metric": 0.0007470052223652601,
          "lr_first": 0.0025894346846633653,
          "lr_last": 0.002576724510279026,
          "seconds": 0.9501138279993029
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007350800820859149,
          "train_loss": 0.008459622971713543,
          "validation_loss": 0.006609209813177586,
          "train_metric": 0.000852022145409137,
          "validation_metric": 0.0006656554178334773,
          "lr_first": 0.0025758714054198526,
          "lr_last": 0.0025629891420691935,
          "seconds": 0.9620081269986258
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.008415525066084228,
          "train_loss": 0.007970809005200863,
          "validation_loss": 0.005994886159896851,
          "train_metric": 0.0008027907460927963,
          "validation_metric": 0.000603782944381237,
          "lr_first": 0.0025621246378738837,
          "lr_last": 0.0025490724928664103,
          "seconds": 0.9746179259982455
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007401110080536455,
          "train_loss": 0.007078518159687519,
          "validation_loss": 0.006062920205295086,
          "train_metric": 0.0007129223085939884,
          "validation_metric": 0.0006106350920163095,
          "lr_first": 0.0025481967374633083,
          "lr_last": 0.002534976947216873,
          "seconds": 0.9867596569984016
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.007194512159912847,
          "train_loss": 0.007242962718009949,
          "validation_loss": 0.00612006476148963,
          "train_metric": 0.0007294846582226455,
          "validation_metric": 0.0006163904909044504,
          "lr_first": 0.002534090090662157,
          "lr_last": 0.002520704920319766,
          "seconds": 0.9988059059996885
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007597973104566336,
          "train_loss": 0.006959767546504736,
          "validation_loss": 0.005865865387022495,
          "train_metric": 0.0007009622640907764,
          "validation_metric": 0.0005907884915359318,
          "lr_first": 0.0025198071145717425,
          "lr_last": 0.0025062588576134272,
          "seconds": 1.011315128998831
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.00794612780737225,
          "train_loss": 0.007703699637204409,
          "validation_loss": 0.0069849309511482716,
          "train_metric": 0.0007758884457871318,
          "validation_metric": 0.0007034966256469488,
          "lr_first": 0.0025053502565064887,
          "lr_last": 0.00249164123435633,
          "seconds": 1.0235275389986782
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.007668186561204493,
          "train_loss": 0.006981272716075182,
          "validation_loss": 0.005637022200971842,
          "train_metric": 0.0007031282875686884,
          "validation_metric": 0.0005677403532899916,
          "lr_first": 0.0024907219935745985,
          "lr_last": 0.0024768545552029684,
          "seconds": 1.0363242819985317
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008293298044009134,
          "train_loss": 0.008070108480751514,
          "validation_loss": 0.006033163517713547,
          "train_metric": 0.0008127918117679656,
          "validation_metric": 0.0006076381541788578,
          "lr_first": 0.0024759248322536198,
          "lr_last": 0.002461901353774695,
          "seconds": 1.05749956299951
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.007782445303746499,
          "train_loss": 0.007456752005964518,
          "validation_loss": 0.006243608891963959,
          "train_metric": 0.0007510168361477554,
          "validation_metric": 0.0006288335425779223,
          "lr_first": 0.0024609613079609708,
          "lr_last": 0.0024467841922255927,
          "seconds": 1.0755310589993314
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.007975669504958205,
          "train_loss": 0.008135255426168442,
          "validation_loss": 0.007355880923569202,
          "train_metric": 0.0008193531539291143,
          "validation_metric": 0.0007408574456349015,
          "lr_first": 0.0024458339846195024,
          "lr_last": 0.0024315056608034706,
          "seconds": 1.087778436998633
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006601853274332825,
          "train_loss": 0.006051723379641771,
          "validation_loss": 0.005073388107120991,
          "train_metric": 0.0006095074350014329,
          "validation_metric": 0.000510973040945828,
          "lr_first": 0.002430545454218197,
          "lr_last": 0.0024160683774060343,
          "seconds": 1.1003014409998286
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006387718196492642,
          "train_loss": 0.006772509776055813,
          "validation_loss": 0.005298936739563942,
          "train_metric": 0.0006821023998782039,
          "validation_metric": 0.0005336894537322223,
          "lr_first": 0.0024150983363680325,
          "lr_last": 0.0024004749871323217,
          "seconds": 1.1124753259991849
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006964447704376653,
          "train_loss": 0.006012565456330776,
          "validation_loss": 0.0047522736713290215,
          "train_metric": 0.0006055635749362409,
          "validation_metric": 0.00047863161307759583,
          "lr_first": 0.0023994952778531296,
          "lr_last": 0.0023847281618294826,
          "seconds": 1.124558043999059
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006433619506424293,
          "train_loss": 0.006046713329851627,
          "validation_loss": 0.00508439727127552,
          "train_metric": 0.0006090027745813131,
          "validation_metric": 0.0005120818968862295,
          "lr_first": 0.002383738952177242,
          "lr_last": 0.002368830599634967,
          "seconds": 1.1364001469992218
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.006567460623045918,
          "train_loss": 0.006074830889701843,
          "validation_loss": 0.005066614132374525,
          "train_metric": 0.0006118346937000751,
          "validation_metric": 0.0005102907889522612,
          "lr_first": 0.002367832059105659,
          "lr_last": 0.002352785024514216,
          "seconds": 1.1481284059991594
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.008052033794228919,
          "train_loss": 0.0065531861037015915,
          "validation_loss": 0.005666631739586592,
          "train_metric": 0.0006600128253921866,
          "validation_metric": 0.0005707223317585886,
          "lr_first": 0.00235177732420262,
          "lr_last": 0.0023365941857939245,
          "seconds": 1.1600038199994742
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.007817708828952163,
          "train_loss": 0.005711706820875406,
          "validation_loss": 0.004860119894146919,
          "train_metric": 0.0005752621800638735,
          "validation_metric": 0.0004894935409538448,
          "lr_first": 0.002335577498364301,
          "lr_last": 0.0023202608576909545,
          "seconds": 1.171852225999828
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.00609383374103345,
          "train_loss": 0.005599590949714184,
          "validation_loss": 0.004856469575315714,
          "train_metric": 0.0005639703595079482,
          "validation_metric": 0.0004891257849521935,
          "lr_first": 0.002319235357347461,
          "lr_last": 0.002303787838836996,
          "seconds": 1.1835632529982831
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006514605433039833,
          "train_loss": 0.005544018000364304,
          "validation_loss": 0.004577497951686382,
          "train_metric": 0.00055837316904217,
          "validation_metric": 0.0004610288597177714,
          "lr_first": 0.002302753701293836,
          "lr_last": 0.002287177951799022,
          "seconds": 1.1957483779988252
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006859620858449489,
          "train_loss": 0.011131823994219303,
          "validation_loss": 0.007415438070893288,
          "train_metric": 0.0011211564997211099,
          "validation_metric": 0.0007468556868843734,
          "lr_first": 0.002286135354250341,
          "lr_last": 0.002270434042595671,
          "seconds": 1.2073474859989801
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.007116153254173696,
          "train_loss": 0.005776462610810995,
          "validation_loss": 0.004323720466345549,
          "train_metric": 0.0005817841156385839,
          "validation_metric": 0.0004354692646302283,
          "lr_first": 0.0022693831636851907,
          "lr_last": 0.002253558980209585,
          "seconds": 1.2195714409990615
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006123938495875336,
          "train_loss": 0.005148808937519789,
          "validation_loss": 0.004176880232989788,
          "train_metric": 0.0005185691989026964,
          "validation_metric": 0.00042068015318363905,
          "lr_first": 0.0022524999999999976,
          "lr_last": 0.0022365556560958325,
          "seconds": 1.2319222969999828
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.006797972426284105,
          "train_loss": 0.005312827881425619,
          "validation_loss": 0.004268969874829054,
          "train_metric": 0.0005350885912775993,
          "validation_metric": 0.0004299549909774214,
          "lr_first": 0.0022354887560379452,
          "lr_last": 0.002219426983686468,
          "seconds": 1.243795352998859
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.005883376172278076,
          "train_loss": 0.005393367260694504,
          "validation_loss": 0.0040777288377285,
          "train_metric": 0.0005432001780718565,
          "validation_metric": 0.00041069387225434184,
          "lr_first": 0.002218352346588113,
          "lr_last": 0.0022021758978913284,
          "seconds": 1.2562651109983562
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.006698817363940179,
          "train_loss": 0.005786505062133074,
          "validation_loss": 0.004949836526066065,
          "train_metric": 0.0005827955901622772,
          "validation_metric": 0.0004985294071957469,
          "lr_first": 0.0022010937078860405,
          "lr_last": 0.002184805354595156,
          "seconds": 1.27074689499932
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0052287028229329735,
          "train_loss": 0.004827258642762899,
          "validation_loss": 0.00377790629863739,
          "train_metric": 0.00048618376604281366,
          "validation_metric": 0.00038049687282182276,
          "lr_first": 0.0021837157971106207,
          "lr_last": 0.0021673183301511214,
          "seconds": 1.2832996929992078
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005545696571061853,
          "train_loss": 0.005202919244766235,
          "validation_loss": 0.004083799198269844,
          "train_metric": 0.0005240190075710416,
          "validation_metric": 0.00041130531462840736,
          "lr_first": 0.002166221591877406,
          "lr_last": 0.0021497178208708407,
          "seconds": 1.2951472729982925
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005636722096824087,
          "train_loss": 0.004761110991239548,
          "validation_loss": 0.004129934124648571,
          "train_metric": 0.00047952166642062366,
          "validation_metric": 0.00041595182847231627,
          "lr_first": 0.0021486140897284006,
          "lr_last": 0.002132006842510969,
          "seconds": 1.3087053009985539
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.005018608499085531,
          "train_loss": 0.005352854263037443,
          "validation_loss": 0.004804822616279125,
          "train_metric": 0.0005391198792494833,
          "validation_metric": 0.0004839241737499833,
          "lr_first": 0.002130896307618455,
          "lr_last": 0.0021141884297564742,
          "seconds": 1.3204901929984771
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005835162926814519,
          "train_loss": 0.005830409936606884,
          "validation_loss": 0.004125657025724649,
          "train_metric": 0.0005872175679542124,
          "validation_metric": 0.00041552100447006524,
          "lr_first": 0.0021130712813983224,
          "lr_last": 0.0020962656357006496,
          "seconds": 1.332296103999397
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0055340087856166065,
          "train_loss": 0.0056274388916790485,
          "validation_loss": 0.0042901961132884026,
          "train_metric": 0.0005667750374414027,
          "validation_metric": 0.0004320928710512817,
          "lr_first": 0.00209514206529448,
          "lr_last": 0.002078241531321991,
          "seconds": 1.3442544679983257
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.005360699011362158,
          "train_loss": 0.004871584940701723,
          "validation_loss": 0.0038027495611459017,
          "train_metric": 0.0004906480899080634,
          "validation_metric": 0.00038299901643767953,
          "lr_first": 0.0020771117313858087,
          "lr_last": 0.0020601192049579968,
          "seconds": 1.3561538709982415
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.00518205020489404,
          "train_loss": 0.004385503940284252,
          "validation_loss": 0.0036300448700785637,
          "train_metric": 0.00044169192551635206,
          "validation_metric": 0.00036560482112690806,
          "lr_first": 0.0020589833690772045,
          "lr_last": 0.0020419017617759943,
          "seconds": 1.3687544639997213
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004735865761176683,
          "train_loss": 0.004402937367558479,
          "validation_loss": 0.0033835191279649734,
          "train_metric": 0.00044344773050397635,
          "validation_metric": 0.00034077567397616804,
          "lr_first": 0.0020407600845702236,
          "lr_last": 0.0020235923232410895,
          "seconds": 1.3811714609983028
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004726823091914412,
          "train_loss": 0.00466553121805191,
          "validation_loss": 0.0035985996946692467,
          "train_metric": 0.00046989513793960214,
          "validation_metric": 0.0003624378005042672,
          "lr_first": 0.0020224450003308516,
          "lr_last": 0.002005194026581318,
          "seconds": 1.4759738209995703
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004633390679373406,
          "train_loss": 0.005355499219149351,
          "validation_loss": 0.0038022235967218876,
          "train_metric": 0.0005393861792981625,
          "validation_metric": 0.0003829460474662483,
          "lr_first": 0.0020040412545544867,
          "lr_last": 0.001986710024250099,
          "seconds": 1.4882528859998274
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.005203351298405323,
          "train_loss": 0.00442679924890399,
          "validation_loss": 0.0037807754706591368,
          "train_metric": 0.00044585097930394113,
          "validation_metric": 0.00038078587385825813,
          "lr_first": 0.0019855520006282273,
          "lr_last": 0.0019681434833860745,
          "seconds": 1.5085035359988979
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0041687689736136235,
          "train_loss": 0.004045642912387848,
          "validation_loss": 0.0034499026369303465,
          "train_metric": 0.00040746235754340887,
          "validation_metric": 0.0003474615514278412,
          "lr_first": 0.0019669804065905482,
          "lr_last": 0.0019494975852704384,
          "seconds": 1.530937201998313
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004888013267191127,
          "train_loss": 0.004711226560175419,
          "validation_loss": 0.00365860085003078,
          "train_metric": 0.0004744974721688777,
          "validation_metric": 0.00036848089075647295,
          "lr_first": 0.0019483296545884795,
          "lr_last": 0.0019307755247818476,
          "seconds": 1.553299008999602
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.006970222399104387,
          "train_loss": 0.006660035811364651,
          "validation_loss": 0.005101521033793688,
          "train_metric": 0.0006707743741571903,
          "validation_metric": 0.0005138065316714346,
          "lr_first": 0.0019296029403323655,
          "lr_last": 0.0019119805098489838,
          "seconds": 1.5732450159994187
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.00467652077350067,
          "train_loss": 0.009409744292497635,
          "validation_loss": 0.006609454285353422,
          "train_metric": 0.0009477148996666074,
          "validation_metric": 0.0006656800396740437,
          "lr_first": 0.001910803472548287,
          "lr_last": 0.0018931157609008988,
          "seconds": 1.5937045439986832
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.006908910087076947,
          "train_loss": 0.004324018023908138,
          "validation_loss": 0.0034331430215388536,
          "train_metric": 0.0004354992706794292,
          "validation_metric": 0.000345773616572842,
          "lr_first": 0.0018919344724282693,
          "lr_last": 0.0018741845103152086,
          "seconds": 1.6132900059983513
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.005428287033282686,
          "train_loss": 0.005857693962752819,
          "validation_loss": 0.005413766019046307,
          "train_metric": 0.0005899654934182763,
          "validation_metric": 0.0005452546756714582,
          "lr_first": 0.0018729991730783457,
          "lr_last": 0.0018551900018642434,
          "seconds": 1.6352124939985515
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.005263964441837743,
          "train_loss": 0.0058121331967413425,
          "validation_loss": 0.0052050030790269375,
          "train_metric": 0.000585376750677824,
          "validation_metric": 0.0005242288461886346,
          "lr_first": 0.00185400081896458,
          "lr_last": 0.0018361354901592413,
          "seconds": 1.6701656189998175
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004712892128736712,
          "train_loss": 0.004062305670231581,
          "validation_loss": 0.0036010961048305035,
          "train_metric": 0.0004091406299266964,
          "validation_metric": 0.0003626891993917525,
          "lr_first": 0.0018349426653571461,
          "lr_last": 0.001817024240092691,
          "seconds": 1.6899098869998852
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003880092735926155,
          "train_loss": 0.004174948204308748,
          "validation_loss": 0.002979012206196785,
          "train_metric": 0.00042048547766171396,
          "validation_metric": 0.0003000352007802576,
          "lr_first": 0.0018158279777725514,
          "lr_last": 0.0017978595262789054,
          "seconds": 1.7136296889984806
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004210228275042027,
          "train_loss": 0.004950962029397488,
          "validation_loss": 0.003584453370422125,
          "train_metric": 0.0004986427375115454,
          "validation_metric": 0.00036101299338042736,
          "lr_first": 0.0017966600314141131,
          "lr_last": 0.001778644632492937,
          "seconds": 1.734690467999826
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0043309659013175406,
          "train_loss": 0.003817247226834297,
          "validation_loss": 0.0032812966965138912,
          "train_metric": 0.00038445915561169386,
          "validation_metric": 0.0003304801939520985,
          "lr_first": 0.0017774421106107614,
          "lr_last": 0.0017593828511079182,
          "seconds": 1.7560467359999166
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.004863278329139575,
          "train_loss": 0.00393854733556509,
          "validation_loss": 0.00290325446985662,
          "train_metric": 0.0003966760414186865,
          "validation_metric": 0.00029240522417239845,
          "lr_first": 0.0017581775082542933,
          "lr_last": 0.0017400774825309276,
          "seconds": 1.7820768139990832
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.003714798978762701,
          "train_loss": 0.003968254197388887,
          "validation_loss": 0.002941090613603592,
          "train_metric": 0.0003996680025011301,
          "validation_metric": 0.00029621587600558996,
          "lr_first": 0.0017388695252351471,
          "lr_last": 0.001720731834637488,
          "seconds": 1.8079870429992297
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.004007485375041142,
          "train_loss": 0.0073453448712825775,
          "validation_loss": 0.005211316514760256,
          "train_metric": 0.000739796319976449,
          "validation_metric": 0.0005248647648841143,
          "lr_first": 0.001719521469876815,
          "lr_last": 0.0017013492222047732,
          "seconds": 1.834839496999848
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.005396296503022313,
          "train_loss": 0.00638968963176012,
          "validation_loss": 0.005135911051183939,
          "train_metric": 0.0006435461109504104,
          "validation_metric": 0.0005172701785340905,
          "lr_first": 0.0017001366573689788,
          "lr_last": 0.0016819329663436428,
          "seconds": 1.861912043999837
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.005496314508491196,
          "train_loss": 0.004452133551239967,
          "validation_loss": 0.003504369640722871,
          "train_metric": 0.00044840259943157434,
          "validation_metric": 0.00035294730332680047,
          "lr_first": 0.0016807184091994691,
          "lr_last": 0.0016624863939295827,
          "seconds": 1.8903088769984606
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0047314640250988305,
          "train_loss": 0.003909853752702475,
          "validation_loss": 0.003353268140926957,
          "train_metric": 0.0003937861474696547,
          "validation_metric": 0.00033772888127714396,
          "lr_first": 0.001661270052585144,
          "lr_last": 0.0016430128370326644,
          "seconds": 1.917754050999065
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004992199668777175,
          "train_loss": 0.004428062587976456,
          "validation_loss": 0.003959628287702799,
          "train_metric": 0.00044597825035452843,
          "validation_metric": 0.0003987992531619966,
          "lr_first": 0.0016417949199017884,
          "lr_last": 0.0016235156323466118,
          "seconds": 1.9415978099987115
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.003974256651417818,
          "train_loss": 0.003643760224804282,
          "validation_loss": 0.002727070590481162,
          "train_metric": 0.000366986176231876,
          "validation_metric": 0.000274660560535267,
          "lr_first": 0.001622296348113129,
          "lr_last": 0.0016039981206170785,
          "seconds": 1.9717954479983746
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0035712123280973174,
          "train_loss": 0.003706492716446519,
          "validation_loss": 0.0032724591437727213,
          "train_metric": 0.0003733043558895588,
          "validation_metric": 0.0003295901115052402,
          "lr_first": 0.0016027776781990647,
          "lr_last": 0.0015844636460692202,
          "seconds": 1.9971863689988822
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003393658160348423,
          "train_loss": 0.0032329305540770292,
          "validation_loss": 0.002596660517156124,
          "train_metric": 0.00032560888212174177,
          "validation_metric": 0.00026152614736929536,
          "lr_first": 0.0015832422545832014,
          "lr_last": 0.0015649155558346884,
          "seconds": 2.025362367998241
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0033488128501630854,
          "train_loss": 0.003315175883471966,
          "validation_loss": 0.0027518272399902344,
          "train_metric": 0.0003338923561386764,
          "validation_metric": 0.0002771540021058172,
          "lr_first": 0.0015636934245598084,
          "lr_last": 0.0015453571993781102,
          "seconds": 2.050391373999446
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0036947680200682953,
          "train_loss": 0.0032753453124314547,
          "validation_loss": 0.002627411624416709,
          "train_metric": 0.0003298807714600116,
          "validation_metric": 0.0002646232896950096,
          "lr_first": 0.0015441345377202708,
          "lr_last": 0.0015257919279231778,
          "seconds": 2.074640050999733
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0035068489814875647,
          "train_loss": 0.003253509756177664,
          "validation_loss": 0.0025772496592253447,
          "train_metric": 0.000327681569615379,
          "validation_metric": 0.0002595711557660252,
          "lr_first": 0.0015245689453791606,
          "lr_last": 0.0015062230938784333,
          "seconds": 2.094616269998369
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.003516692180710379,
          "train_loss": 0.00355373602360487,
          "validation_loss": 0.0030627914238721132,
          "train_metric": 0.00035791925620287657,
          "validation_metric": 0.00030847315792925656,
          "lr_first": 0.001505000000000001,
          "lr_last": 0.0014866540502628492,
          "seconds": 2.114623674999166
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.003644773649284616,
          "train_loss": 0.0031912808772176504,
          "validation_loss": 0.002469472587108612,
          "train_metric": 0.00032141414703801274,
          "validation_metric": 0.0002487162419129163,
          "lr_first": 0.0014854310546208404,
          "lr_last": 0.0014670881501313076,
          "seconds": 2.140133846000026
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0034532147692516446,
          "train_loss": 0.0031286755111068487,
          "validation_loss": 0.0025513777509331703,
          "train_metric": 0.00031510871485807,
          "validation_metric": 0.00025696540251374245,
          "lr_first": 0.0014658654622797308,
          "lr_last": 0.0014475287460000708,
          "seconds": 2.161146798998743
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.003424627248023171,
          "train_loss": 0.0033916491083800793,
          "validation_loss": 0.002623584121465683,
          "train_metric": 0.0003415944520384073,
          "validation_metric": 0.00026423780946061015,
          "lr_first": 0.0014463065754401938,
          "lr_last": 0.001427979189272343,
          "seconds": 2.18193652699847
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0036477104949881323,
          "train_loss": 0.0031721247360110283,
          "validation_loss": 0.002402833430096507,
          "train_metric": 0.00031948479590937495,
          "validation_metric": 0.00024200456391554326,
          "lr_first": 0.0014267577454168008,
          "lr_last": 0.0014084428296640297,
          "seconds": 2.2145392449983774
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.003215632910723798,
          "train_loss": 0.0031586105469614267,
          "validation_loss": 0.002524451119825244,
          "train_metric": 0.0003181236970704049,
          "validation_metric": 0.0002542534493841231,
          "lr_first": 0.001407222321800938,
          "lr_last": 0.0013889230146297755,
          "seconds": 2.2528194970000186
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0033229738000954967,
          "train_loss": 0.003072583582252264,
          "validation_loss": 0.002369454363361001,
          "train_metric": 0.0003094593994319439,
          "validation_metric": 0.00023864279501140118,
          "lr_first": 0.0013877036518868736,
          "lr_last": 0.0013694230887893984,
          "seconds": 2.286699462998513
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.003138683969154954,
          "train_loss": 0.003117284970358014,
          "validation_loss": 0.0023705242201685905,
          "train_metric": 0.00031396152917295694,
          "validation_metric": 0.00023875049373600632,
          "lr_first": 0.001368205080098215,
          "lr_last": 0.0013499463933548063,
          "seconds": 2.312183547999666
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0034827872732421383,
          "train_loss": 0.0034129652194678783,
          "validation_loss": 0.002960867015644908,
          "train_metric": 0.00034374138340353966,
          "validation_metric": 0.00029820765485055745,
          "lr_first": 0.00134872994741486,
          "lr_last": 0.0013304962655574931,
          "seconds": 2.3420957069993165
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0040791874343995005,
          "train_loss": 0.0032201341819018126,
          "validation_loss": 0.00268260994926095,
          "train_metric": 0.00032432013540528715,
          "validation_metric": 0.00027018264518119395,
          "lr_first": 0.0013292815908005342,
          "lr_last": 0.0013110760380767297,
          "seconds": 2.5079703189985594
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0035355062063899823,
          "train_loss": 0.0029585000593215227,
          "validation_loss": 0.0024394821375608444,
          "train_metric": 0.00029796926537528634,
          "validation_metric": 0.00024569572997279465,
          "lr_first": 0.0013098633426310243,
          "lr_last": 0.0012916890384685182,
          "seconds": 2.542612960000042
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.003006873906997498,
          "train_loss": 0.004101315978914499,
          "validation_loss": 0.0035115338396281004,
          "train_metric": 0.0004130695015192032,
          "validation_metric": 0.00035366881638765335,
          "lr_first": 0.0012904785301231876,
          "lr_last": 0.0012723385885954383,
          "seconds": 2.5612486069985607
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.003119390174106229,
          "train_loss": 0.0030347222927957773,
          "validation_loss": 0.0022969890851527452,
          "train_metric": 0.0003056460991501808,
          "validation_metric": 0.00023134431103244424,
          "lr_first": 0.001271130474764856,
          "lr_last": 0.0012530280040574575,
          "seconds": 2.5882947339996463
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.00306600996555062,
          "train_loss": 0.003289906308054924,
          "validation_loss": 0.0028566867113113403,
          "train_metric": 0.00033134734258055687,
          "validation_metric": 0.00028771505458280444,
          "lr_first": 0.0012518224917457093,
          "lr_last": 0.001233760593623827,
          "seconds": 2.6055066659992008
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0033721491126925685,
          "train_loss": 0.002949667861685157,
          "validation_loss": 0.002454799832776189,
          "train_metric": 0.00029707973590120673,
          "validation_metric": 0.00024723843671381474,
          "lr_first": 0.001232557889389241,
          "lr_last": 0.0012145396586661337,
          "seconds": 2.6225459199995385
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004319005856814329,
          "train_loss": 0.00596557417884469,
          "validation_loss": 0.004319369327276945,
          "train_metric": 0.0006008309428580105,
          "validation_metric": 0.00043503110646270216,
          "lr_first": 0.0012133399685858898,
          "lr_last": 0.0011953684925926351,
          "seconds": 2.6393679499997234
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003817568263912108,
          "train_loss": 0.002915665740147233,
          "validation_loss": 0.002378517296165228,
          "train_metric": 0.00029365517548285425,
          "validation_metric": 0.00023955553479027003,
          "lr_first": 0.0011941720222274513,
          "lr_last": 0.0011762503802839421,
          "seconds": 2.6562612279994937
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0030874836338625755,
          "train_loss": 0.0032590851187705994,
          "validation_loss": 0.0028447555378079414,
          "train_metric": 0.0003282431571278721,
          "validation_metric": 0.0002865133574232459,
          "lr_first": 0.001175057334642857,
          "lr_last": 0.001157188597530181,
          "seconds": 2.6734334039992973
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.003251107300457079,
          "train_loss": 0.003506576409563422,
          "validation_loss": 0.003007665043696761,
          "train_metric": 0.00035316956927999854,
          "validation_metric": 0.000302920991089195,
          "lr_first": 0.0011559991810354228,
          "lr_last": 0.0011381864104696943,
          "seconds": 2.6907985449997796
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.002968494958622614,
          "train_loss": 0.0029028260614722967,
          "validation_loss": 0.002140981610864401,
          "train_metric": 0.00029236203408800066,
          "validation_metric": 0.000215631807805039,
          "lr_first": 0.0011370008269216568,
          "lr_last": 0.0011192470750294122,
          "seconds": 2.708692379999775
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.002939648875326384,
          "train_loss": 0.002760025206953287,
          "validation_loss": 0.002156417351216078,
          "train_metric": 0.00027797967777587473,
          "validation_metric": 0.00021718646166846156,
          "lr_first": 0.0011180655275717327,
          "lr_last": 0.0011003738363669622,
          "seconds": 2.725635766999403
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.003048853417567443,
          "train_loss": 0.0029054409824311733,
          "validation_loss": 0.002410494489595294,
          "train_metric": 0.0002926254237536341,
          "validation_metric": 0.00024277616466861218,
          "lr_first": 0.001099196527451715,
          "lr_last": 0.0010815699283146308,
          "seconds": 2.741981297998791
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0027066909096902236,
          "train_loss": 0.0042363908141851425,
          "validation_loss": 0.002941545331850648,
          "train_metric": 0.00042667367961257696,
          "validation_metric": 0.0002962617145385593,
          "lr_first": 0.0010803970596676372,
          "lr_last": 0.0010628385728252573,
          "seconds": 2.75902340599896
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0029628510237671435,
          "train_loss": 0.003495758632197976,
          "validation_loss": 0.0030215184669941664,
          "train_metric": 0.0003520800091791898,
          "validation_metric": 0.00030431628692895174,
          "lr_first": 0.0010616703454115225,
          "lr_last": 0.0010441829794201776,
          "seconds": 2.7761078529983934
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.003861302826408064,
          "train_loss": 0.0030400841496884823,
          "validation_loss": 0.0024583975318819284,
          "train_metric": 0.0003061861789319664,
          "validation_metric": 0.000247600837610662,
          "lr_first": 0.001043019593409455,
          "lr_last": 0.0010256063446392812,
          "seconds": 2.792974112999218
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.004394677271193359,
          "train_loss": 0.002955424366518855,
          "validation_loss": 0.0023470912128686905,
          "train_metric": 0.0002976595133077353,
          "validation_metric": 0.00023639039136469364,
          "lr_first": 0.0010244479993717752,
          "lr_last": 0.001007111851493306,
          "seconds": 2.809781114998259
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0033407504233764485,
          "train_loss": 0.0030487587209790945,
          "validation_loss": 0.002125398488715291,
          "train_metric": 0.0003070598468184471,
          "validation_metric": 0.00021406234009191394,
          "lr_first": 0.0010059587454455136,
          "lr_last": 0.0009887026689184423,
          "seconds": 2.827311800998359
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0032240250875474885,
          "train_loss": 0.0031748032197356224,
          "validation_loss": 0.0023431945592164993,
          "train_metric": 0.00031975453021004796,
          "validation_metric": 0.0002359979844186455,
          "lr_first": 0.0009875549996691494,
          "lr_last": 0.0009703819512333538,
          "seconds": 2.8442439119990013
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0029311292128113564,
          "train_loss": 0.0028358709532767534,
          "validation_loss": 0.002442800672724843,
          "train_metric": 0.00028561853105202317,
          "validation_metric": 0.0002460299583617598,
          "lr_first": 0.0009692399154297775,
          "lr_last": 0.0009521528375986995,
          "seconds": 2.860965039999428
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0027739924917113967,
          "train_loss": 0.002583951922133565,
          "validation_loss": 0.00207475945353508,
          "train_metric": 0.00026024619000963867,
          "validation_metric": 0.0002089621702907607,
          "lr_first": 0.0009510166309227964,
          "lr_last": 0.0009340184514792582,
          "seconds": 2.878215957998691
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.002734052002779208,
          "train_loss": 0.0036991776432842016,
          "validation_loss": 0.0032592513598501682,
          "train_metric": 0.00037256762152537704,
          "validation_metric": 0.0003282598336227238,
          "lr_first": 0.0009328882686141923,
          "lr_last": 0.0009159819001087345,
          "seconds": 2.895030861998748
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0032116790644067805,
          "train_loss": 0.0029342237394303083,
          "validation_loss": 0.0024957177229225636,
          "train_metric": 0.0002955242816824466,
          "validation_metric": 0.00025135951000265777,
          "lr_first": 0.0009148579347055221,
          "lr_last": 0.0008980462739573539,
          "seconds": 2.9115737449992594
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0028395533518050797,
          "train_loss": 0.0029328581877052784,
          "validation_loss": 0.0025123944506049156,
          "train_metric": 0.000295386795187369,
          "validation_metric": 0.00025303917936980724,
          "lr_first": 0.0008969287186016802,
          "lr_last": 0.000880214646202324,
          "seconds": 2.92821063899828
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0027480675071274163,
          "train_loss": 0.002509523881599307,
          "validation_loss": 0.0019505320815369487,
          "train_metric": 0.00025275006191805005,
          "validation_metric": 0.00019645046268124133,
          "lr_first": 0.0008791036923815459,
          "lr_last": 0.0008624900722012644,
          "seconds": 2.9474545049997687
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0026612534493324347,
          "train_loss": 0.0025518920738250017,
          "validation_loss": 0.001977490494027734,
          "train_metric": 0.00025701720733195543,
          "validation_metric": 0.00019916558812838048,
          "lr_first": 0.0008613859102716003,
          "lr_last": 0.0008448755889686828,
          "seconds": 2.9909188199999335
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.002713095127546694,
          "train_loss": 0.0035474535543471575,
          "validation_loss": 0.0025723858270794153,
          "train_metric": 0.0003572865389287472,
          "validation_metric": 0.000259081250987947,
          "lr_first": 0.0008437784081225947,
          "lr_last": 0.000827374214655601,
          "seconds": 3.0371634119983355
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.00329310277265904,
          "train_loss": 0.0029917380306869745,
          "validation_loss": 0.0022594367619603872,
          "train_metric": 0.0003013169043697417,
          "validation_metric": 0.00022756218095310032,
          "lr_first": 0.0008262842028893788,
          "lr_last": 0.0008099889480324116,
          "seconds": 3.0809380629998486
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0034204627663712017,
          "train_loss": 0.003356536151841283,
          "validation_loss": 0.0024666786193847656,
          "train_metric": 0.00033805801649577916,
          "validation_metric": 0.00024843483697623014,
          "lr_first": 0.0008089062921139598,
          "lr_last": 0.0007927227679750494,
          "seconds": 3.122957546998805
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0030455340383923613,
          "train_loss": 0.0030492336954921484,
          "validation_loss": 0.002322802087292075,
          "train_metric": 0.00030710763530805707,
          "validation_metric": 0.0002339441271033138,
          "lr_first": 0.0007916476534118866,
          "lr_last": 0.0007755786329545818,
          "seconds": 3.140208838000035
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.003027313003258314,
          "train_loss": 0.0025738277472555637,
          "validation_loss": 0.002113108290359378,
          "train_metric": 0.0002592264791019261,
          "validation_metric": 0.00021282455418258905,
          "lr_first": 0.0007745112439620533,
          "lr_last": 0.0007585594805302867,
          "seconds": 3.1566978709997784
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0036789439764106646,
          "train_loss": 0.004639043938368559,
          "validation_loss": 0.0033806327264755964,
          "train_metric": 0.0004672275099437684,
          "validation_metric": 0.00034048492670990527,
          "lr_first": 0.0007575000000000017,
          "lr_last": 0.0007416682268463151,
          "seconds": 3.174248389999775
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0037848357533221133,
          "train_loss": 0.0035466451663523912,
          "validation_loss": 0.002658623270690441,
          "train_metric": 0.0003572051355149597,
          "validation_metric": 0.0002677667944226414,
          "lr_first": 0.0007406168363148086,
          "lr_last": 0.000724907766132032,
          "seconds": 3.191609146999326
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.002614121898659505,
          "train_loss": 0.003248188178986311,
          "validation_loss": 0.0028338392730802298,
          "train_metric": 0.0003271456225775182,
          "validation_metric": 0.0002854138729162514,
          "lr_first": 0.0007238646457496579,
          "lr_last": 0.0007082809702060954,
          "seconds": 3.2087794429990026
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0028665543322858866,
          "train_loss": 0.0025460051838308573,
          "validation_loss": 0.0020728763192892075,
          "train_metric": 0.0002564243332017213,
          "validation_metric": 0.0002087725151795894,
          "lr_first": 0.000707246298706162,
          "lr_last": 0.0006917906879843936,
          "seconds": 3.225626444998852
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.002743386634392664,
          "train_loss": 0.0025817351415753365,
          "validation_loss": 0.002095792442560196,
          "train_metric": 0.00026002287631854415,
          "validation_metric": 0.00021108052169438452,
          "lr_first": 0.0006907646426525363,
          "lr_last": 0.0006754397449918927,
          "seconds": 3.2423109829996974
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0026111022853001487,
          "train_loss": 0.002630198374390602,
          "validation_loss": 0.001998075982555747,
          "train_metric": 0.00026490396703593433,
          "validation_metric": 0.00020123888680245727,
          "lr_first": 0.0006744225016356963,
          "lr_last": 0.0006592309428784994,
          "seconds": 3.3571252899982937
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0024792956210148986,
          "train_loss": 0.002354980679228902,
          "validation_loss": 0.0019058197503909469,
          "train_metric": 0.00023718502779956907,
          "validation_metric": 0.00019194718333892524,
          "lr_first": 0.0006582226757973757,
          "lr_last": 0.0006431670589390128,
          "seconds": 3.3758501639986207
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0029319440545805264,
          "train_loss": 0.00298936921171844,
          "validation_loss": 0.0025472063571214676,
          "train_metric": 0.0003010783111676574,
          "validation_metric": 0.00025654525961726904,
          "lr_first": 0.0006421679408943365,
          "lr_last": 0.0006272508456372499,
          "seconds": 3.3932558950000384
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0027931310469284654,
          "train_loss": 0.0024570273235440254,
          "validation_loss": 0.0020072765182703733,
          "train_metric": 0.00024746276903897524,
          "validation_metric": 0.00020216553821228445,
          "lr_first": 0.0006262610478227541,
          "lr_last": 0.0006114850301344212,
          "seconds": 3.4106264259989985
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.002518303706892766,
          "train_loss": 0.002342221327126026,
          "validation_loss": 0.0019118062919005752,
          "train_metric": 0.00023589991906192154,
          "validation_metric": 0.00019255014194641262,
          "lr_first": 0.0006105047221468661,
          "lr_last": 0.0005958723138218511,
          "seconds": 3.427602379999371
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0025580985384294763,
          "train_loss": 0.0023953760974109173,
          "validation_loss": 0.001873932546004653,
          "train_metric": 0.0002412535104667768,
          "validation_metric": 0.0001887356338556856,
          "lr_first": 0.0005949016636319636,
          "lr_last": 0.0005804153718581059,
          "seconds": 3.44530077699892
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0024701295915292576,
          "train_loss": 0.0024515895638614893,
          "validation_loss": 0.001898866961710155,
          "train_metric": 0.0002469151222612709,
          "validation_metric": 0.00019124694517813623,
          "lr_first": 0.0005794545457817997,
          "lr_last": 0.0005651168527106182,
          "seconds": 3.4632670119990507
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0025085636079893447,
          "train_loss": 0.002292698947712779,
          "validation_loss": 0.001830752706155181,
          "train_metric": 0.0002309122501173988,
          "validation_metric": 0.0001843867066781968,
          "lr_first": 0.0005641660153804942,
          "lr_last": 0.0005499793777018868,
          "seconds": 3.4821196049997525
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.002420762219117023,
          "train_loss": 0.0026058123912662268,
          "validation_loss": 0.0022025678772479296,
          "train_metric": 0.00026244789478369057,
          "validation_metric": 0.00022183454711921513,
          "lr_first": 0.000549038692039027,
          "lr_last": 0.0005350055405603265,
          "seconds": 3.5007985419997567
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.002312557589903008,
          "train_loss": 0.002927312394604087,
          "validation_loss": 0.002200644463300705,
          "train_metric": 0.000294828147161752,
          "validation_metric": 0.0002216408756794408,
          "lr_first": 0.0005340751677463768,
          "lr_last": 0.0005201979069758457,
          "seconds": 3.5181091079994076
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0027822606461995747,
          "train_loss": 0.0025131353177130222,
          "validation_loss": 0.0021088402718305588,
          "train_metric": 0.0002531138015910983,
          "validation_metric": 0.00021239467605482787,
          "lr_first": 0.0005192780064253987,
          "lr_last": 0.0005055590141602302,
          "seconds": 3.535034820999499
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0026371885069238488,
          "train_loss": 0.0023743906058371067,
          "validation_loss": 0.001825625658966601,
          "train_metric": 0.00023913993209134787,
          "validation_metric": 0.00018387031741440296,
          "lr_first": 0.0005046497434935084,
          "lr_last": 0.0004910913704124016,
          "seconds": 3.5534591139985423
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0023845076939323917,
          "train_loss": 0.0022997099440544844,
          "validation_loss": 0.0018213592702522874,
          "train_metric": 0.00023161839635577053,
          "validation_metric": 0.00018344062846153975,
          "lr_first": 0.0004901928854282528,
          "lr_last": 0.0004767974546886371,
          "seconds": 3.5744833839999046
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0024717074302316178,
          "train_loss": 0.002476637251675129,
          "validation_loss": 0.0018530895467847586,
          "train_metric": 0.0002494378131814301,
          "validation_metric": 0.0001866363745648414,
          "lr_first": 0.0004759099093378383,
          "lr_last": 0.00046267971617781114,
          "seconds": 3.5933472259985137
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0027860086993314326,
          "train_loss": 0.002480548806488514,
          "validation_loss": 0.0018625942757353187,
          "train_metric": 0.0002498317917343229,
          "validation_metric": 0.0001875936723081395,
          "lr_first": 0.0004618032625366892,
          "lr_last": 0.00044874057388173796,
          "seconds": 3.6115508079983556
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0024158224168786546,
          "train_loss": 0.0023188265040516853,
          "validation_loss": 0.0018727859714999795,
          "train_metric": 0.00023354370205197483,
          "validation_metric": 0.00018862013530451804,
          "lr_first": 0.0004478753621261122,
          "lr_last": 0.00043498241620069174,
          "seconds": 3.6302337009983603
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.002359358437388437,
          "train_loss": 0.0022930495906621218,
          "validation_loss": 0.0018088356591761112,
          "train_metric": 0.00023094756761565804,
          "validation_metric": 0.00018217934120912105,
          "lr_first": 0.00043412859458014243,
          "lr_last": 0.0004214076005241641,
          "seconds": 3.6496949739994307
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0024179663851100486,
          "train_loss": 0.0023970913607627153,
          "validation_loss": 0.002001143991947174,
          "train_metric": 0.00024142631446011364,
          "validation_metric": 0.0002015479258261621,
          "lr_first": 0.0004205653153366304,
          "lr_last": 0.00040801845282693807,
          "seconds": 3.6685884459984663
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0025717349999467842,
          "train_loss": 0.002264058217406273,
          "validation_loss": 0.0018484595930203795,
          "train_metric": 0.00022802763851359487,
          "validation_metric": 0.00018617008754517883,
          "lr_first": 0.0004071878483936508,
          "lr_last": 0.0003948172672705435,
          "seconds": 3.6880846549993294
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0022906020858499687,
          "train_loss": 0.0022313471417874098,
          "validation_loss": 0.0017723499331623316,
          "train_metric": 0.0002247330849058926,
          "validation_metric": 0.00017850458971224725,
          "lr_first": 0.00039399848591129634,
          "lr_last": 0.0003818063058101653,
          "seconds": 3.707885408999573
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.00229546213085996,
          "train_loss": 0.0022379790898412466,
          "validation_loss": 0.0017922802362591028,
          "train_metric": 0.0002254010469187051,
          "validation_metric": 0.0001805118954507634,
          "lr_first": 0.0003809994878189294,
          "lr_last": 0.00036898779780706985,
          "seconds": 3.726706404999277
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.002360801870963769,
          "train_loss": 0.002218174748122692,
          "validation_loss": 0.001776658813469112,
          "train_metric": 0.0002234064304502681,
          "validation_metric": 0.00017893858603201807,
          "lr_first": 0.00036819308142795417,
          "lr_last": 0.000356363939646614,
          "seconds": 3.753142907999063
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.002485907603841042,
          "train_loss": 0.00222514639608562,
          "validation_loss": 0.00179762439802289,
          "train_metric": 0.00022410858946386725,
          "validation_metric": 0.000181050127139315,
          "lr_first": 0.0003555814610501785,
          "lr_last": 0.0003439368943619046,
          "seconds": 3.7763032259990723
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0023105881045921706,
          "train_loss": 0.002199680544435978,
          "validation_loss": 0.0017522149719297886,
          "train_metric": 0.00022154378530103713,
          "validation_metric": 0.0001764766639098525,
          "lr_first": 0.0003431667876218289,
          "lr_last": 0.0003317087912631771,
          "seconds": 3.7995259369999985
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.002224820786068449,
          "train_loss": 0.0022037914022803307,
          "validation_loss": 0.0017476413631811738,
          "train_metric": 0.00022195778728928417,
          "validation_metric": 0.00017601603758521378,
          "lr_first": 0.00033095118833328664,
          "lr_last": 0.00031968172557294704,
          "seconds": 3.825912204998531
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.002291422693815548,
          "train_loss": 0.0022541636135429144,
          "validation_loss": 0.001757005462422967,
          "train_metric": 0.00022703110880684108,
          "validation_metric": 0.00017695914721116424,
          "lr_first": 0.00031893675626460375,
          "lr_last": 0.00030785775806700536,
          "seconds": 3.851072427998588
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002329283430299256,
          "train_loss": 0.0022861564066261053,
          "validation_loss": 0.0018805369036272168,
          "train_metric": 0.0002302533102920279,
          "validation_metric": 0.00018940077279694378,
          "lr_first": 0.000307125550026865,
          "lr_last": 0.00029623891472131333,
          "seconds": 3.874739924998721
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.002335441044124309,
          "train_loss": 0.002353876829147339,
          "validation_loss": 0.0018122349865734577,
          "train_metric": 0.0002370738802710548,
          "validation_metric": 0.0001825216895667836,
          "lr_first": 0.0002955195934094543,
          "lr_last": 0.00028482718636486487,
          "seconds": 3.896686570999009
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0022961368777032476,
          "train_loss": 0.002239790977910161,
          "validation_loss": 0.0018313180189579725,
          "train_metric": 0.00022558355703949928,
          "validation_metric": 0.0001844436483224854,
          "lr_first": 0.0002841208750332906,
          "lr_last": 0.0002736245283385659,
          "seconds": 3.91924771199956
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0022147397394292057,
          "train_loss": 0.002264559268951416,
          "validation_loss": 0.001763493288308382,
          "train_metric": 0.00022807810455560684,
          "validation_metric": 0.00017761258641257882,
          "lr_first": 0.00027293134801008756,
          "lr_last": 0.00026263286016019817,
          "seconds": 3.9403579289992194
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.002270019431307446,
          "train_loss": 0.0022198213264346123,
          "validation_loss": 0.0018116467399522662,
          "train_metric": 0.00022357224952429533,
          "validation_metric": 0.00018246246327180415,
          "lr_first": 0.0002619529296076959,
          "lr_last": 0.0002518540651955178,
          "seconds": 3.962296019999485
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0022135970466479193,
          "train_loss": 0.002194121014326811,
          "validation_loss": 0.0017261960310861468,
          "train_metric": 0.00022098385670688003,
          "validation_metric": 0.00017385616956744343,
          "lr_first": 0.00025118750092159196,
          "lr_last": 0.0002412899903355538,
          "seconds": 3.989907488999961
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.002399703356786631,
          "train_loss": 0.0022429325617849827,
          "validation_loss": 0.0018411378841847181,
          "train_metric": 0.0002258999738842249,
          "validation_metric": 0.00018543264013715088,
          "lr_first": 0.00024063690655256104,
          "lr_last": 0.00023094244568014875,
          "seconds": 4.013400946998445
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.002273384314321447,
          "train_loss": 0.0021859407424926758,
          "validation_loss": 0.0017283034976571798,
          "train_metric": 0.00022015995637048036,
          "validation_metric": 0.00017406843835487962,
          "lr_first": 0.0002303029542906331,
          "lr_last": 0.00022081320422781038,
          "seconds": 4.040978132999953
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.002199912974901963,
          "train_loss": 0.00216089840978384,
          "validation_loss": 0.0017260786844417453,
          "train_metric": 0.00021763777476735413,
          "validation_metric": 0.0001738443097565323,
          "lr_first": 0.00022018741480532905,
          "lr_last": 0.00021090400157191562,
          "seconds": 4.063952216998587
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0021946009546809364,
          "train_loss": 0.0021554026752710342,
          "validation_loss": 0.0017206420889124274,
          "train_metric": 0.0002170842490158975,
          "validation_metric": 0.00017329676484223455,
          "lr_first": 0.00021029202134226492,
          "lr_last": 0.00020121653560332717,
          "seconds": 4.091141302998949
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0023049053124850616,
          "train_loss": 0.002228357130661607,
          "validation_loss": 0.0018316099885851145,
          "train_metric": 0.00022443196212407202,
          "validation_metric": 0.00018447305774316192,
          "lr_first": 0.00020061846942616898,
          "lr_last": 0.00019175246621946768,
          "seconds": 4.294875539999339
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0022133774400572293,
          "train_loss": 0.0021844953298568726,
          "validation_loss": 0.001714290352538228,
          "train_metric": 0.00022001436445862055,
          "validation_metric": 0.00017265706264879555,
          "lr_first": 0.0001911684165703626,
          "lr_last": 0.00018251341503990455,
          "seconds": 4.313270338998336
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0022122326845419593,
          "train_loss": 0.0021475418470799923,
          "validation_loss": 0.0017099075485020876,
          "train_metric": 0.00021629252296406776,
          "validation_metric": 0.0001722156157484278,
          "lr_first": 0.00018194348199275317,
          "lr_last": 0.00017350096512849466,
          "seconds": 4.3316137099991465
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.002206783519795863,
          "train_loss": 0.0021435795351862907,
          "validation_loss": 0.0017114487709477544,
          "train_metric": 0.00021589346579276025,
          "validation_metric": 0.0001723708410281688,
          "lr_first": 0.00017294524633839077,
          "lr_last": 0.00016471666072213429,
          "seconds": 4.349170718998721
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0021642813881044276,
          "train_loss": 0.0021471832878887653,
          "validation_loss": 0.0017022668616846204,
          "train_metric": 0.00021625644876621664,
          "validation_metric": 0.00017144606681540608,
          "lr_first": 0.00016417525140863167,
          "lr_last": 0.00015616200696616197,
          "seconds": 4.367266961999121
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0021748325398220913,
          "train_loss": 0.0021598103921860456,
          "validation_loss": 0.0017564867157489061,
          "train_metric": 0.00021752816974185407,
          "validation_metric": 0.00017690689128357917,
          "lr_first": 0.00015563499989695927,
          "lr_last": 0.00014783846965645865,
          "seconds": 4.385410099999717
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0021904300956521183,
          "train_loss": 0.0021475667599588633,
          "validation_loss": 0.001693905214779079,
          "train_metric": 0.00021629504044540226,
          "validation_metric": 0.00017060391837731004,
          "lr_first": 0.00014732595513150429,
          "lr_last": 0.0001397474749882931,
          "seconds": 4.403216000999237
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.002151660439267289,
          "train_loss": 0.002133498899638653,
          "validation_loss": 0.0017005120171234012,
          "train_metric": 0.0002148781786672771,
          "validation_metric": 0.00017126937746070325,
          "lr_first": 0.0001392495408243121,
          "lr_last": 0.00013189040931194934,
          "seconds": 4.420706248998613
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.00221471816439589,
          "train_loss": 0.002158173592761159,
          "validation_loss": 0.0017519978573545814,
          "train_metric": 0.00021736336930189282,
          "validation_metric": 0.0001764548069331795,
          "lr_first": 0.00013140714082739743,
          "lr_last": 0.00012426861889518266,
          "seconds": 4.437572591999924
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.002315778943739133,
          "train_loss": 0.00216308212839067,
          "validation_loss": 0.0016967228148132563,
          "train_metric": 0.00021785769786220044,
          "validation_metric": 0.00017088769527617842,
          "lr_first": 0.00012380009889562699,
          "lr_last": 0.00011688340969254351,
          "seconds": 4.45409762700001
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0021654172087437473,
          "train_loss": 0.002140970667824149,
          "validation_loss": 0.001729026436805725,
          "train_metric": 0.00021563074551522732,
          "validation_metric": 0.0001741412270348519,
          "lr_first": 0.0001164297184564739,
          "lr_last": 0.00010973604712160967,
          "seconds": 4.47042425699874
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.00216410690882185,
          "train_loss": 0.002128692576661706,
          "validation_loss": 0.0017039309022948146,
          "train_metric": 0.00021439415286295116,
          "validation_metric": 0.00017161369032692164,
          "lr_first": 0.00010929726238668399,
          "lr_last": 0.00010282775584616308,
          "seconds": 4.512438891999409
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.00213375535531668,
          "train_loss": 0.0021395517978817225,
          "validation_loss": 0.0016892293933779001,
          "train_metric": 0.00021548784570768476,
          "validation_metric": 0.00017013300384860486,
          "lr_first": 0.00010240395279588673,
          "lr_last": 9.615971956635015e-05,
          "seconds": 4.542236454999511
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0021542968170251697,
          "train_loss": 0.002129088621586561,
          "validation_loss": 0.001688264892436564,
          "train_metric": 0.00021443399600684643,
          "validation_metric": 0.00017003584071062505,
          "lr_first": 9.575097081719392e-05,
          "lr_last": 8.973308081585969e-05,
          "seconds": 4.560384905998944
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.002131509580067359,
          "train_loss": 0.0021226091776043177,
          "validation_loss": 0.0016998491482809186,
          "train_metric": 0.00021378144447226077,
          "validation_metric": 0.00017120258416980505,
          "lr_first": 8.933945640481766e-05,
          "lr_last": 8.354894076615733e-05,
          "seconds": 4.577061991998562
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.002142364282917697,
          "train_loss": 0.002124006627127528,
          "validation_loss": 0.0017063344130292535,
          "train_metric": 0.00021392214694060385,
          "validation_metric": 0.00017185577598866075,
          "lr_first": 8.317050813874592e-05,
          "lr_last": 7.760835903780457e-05,
          "seconds": 4.593843865999588
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.002172871201764792,
          "train_loss": 0.0021281898953020573,
          "validation_loss": 0.0017168376361951232,
          "train_metric": 0.00021434352674987167,
          "validation_metric": 0.0001729135838104412,
          "lr_first": 7.724518303650655e-05,
          "lr_last": 7.191235351889875e-05,
          "seconds": 4.625380609999411
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0021394713839981705,
          "train_loss": 0.0021206638775765896,
          "validation_loss": 0.0017027802532538772,
          "train_metric": 0.0002135855029337108,
          "validation_metric": 0.0001714977843221277,
          "lr_first": 7.156449637205192e-05,
          "lr_last": 6.646190019066193e-05,
          "seconds": 4.642337689998385
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0021388508939708117,
          "train_loss": 0.0021309417206794024,
          "validation_loss": 0.0016832288820296526,
          "train_metric": 0.00021462066797539592,
          "validation_metric": 0.00016952863370534033,
          "lr_first": 6.61294215017977e-05,
          "lr_last": 6.125793296021323e-05,
          "seconds": 4.660697153998626
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.002191970153944567,
          "train_loss": 0.002119583310559392,
          "validation_loss": 0.0017055608332157135,
          "train_metric": 0.00021347666915971786,
          "validation_metric": 0.00017177785048261285,
          "lr_first": 6.094088969784328e-05,
          "lr_last": 5.630134350054723e-05,
          "seconds": 4.677895139999237
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.002129591284756316,
          "train_loss": 0.0021194263827055693,
          "validation_loss": 0.0016821075696498156,
          "train_metric": 0.0002134608366759494,
          "validation_metric": 0.00016941572539508343,
          "lr_first": 5.599978998840269e-05,
          "lr_last": 5.1592981097751505e-05,
          "seconds": 4.695412620998468
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.0021237230102997273,
          "train_loss": 0.00211505638435483,
          "validation_loss": 0.0016915512969717383,
          "train_metric": 0.00021302075765561312,
          "validation_metric": 0.00017036686767823994,
          "lr_first": 5.130696900547388e-05,
          "lr_last": 4.7133652505484666e-05,
          "seconds": 4.712881212999491
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.002134475034836214,
          "train_loss": 0.002114285947754979,
          "validation_loss": 0.0016951309517025948,
          "train_metric": 0.0002129431231878698,
          "validation_metric": 0.00017072736227419227,
          "lr_first": 4.686323083977387e-05,
          "lr_last": 4.292412180674455e-05,
          "seconds": 4.729540756999995
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0021255617029964924,
          "train_loss": 0.0021148892119526863,
          "validation_loss": 0.0016991904703900218,
          "train_metric": 0.00021300387743394822,
          "validation_metric": 0.00017113624198827893,
          "lr_first": 4.266933690296076e-05,
          "lr_last": 3.896511028294589e-05,
          "seconds": 4.746122409998861
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0021269383214530535,
          "train_loss": 0.002115969778969884,
          "validation_loss": 0.0016826189821586013,
          "train_metric": 0.00021311271120794117,
          "validation_metric": 0.0001694672100711614,
          "lr_first": 3.872600579717075e-05,
          "lr_last": 3.52572962903319e-05,
          "seconds": 4.762623500999325
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.002134740680048708,
          "train_loss": 0.002114133909344673,
          "validation_loss": 0.0017001949017867446,
          "train_metric": 0.00021292781457304955,
          "validation_metric": 0.0001712373923510313,
          "lr_first": 3.503391319188826e-05,
          "lr_last": 3.18013151437425e-05,
          "seconds": 4.779396923999229
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0021262478330754675,
          "train_loss": 0.0021140044555068016,
          "validation_loss": 0.001680998015217483,
          "train_metric": 0.00021291480516083539,
          "validation_metric": 0.00016930396668612957,
          "lr_first": 3.159369170817401e-05,
          "lr_last": 2.8597759007755454e-05,
          "seconds": 4.797612238999136
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.002128898817318259,
          "train_loss": 0.0021117599681019783,
          "validation_loss": 0.0016938382759690285,
          "train_metric": 0.00021268875570967793,
          "validation_metric": 0.0001705971808405593,
          "lr_first": 2.8405930810269434e-05,
          "lr_last": 2.5647176795223176e-05,
          "seconds": 4.815364694999516
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0021223886287771165,
          "train_loss": 0.0021118326112627983,
          "validation_loss": 0.0016840638127177954,
          "train_metric": 0.00021269606077112257,
          "validation_metric": 0.00016961274377536029,
          "lr_first": 2.5471176704593375e-05,
          "lr_last": 2.295007407321785e-05,
          "seconds": 4.832604836999963
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.0021174695903027896,
          "train_loss": 0.0021106519270688295,
          "validation_loss": 0.0016912964638322592,
          "train_metric": 0.00021257714251987636,
          "validation_metric": 0.00017034118354786187,
          "lr_first": 2.2789932246153667e-05,
          "lr_last": 2.0506912976406118e-05,
          "seconds": 4.850820017998558
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.002126044699252816,
          "train_loss": 0.0021114086266607046,
          "validation_loss": 0.0016962566878646612,
          "train_metric": 0.00021265336545184255,
          "validation_metric": 0.00017084076534956694,
          "lr_first": 2.036265685238534e-05,
          "lr_last": 1.831811212786458e-05,
          "seconds": 4.868252446998667
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.0021111187925271224,
          "train_loss": 0.0021099571604281664,
          "validation_loss": 0.001688731019385159,
          "train_metric": 0.0002125071914633736,
          "validation_metric": 0.000170082799741067,
          "lr_first": 1.818976642443138e-05,
          "lr_last": 1.6384046567350722e-05,
          "seconds": 4.885934405998341
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.002115772800607374,
          "train_loss": 0.0021097331773489714,
          "validation_loss": 0.0016894945874810219,
          "train_metric": 0.00021248463599476963,
          "validation_metric": 0.00017015969206113368,
          "lr_first": 1.6271633275880907e-05,
          "lr_last": 1.4705047687041804e-05,
          "seconds": 4.903002563998598
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.00211269169813022,
          "train_loss": 0.002109585329890251,
          "validation_loss": 0.0016883870121091604,
          "train_metric": 0.00021246970572974533,
          "validation_metric": 0.00017004815163090825,
          "lr_first": 1.4608586068973758e-05,
          "lr_last": 1.3281403174752126e-05,
          "seconds": 4.92016910899838
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.002111238587531261,
          "train_loss": 0.002109318971633911,
          "validation_loss": 0.001686337636783719,
          "train_metric": 0.00021244288654997945,
          "validation_metric": 0.0001698417472653091,
          "lr_first": 1.3200909758287839e-05,
          "lr_last": 1.2113356964640166e-05,
          "seconds": 4.937285191999763
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.002114139060722664,
          "train_loss": 0.0021097108256071806,
          "validation_loss": 0.0016838343581184745,
          "train_metric": 0.00021248235134407878,
          "validation_metric": 0.0001695896207820624,
          "lr_first": 1.2048845541912142e-05,
          "lr_last": 1.120110919541105e-05,
          "seconds": 4.954638330998932
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.002113729493430583,
          "train_loss": 0.002108914777636528,
          "validation_loss": 0.001687831012532115,
          "train_metric": 0.0002124021848430857,
          "validation_metric": 0.0001699921558611095,
          "lr_first": 1.1152590820119275e-05,
          "lr_last": 1.0544816176023438e-05,
          "seconds": 5.073793740999463
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.002110762223310303,
          "train_loss": 0.002108845626935363,
          "validation_loss": 0.0016853371635079384,
          "train_metric": 0.00021239521447569132,
          "validation_metric": 0.00016974098980426788,
          "lr_first": 1.0512299161541902e-05,
          "lr_last": 1.0144590358907933e-05,
          "seconds": 5.0914349790000415
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.0021135928473086096,
          "train_loss": 0.0021086763590574265,
          "validation_loss": 0.0016850439133122563,
          "train_metric": 0.00021237815963104367,
          "validation_metric": 0.00016971146396826953,
          "lr_first": 1.0128080276859537e-05,
          "lr_last": 1.0000500320697657e-05,
          "seconds": 5.110815218999051
        }
      ],
      "best": {
        "epoch": 228,
        "updates": 3648
      },
      "best_validation_metric": 0.00016930396668612957,
      "config": {
        "label": "Cosine decay",
        "batch": 64,
        "lr": 0.003,
        "schedule": "cosine",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.013011685774185048
    },
    "9967999fa0f8_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.1153237980324775,
          "train_loss": 0.5487765669822693,
          "validation_loss": 0.39034348726272583,
          "train_metric": 0.05527076870203018,
          "validation_metric": 0.03931397199630737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.025009290000525652
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.06490567768923938,
          "train_loss": 0.16639138758182526,
          "validation_loss": 0.12541139125823975,
          "train_metric": 0.016758333891630173,
          "validation_metric": 0.012630975805222988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03981699900032254
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.07953184633515775,
          "train_loss": 0.04384057596325874,
          "validation_loss": 0.0363566130399704,
          "train_metric": 0.004415462724864483,
          "validation_metric": 0.003661705646663904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05499706499904278
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.06982522364705801,
          "train_loss": 0.04148896783590317,
          "validation_loss": 0.03108304552733898,
          "train_metric": 0.004178617149591446,
          "validation_metric": 0.0031305705197155476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14301979700030643
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.05738291470333934,
          "train_loss": 0.058124788105487823,
          "validation_loss": 0.04193168133497238,
          "train_metric": 0.005854116752743721,
          "validation_metric": 0.004223205149173737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2204944360000809
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.051650868845172226,
          "train_loss": 0.028621364384889603,
          "validation_loss": 0.0234699584543705,
          "train_metric": 0.002882639179006219,
          "validation_metric": 0.002363807987421751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30101048200049263
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.04633639502571896,
          "train_loss": 0.0266619473695755,
          "validation_loss": 0.02151239477097988,
          "train_metric": 0.0026852935552597046,
          "validation_metric": 0.0021666495595127344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3791690029993333
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04296509822597727,
          "train_loss": 0.023144185543060303,
          "validation_loss": 0.01955394633114338,
          "train_metric": 0.0023309974931180477,
          "validation_metric": 0.0019694019574671984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47575214700009383
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04538457409944385,
          "train_loss": 0.050588663667440414,
          "validation_loss": 0.03838120773434639,
          "train_metric": 0.005095105152577162,
          "validation_metric": 0.0038656145334243774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4951839610002935
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05338413640856743,
          "train_loss": 0.03302048519253731,
          "validation_loss": 0.020742932334542274,
          "train_metric": 0.0033257019240409136,
          "validation_metric": 0.002089152345433831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5101956729995436
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.044860841357149184,
          "train_loss": 0.021277448162436485,
          "validation_loss": 0.016654128208756447,
          "train_metric": 0.002142986748367548,
          "validation_metric": 0.0016773426905274391,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5261370490006811
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.037308583210688084,
          "train_loss": 0.018689453601837158,
          "validation_loss": 0.015273457393050194,
          "train_metric": 0.0018823331920430064,
          "validation_metric": 0.0015382866840809584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.542283344999305
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03861196362413466,
          "train_loss": 0.02450603060424328,
          "validation_loss": 0.021142413839697838,
          "train_metric": 0.002468157559633255,
          "validation_metric": 0.002129386644810438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6455353860001196
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03728284849785268,
          "train_loss": 0.019097961485385895,
          "validation_loss": 0.01598045416176319,
          "train_metric": 0.001923476462252438,
          "validation_metric": 0.0016094926977530122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6600131929990312
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.053291109565179795,
          "train_loss": 0.015769461169838905,
          "validation_loss": 0.012095989659428596,
          "train_metric": 0.0015882423613220453,
          "validation_metric": 0.0012182638747617602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6749038950001705
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.035287767415866256,
          "train_loss": 0.014445947483181953,
          "validation_loss": 0.010992733761668205,
          "train_metric": 0.0014549429761245847,
          "validation_metric": 0.0011071480112150311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6899013040001591
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.0385514683148358,
          "train_loss": 0.03535957261919975,
          "validation_loss": 0.026684265583753586,
          "train_metric": 0.003561286721378565,
          "validation_metric": 0.0026875415351241827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7040173219993449
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.05318083695601672,
          "train_loss": 0.025026556104421616,
          "validation_loss": 0.018414661288261414,
          "train_metric": 0.002520583337172866,
          "validation_metric": 0.00185465719550848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7182495069992001
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.04567699210019782,
          "train_loss": 0.016098229214549065,
          "validation_loss": 0.013184873387217522,
          "train_metric": 0.0016213544877246022,
          "validation_metric": 0.0013279321137815714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7323472539992508
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.04040154907852411,
          "train_loss": 0.020340681076049805,
          "validation_loss": 0.01423314306885004,
          "train_metric": 0.002048639114946127,
          "validation_metric": 0.0014335098676383495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7465346389999468
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.03722978284349665,
          "train_loss": 0.013304292224347591,
          "validation_loss": 0.010463152080774307,
          "train_metric": 0.0013399594463407993,
          "validation_metric": 0.0010538104688748717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7612054149994947
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.028318025695625693,
          "train_loss": 0.022997111082077026,
          "validation_loss": 0.019014567136764526,
          "train_metric": 0.0023161848075687885,
          "validation_metric": 0.001915077562443912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7755673070005287
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.027374538010917604,
          "train_loss": 0.012311378493905067,
          "validation_loss": 0.009378034621477127,
          "train_metric": 0.0012399570550769567,
          "validation_metric": 0.0009445211035199463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7903649340005359
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.04241672868374735,
          "train_loss": 0.02188965678215027,
          "validation_loss": 0.018242280930280685,
          "train_metric": 0.0022046458907425404,
          "validation_metric": 0.0018372959457337856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8046833490006975
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.039926229597767815,
          "train_loss": 0.020444000139832497,
          "validation_loss": 0.014711299911141396,
          "train_metric": 0.0020590450149029493,
          "validation_metric": 0.0014816683251410723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8188317820004158
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.03899700823239982,
          "train_loss": 0.014024612493813038,
          "validation_loss": 0.011065952479839325,
          "train_metric": 0.0014125076122581959,
          "validation_metric": 0.0011145222233608365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8334366050003155
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.04584402992622927,
          "train_loss": 0.01520374696701765,
          "validation_loss": 0.01195753738284111,
          "train_metric": 0.0015312659088522196,
          "validation_metric": 0.0012043192982673645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8474850580005295
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.032777938642539084,
          "train_loss": 0.015821393579244614,
          "validation_loss": 0.012008905410766602,
          "train_metric": 0.001593472552485764,
          "validation_metric": 0.0012094930279999971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8612639189996116
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.03676655888557434,
          "train_loss": 0.014672689139842987,
          "validation_loss": 0.01138561312109232,
          "train_metric": 0.0014777793549001217,
          "validation_metric": 0.001146717113442719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8755619559997285
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.040524902986362576,
          "train_loss": 0.016085496172308922,
          "validation_loss": 0.012630436569452286,
          "train_metric": 0.0016200721729546785,
          "validation_metric": 0.0012720912927761674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.889838637000139
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.02926892414689064,
          "train_loss": 0.022279072552919388,
          "validation_loss": 0.016578340902924538,
          "train_metric": 0.002243866678327322,
          "validation_metric": 0.0016697098035365343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9038494060005178
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.041043392731808126,
          "train_loss": 0.01238026563078165,
          "validation_loss": 0.010411061346530914,
          "train_metric": 0.0012468950590118766,
          "validation_metric": 0.001048564095981419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9177891050003382
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.029928818461485207,
          "train_loss": 0.013320711441338062,
          "validation_loss": 0.00985651183873415,
          "train_metric": 0.0013416132424026728,
          "validation_metric": 0.0009927117498591542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9316291819995968
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.04110606200993061,
          "train_loss": 0.01899074763059616,
          "validation_loss": 0.01588241197168827,
          "train_metric": 0.001912678824737668,
          "validation_metric": 0.0015996183501556516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9456291639999108
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.026568253349978477,
          "train_loss": 0.013547425158321857,
          "validation_loss": 0.010985392145812511,
          "train_metric": 0.0013644471764564514,
          "validation_metric": 0.0011064085410907865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9593189469997014
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.035897616122383624,
          "train_loss": 0.010429147630929947,
          "validation_loss": 0.008227070793509483,
          "train_metric": 0.0010503856465220451,
          "validation_metric": 0.0008286003721877933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9739117310000438
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.05127669312059879,
          "train_loss": 0.013014156371355057,
          "validation_loss": 0.009743716567754745,
          "train_metric": 0.0013107381528243423,
          "validation_metric": 0.0009813514770939946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9881875659993966
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.036279824387747794,
          "train_loss": 0.011671734973788261,
          "validation_loss": 0.009263986721634865,
          "train_metric": 0.0011755345622077584,
          "validation_metric": 0.0009330347529612482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0022870709999552
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.044308391283266246,
          "train_loss": 0.015491495840251446,
          "validation_loss": 0.012892412953078747,
          "train_metric": 0.001560246804729104,
          "validation_metric": 0.0012984767090529203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0166651090003143
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.03142931987531483,
          "train_loss": 0.015530009754002094,
          "validation_loss": 0.013506130315363407,
          "train_metric": 0.0015641256468370557,
          "validation_metric": 0.0013602878898382187,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0304804789993796
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.025552454957505688,
          "train_loss": 0.010796266607940197,
          "validation_loss": 0.008493127301335335,
          "train_metric": 0.001087360316887498,
          "validation_metric": 0.0008553967345505953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0444202779999614
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.028113312495406717,
          "train_loss": 0.015652228146791458,
          "validation_loss": 0.013093183748424053,
          "train_metric": 0.0015764350537210703,
          "validation_metric": 0.0013186978176236153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0583610869998665
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.03768110822420567,
          "train_loss": 0.022739609703421593,
          "validation_loss": 0.015673093497753143,
          "train_metric": 0.0022902502678334713,
          "validation_metric": 0.001578536699526012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0724870009998995
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.04467264303821139,
          "train_loss": 0.0191511083394289,
          "validation_loss": 0.015441679395735264,
          "train_metric": 0.0019288293551653624,
          "validation_metric": 0.0015552295371890068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.086766914000691
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.04263714700937271,
          "train_loss": 0.012109002098441124,
          "validation_loss": 0.00962262973189354,
          "train_metric": 0.001219574362039566,
          "validation_metric": 0.0009691560408100486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1006179390005855
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.03301180002745241,
          "train_loss": 0.011301705613732338,
          "validation_loss": 0.009314539842307568,
          "train_metric": 0.0011382662924006581,
          "validation_metric": 0.0009381263516843319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1144014620003873
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.03886362974299118,
          "train_loss": 0.008863769471645355,
          "validation_loss": 0.006436891853809357,
          "train_metric": 0.0008927264134399593,
          "validation_metric": 0.0006483001634478569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.129270784000255
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.03021861833985895,
          "train_loss": 0.011773141101002693,
          "validation_loss": 0.009424152784049511,
          "train_metric": 0.0011857477948069572,
          "validation_metric": 0.0009491662494838238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1434564489991317
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.02948287274921313,
          "train_loss": 0.00958342757076025,
          "validation_loss": 0.007140933070331812,
          "train_metric": 0.0009652076987549663,
          "validation_metric": 0.0007192085613496602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2355774520001432
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.02246356153045781,
          "train_loss": 0.008634931407868862,
          "validation_loss": 0.006989111192524433,
          "train_metric": 0.0008696787408553064,
          "validation_metric": 0.0007039176998659968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2499162570002227
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.02831532701384276,
          "train_loss": 0.011877615936100483,
          "validation_loss": 0.008927795104682446,
          "train_metric": 0.001196269877254963,
          "validation_metric": 0.0008991746581159532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2639441320006881
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.029570034705102444,
          "train_loss": 0.01006936002522707,
          "validation_loss": 0.008113945834338665,
          "train_metric": 0.0010141490492969751,
          "validation_metric": 0.0008172068628482521,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2779613229995448
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.029297701257746667,
          "train_loss": 0.010209252126514912,
          "validation_loss": 0.00812752079218626,
          "train_metric": 0.0010282384464517236,
          "validation_metric": 0.0008185741025954485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2918942340002104
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.024907023354899138,
          "train_loss": 0.011722237803041935,
          "validation_loss": 0.009019540622830391,
          "train_metric": 0.0011806209804490209,
          "validation_metric": 0.0009084150660783052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3057774619992415
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.0361407520249486,
          "train_loss": 0.010423202998936176,
          "validation_loss": 0.007892029359936714,
          "train_metric": 0.0010497866896912456,
          "validation_metric": 0.0007948562852106988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3197745809993648
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.030154311971273273,
          "train_loss": 0.012033813633024693,
          "validation_loss": 0.008939504623413086,
          "train_metric": 0.0012120016617700458,
          "validation_metric": 0.0009003541199490428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.333782790999976
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.025249079873901792,
          "train_loss": 0.009723431430757046,
          "validation_loss": 0.0067979395389556885,
          "train_metric": 0.0009793085046112537,
          "validation_metric": 0.0006846635951660573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3473759579992475
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.029496450151782483,
          "train_loss": 0.014346702955663204,
          "validation_loss": 0.012257054448127747,
          "train_metric": 0.0014449474401772022,
          "validation_metric": 0.0012344856513664126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.361148112000592
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.03546947590075433,
          "train_loss": 0.01361977867782116,
          "validation_loss": 0.012962495908141136,
          "train_metric": 0.0013717340771108866,
          "validation_metric": 0.0013055350864306092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.374956296000164
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.024128404969815165,
          "train_loss": 0.008195634931325912,
          "validation_loss": 0.006703309714794159,
          "train_metric": 0.0008254343410953879,
          "validation_metric": 0.0006751327309757471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3887257349997526
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.02903458318905905,
          "train_loss": 0.007430072873830795,
          "validation_loss": 0.005720633547753096,
          "train_metric": 0.0007483296794816852,
          "validation_metric": 0.000576161255594343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.403288731999055
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.029824188153725117,
          "train_loss": 0.011464345268905163,
          "validation_loss": 0.0079294852912426,
          "train_metric": 0.0011546468595042825,
          "validation_metric": 0.0007986286655068398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4173996339995938
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.03215248300693929,
          "train_loss": 0.01514676958322525,
          "validation_loss": 0.012725308537483215,
          "train_metric": 0.001525527099147439,
          "validation_metric": 0.0012816465459764004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.431410085000607
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.035545307968277484,
          "train_loss": 0.013589169830083847,
          "validation_loss": 0.01061139814555645,
          "train_metric": 0.0013686513993889093,
          "validation_metric": 0.0010687410831451416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4455667879992689
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.02980829670559615,
          "train_loss": 0.011178470216691494,
          "validation_loss": 0.008337276987731457,
          "train_metric": 0.0011258546728640795,
          "validation_metric": 0.0008396999328397214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4592459020004753
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.02462330777780153,
          "train_loss": 0.008284825831651688,
          "validation_loss": 0.006651068106293678,
          "train_metric": 0.0008344173547811806,
          "validation_metric": 0.0006698711076751351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.473878593000336
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.03150128939887509,
          "train_loss": 0.009746055118739605,
          "validation_loss": 0.007086253724992275,
          "train_metric": 0.0009815869852900505,
          "validation_metric": 0.000713701534550637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4876878800005215
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.02691948553547263,
          "train_loss": 0.010426592081785202,
          "validation_loss": 0.008561599999666214,
          "train_metric": 0.001050128135830164,
          "validation_metric": 0.0008622930618003011,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5016808410000522
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0373539833817631,
          "train_loss": 0.010982412844896317,
          "validation_loss": 0.009557723999023438,
          "train_metric": 0.0011061084223911166,
          "validation_metric": 0.0009626190294511616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5153772699995898
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.020326363475760445,
          "train_loss": 0.008454208262264729,
          "validation_loss": 0.005955847911536694,
          "train_metric": 0.0008514769142493606,
          "validation_metric": 0.000599851191509515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5291801940002188
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.03659987694118172,
          "train_loss": 0.01151164062321186,
          "validation_loss": 0.00993740651756525,
          "train_metric": 0.0011594102252274752,
          "validation_metric": 0.001000859192572534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5429872009990504
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.03324725152924657,
          "train_loss": 0.00999743863940239,
          "validation_loss": 0.00928197056055069,
          "train_metric": 0.001006905222311616,
          "validation_metric": 0.0009348460589535534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5569593460004398
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.036985346232540905,
          "train_loss": 0.012149623595178127,
          "validation_loss": 0.01056765578687191,
          "train_metric": 0.00122366554569453,
          "validation_metric": 0.001064335578121245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5709040889996686
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.03201202052878216,
          "train_loss": 0.01054903119802475,
          "validation_loss": 0.008227379061281681,
          "train_metric": 0.0010624598944559693,
          "validation_metric": 0.000828631455078721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5848520649997226
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.028197633509989828,
          "train_loss": 0.010868635028600693,
          "validation_loss": 0.007757158949971199,
          "train_metric": 0.0010946490801870823,
          "validation_metric": 0.0007812726544216275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.598851757000375
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.028424554970115423,
          "train_loss": 0.009804504923522472,
          "validation_loss": 0.008423903025686741,
          "train_metric": 0.000987473875284195,
          "validation_metric": 0.000848424737341702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6126983310005016
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.035099522094242275,
          "train_loss": 0.010938765481114388,
          "validation_loss": 0.00905358511954546,
          "train_metric": 0.001101712347008288,
          "validation_metric": 0.0009118440211750567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6266514210001333
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.031177557248156518,
          "train_loss": 0.00824183039367199,
          "validation_loss": 0.006056551355868578,
          "train_metric": 0.0008300869958475232,
          "validation_metric": 0.000609993701800704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6784751049999613
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.02660051637212746,
          "train_loss": 0.009901618584990501,
          "validation_loss": 0.005773098673671484,
          "train_metric": 0.0009972548577934504,
          "validation_metric": 0.0005814454052597284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6953456419996655
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.02945774930412881,
          "train_loss": 0.010233461856842041,
          "validation_loss": 0.008965086191892624,
          "train_metric": 0.001030676648952067,
          "validation_metric": 0.0009029305656440556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7109973219994572
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.033217832824448124,
          "train_loss": 0.011439603753387928,
          "validation_loss": 0.009311413392424583,
          "train_metric": 0.0011521551059558988,
          "validation_metric": 0.0009378112736158073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7263175659991248
        }
      ],
      "best": {
        "epoch": 61,
        "updates": 976
      },
      "best_validation_metric": 0.000576161255594343,
      "config": {
        "label": "BatchNorm",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "batch",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.024003359256452896
    },
    "5c044ad193b4_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.5785819953307509,
          "train_loss": 0.24066151678562164,
          "validation_loss": 0.15968576073646545,
          "train_metric": 0.024238549172878265,
          "validation_metric": 0.016082964837551117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03456230999836407
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.20281902188435197,
          "train_loss": 0.14825578033924103,
          "validation_loss": 0.12119951099157333,
          "train_metric": 0.014931779354810715,
          "validation_metric": 0.012206770479679108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0832196479987033
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13218980515375733,
          "train_loss": 0.11559955775737762,
          "validation_loss": 0.09988148510456085,
          "train_metric": 0.011642763391137123,
          "validation_metric": 0.010059698484838009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10179930199956289
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10902172443456948,
          "train_loss": 0.09608157724142075,
          "validation_loss": 0.08105382323265076,
          "train_metric": 0.009676986373960972,
          "validation_metric": 0.008163443766534328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11974058799933118
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08852994372136891,
          "train_loss": 0.07870391011238098,
          "validation_loss": 0.06390558928251266,
          "train_metric": 0.00792676955461502,
          "validation_metric": 0.0064363377168774605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14432031299838854
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.0738776980433613,
          "train_loss": 0.06744672358036041,
          "validation_loss": 0.05580252408981323,
          "train_metric": 0.006792987231165171,
          "validation_metric": 0.005620225332677364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16680799999994633
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.0662799112033099,
          "train_loss": 0.062105804681777954,
          "validation_loss": 0.051005952060222626,
          "train_metric": 0.006255070213228464,
          "validation_metric": 0.0051371329464018345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19084788199870673
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06061952211894095,
          "train_loss": 0.058522891253232956,
          "validation_loss": 0.04748249426484108,
          "train_metric": 0.005894211120903492,
          "validation_metric": 0.0047822631895542145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21352853600001254
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0586743209278211,
          "train_loss": 0.05518710985779762,
          "validation_loss": 0.04696771502494812,
          "train_metric": 0.00555824302136898,
          "validation_metric": 0.0047304159961640835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25539982599912037
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05452180642168969,
          "train_loss": 0.05012526735663414,
          "validation_loss": 0.04216490313410759,
          "train_metric": 0.005048433784395456,
          "validation_metric": 0.004246694967150688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32100924599944847
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0489687561057508,
          "train_loss": 0.0455508716404438,
          "validation_loss": 0.0387922078371048,
          "train_metric": 0.004587717354297638,
          "validation_metric": 0.003907009027898312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.426063526998405
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.045328833162784576,
          "train_loss": 0.04298621788620949,
          "validation_loss": 0.03603599593043327,
          "train_metric": 0.004329414572566748,
          "validation_metric": 0.0036294138990342617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49308858799849986
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.040962925646454096,
          "train_loss": 0.03840823099017143,
          "validation_loss": 0.03274441510438919,
          "train_metric": 0.003868336323648691,
          "validation_metric": 0.003297897521406412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.560810731998572
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.035591799300163984,
          "train_loss": 0.03237031772732735,
          "validation_loss": 0.027760149911046028,
          "train_metric": 0.003260219469666481,
          "validation_metric": 0.0027959002181887627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6278492389992607
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.034175259876064956,
          "train_loss": 0.030200820416212082,
          "validation_loss": 0.02461184561252594,
          "train_metric": 0.0030417158268392086,
          "validation_metric": 0.0024788149166852236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7699930939998012
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.029044358525425196,
          "train_loss": 0.02824326977133751,
          "validation_loss": 0.02484940178692341,
          "train_metric": 0.0028445585630834103,
          "validation_metric": 0.0025027410592883825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8819259399988368
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.027968548878561705,
          "train_loss": 0.026699528098106384,
          "validation_loss": 0.022850092500448227,
          "train_metric": 0.0026890786830335855,
          "validation_metric": 0.0023013774771243334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1298645199985913
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.029929061303846538,
          "train_loss": 0.028845902532339096,
          "validation_loss": 0.023522071540355682,
          "train_metric": 0.002905253553763032,
          "validation_metric": 0.002369056921452284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1424046369993448
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.0253515514486935,
          "train_loss": 0.022620845586061478,
          "validation_loss": 0.019433408975601196,
          "train_metric": 0.0022782885935157537,
          "validation_metric": 0.001957261934876442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1545746409992717
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02046690188581124,
          "train_loss": 0.019571326673030853,
          "validation_loss": 0.016824472695589066,
          "train_metric": 0.0019711521454155445,
          "validation_metric": 0.00169449916575104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1668775979997008
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.021430058928672224,
          "train_loss": 0.019809599965810776,
          "validation_loss": 0.01710662618279457,
          "train_metric": 0.001995150465518236,
          "validation_metric": 0.0017229169607162476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1787090629986778
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.019933807488996536,
          "train_loss": 0.018322203308343887,
          "validation_loss": 0.015159186907112598,
          "train_metric": 0.0018453451339155436,
          "validation_metric": 0.0015267777489498258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1912519229990721
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.0184865053743124,
          "train_loss": 0.018727105110883713,
          "validation_loss": 0.016270481050014496,
          "train_metric": 0.0018861254211515188,
          "validation_metric": 0.001638703397475183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2029774989987345
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019489634316414595,
          "train_loss": 0.018848121166229248,
          "validation_loss": 0.015607197768986225,
          "train_metric": 0.0018983136396855116,
          "validation_metric": 0.0015718997456133366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2149289739991218
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.025991787144448608,
          "train_loss": 0.022055301815271378,
          "validation_loss": 0.01751544512808323,
          "train_metric": 0.0022213291376829147,
          "validation_metric": 0.0017640916630625725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2269664119994559
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.020158647326752543,
          "train_loss": 0.019878527149558067,
          "validation_loss": 0.015789136290550232,
          "train_metric": 0.0020020923111587763,
          "validation_metric": 0.0015902240993455052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.238922806998744
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01832235194160603,
          "train_loss": 0.015337939374148846,
          "validation_loss": 0.013187260366976261,
          "train_metric": 0.0015447810292243958,
          "validation_metric": 0.001328172511421144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2511755919986172
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015697864320827648,
          "train_loss": 0.01441369578242302,
          "validation_loss": 0.012476164847612381,
          "train_metric": 0.0014516946393996477,
          "validation_metric": 0.0012565534561872482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2634009349985718
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.01796175292111002,
          "train_loss": 0.015583343803882599,
          "validation_loss": 0.012778116390109062,
          "train_metric": 0.0015694972826167941,
          "validation_metric": 0.0012869652127847075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2755708770000638
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.017315161181613803,
          "train_loss": 0.026303283870220184,
          "validation_loss": 0.022693904116749763,
          "train_metric": 0.0026491701137274504,
          "validation_metric": 0.002285646740347147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.287456342999576
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.018650115351192653,
          "train_loss": 0.015033909119665623,
          "validation_loss": 0.012586742639541626,
          "train_metric": 0.0015141601907089353,
          "validation_metric": 0.0012676906771957874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.299320229998557
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01525422916165553,
          "train_loss": 0.012996863573789597,
          "validation_loss": 0.011027900502085686,
          "train_metric": 0.0013089965796098113,
          "validation_metric": 0.0011106899473816156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.311824328999137
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.017277867038501427,
          "train_loss": 0.018468111753463745,
          "validation_loss": 0.01605597883462906,
          "train_metric": 0.0018600407056510448,
          "validation_metric": 0.001617099391296506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3238100290000148
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.017836909275501966,
          "train_loss": 0.015464769676327705,
          "validation_loss": 0.013479636050760746,
          "train_metric": 0.001557554816827178,
          "validation_metric": 0.0013576196506619453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3353904409996176
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014926508301869035,
          "train_loss": 0.01525525376200676,
          "validation_loss": 0.012201250530779362,
          "train_metric": 0.0015364533755928278,
          "validation_metric": 0.00122886523604393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3473869230001583
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.013606283508124761,
          "train_loss": 0.01189349964261055,
          "validation_loss": 0.010086850263178349,
          "train_metric": 0.0011978696566075087,
          "validation_metric": 0.0010159105295315385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.363959979998981
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013865286426153034,
          "train_loss": 0.01421459298580885,
          "validation_loss": 0.012337259016931057,
          "train_metric": 0.0014316417509689927,
          "validation_metric": 0.00124256347771734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4074436339997192
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.013906719570513815,
          "train_loss": 0.01308034174144268,
          "validation_loss": 0.011206195689737797,
          "train_metric": 0.0013174042105674744,
          "validation_metric": 0.0011286470107734203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4503923039992515
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.015102692559594288,
          "train_loss": 0.012156970798969269,
          "validation_loss": 0.009952702559530735,
          "train_metric": 0.0012244055978953838,
          "validation_metric": 0.00100239971652627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4983287539998855
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.011600760830333456,
          "train_loss": 0.01074554119259119,
          "validation_loss": 0.009170734323561192,
          "train_metric": 0.0010822515469044447,
          "validation_metric": 0.0009236427722498775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5197546359995613
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011461123736808077,
          "train_loss": 0.011186208575963974,
          "validation_loss": 0.008815637789666653,
          "train_metric": 0.001126633957028389,
          "validation_metric": 0.000887878704816103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5371598490000906
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.012731426904792897,
          "train_loss": 0.011028323322534561,
          "validation_loss": 0.008532376028597355,
          "train_metric": 0.0011107323225587606,
          "validation_metric": 0.0008593496750108898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5548930529985228
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011592802344239317,
          "train_loss": 0.01758863590657711,
          "validation_loss": 0.0152521887794137,
          "train_metric": 0.0017714631976559758,
          "validation_metric": 0.0015361446421593428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5729601129987714
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014004348166054115,
          "train_loss": 0.011498655192553997,
          "validation_loss": 0.009649231098592281,
          "train_metric": 0.0011581024155020714,
          "validation_metric": 0.0009718352230265737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6052912189989001
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.011971019470365718,
          "train_loss": 0.010247869417071342,
          "validation_loss": 0.008546342141926289,
          "train_metric": 0.0010321278823539615,
          "validation_metric": 0.0008607562049292028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6230805089999194
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010324565839255229,
          "train_loss": 0.009410005994141102,
          "validation_loss": 0.008049655705690384,
          "train_metric": 0.0009477414423599839,
          "validation_metric": 0.0008107318426482379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6404385559999355
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0099878313485533,
          "train_loss": 0.009574130177497864,
          "validation_loss": 0.008396321907639503,
          "train_metric": 0.0009642713703215122,
          "validation_metric": 0.0008456468349322677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6572798050001438
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010388583701569587,
          "train_loss": 0.010846927762031555,
          "validation_loss": 0.009024742990732193,
          "train_metric": 0.0010924628004431725,
          "validation_metric": 0.0009089390514418483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6958791649994964
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.011729866499081254,
          "train_loss": 0.008839637972414494,
          "validation_loss": 0.007451624143868685,
          "train_metric": 0.0008902959525585175,
          "validation_metric": 0.0007505002431571484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7311801249998098
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009075396141270176,
          "train_loss": 0.008698005229234695,
          "validation_loss": 0.007412043400108814,
          "train_metric": 0.0008760312921367586,
          "validation_metric": 0.0007465138914994895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7490257069985091
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008586802636273205,
          "train_loss": 0.009239615872502327,
          "validation_loss": 0.00846143253147602,
          "train_metric": 0.0009305804269388318,
          "validation_metric": 0.000852204451803118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7656761889993504
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.009399648464750499,
          "train_loss": 0.008207714185118675,
          "validation_loss": 0.007094361819326878,
          "train_metric": 0.0008266508812084794,
          "validation_metric": 0.0007145180716179311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.782969754000078
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.009622707089874893,
          "train_loss": 0.008783034980297089,
          "validation_loss": 0.007847893983125687,
          "train_metric": 0.0008845950360409915,
          "validation_metric": 0.0007904112571850419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7995357319996401
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009374020868563093,
          "train_loss": 0.008376436308026314,
          "validation_loss": 0.007370653096586466,
          "train_metric": 0.0008436439675278962,
          "validation_metric": 0.0007423452916555107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8158730729992385
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008409974529058672,
          "train_loss": 0.008557048626244068,
          "validation_loss": 0.006746719125658274,
          "train_metric": 0.0008618345018476248,
          "validation_metric": 0.0006795047665946186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8330296579988499
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.008126632194034755,
          "train_loss": 0.007593871559947729,
          "validation_loss": 0.006299670785665512,
          "train_metric": 0.000764826952945441,
          "validation_metric": 0.0006344797438941896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.850268630998471
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008103969594230875,
          "train_loss": 0.007627340033650398,
          "validation_loss": 0.006149530876427889,
          "train_metric": 0.0007681977585889399,
          "validation_metric": 0.0006193582667037845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8675733789987135
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.007666467514354736,
          "train_loss": 0.007350016385316849,
          "validation_loss": 0.0060503422282636166,
          "train_metric": 0.0007402666378766298,
          "validation_metric": 0.0006093683768995106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8849525139994512
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.008778885996434838,
          "train_loss": 0.007884827442467213,
          "validation_loss": 0.007127765100449324,
          "train_metric": 0.0007941309013403952,
          "validation_metric": 0.0007178823580034077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9016449229984573
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007124831900000572,
          "train_loss": 0.008507896214723587,
          "validation_loss": 0.006483363453298807,
          "train_metric": 0.0008568841149099171,
          "validation_metric": 0.0006529804668389261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.917816874998607
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.00802242913050577,
          "train_loss": 0.0072455033659935,
          "validation_loss": 0.005634790752083063,
          "train_metric": 0.000729740597307682,
          "validation_metric": 0.0005675156135112047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9349848089987063
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007499659288441762,
          "train_loss": 0.006905286107212305,
          "validation_loss": 0.005780530162155628,
          "train_metric": 0.0006954751443117857,
          "validation_metric": 0.0005821938975714147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9513606049986265
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006983088416745886,
          "train_loss": 0.00676384475082159,
          "validation_loss": 0.005742666777223349,
          "train_metric": 0.0006812297506257892,
          "validation_metric": 0.0005783804226666689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9682599729985668
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007186421637015883,
          "train_loss": 0.006360976025462151,
          "validation_loss": 0.0053828442469239235,
          "train_metric": 0.0006406542379409075,
          "validation_metric": 0.000542140391189605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9855079079989082
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008146895095705986,
          "train_loss": 0.00710770720615983,
          "validation_loss": 0.0064037274569272995,
          "train_metric": 0.0007158622611314058,
          "validation_metric": 0.0006449599750339985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0018080599984387
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0070507670170627534,
          "train_loss": 0.006784606724977493,
          "validation_loss": 0.005462770350277424,
          "train_metric": 0.0006833208026364446,
          "validation_metric": 0.0005501902778632939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0182304179998027
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007892993970017415,
          "train_loss": 0.007619489450007677,
          "validation_loss": 0.005710892844945192,
          "train_metric": 0.0007674070075154305,
          "validation_metric": 0.0005751801654696465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.035295462999784
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0075215069082332775,
          "train_loss": 0.008999602869153023,
          "validation_loss": 0.007687784265726805,
          "train_metric": 0.0009064070181921124,
          "validation_metric": 0.0007742854650132358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.052044150999791
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.008329692267579958,
          "train_loss": 0.006699476856738329,
          "validation_loss": 0.005957436747848988,
          "train_metric": 0.000674746697768569,
          "validation_metric": 0.0006000111461617053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.06854009899871
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006823713432822842,
          "train_loss": 0.006718658842146397,
          "validation_loss": 0.005331420339643955,
          "train_metric": 0.0006766787264496088,
          "validation_metric": 0.0005369611317291856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0863856699997996
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0063610283978050575,
          "train_loss": 0.006560879293829203,
          "validation_loss": 0.005013317335397005,
          "train_metric": 0.0006607878021895885,
          "validation_metric": 0.0005049229366704822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1039569259992277
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006628050803556107,
          "train_loss": 0.005550465080887079,
          "validation_loss": 0.00450581219047308,
          "train_metric": 0.0005590225337073207,
          "validation_metric": 0.00045380889787338674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1211918569988484
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006065987341571599,
          "train_loss": 0.006213136482983828,
          "validation_loss": 0.004682971630245447,
          "train_metric": 0.000625764369033277,
          "validation_metric": 0.0004716518160421401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.138157751000108
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0066800750500988215,
          "train_loss": 0.007923943921923637,
          "validation_loss": 0.00671512121334672,
          "train_metric": 0.0007980706868693233,
          "validation_metric": 0.0006763224373571575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.154897185999289
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.008286909156595357,
          "train_loss": 0.005879192613065243,
          "validation_loss": 0.005012104287743568,
          "train_metric": 0.0005921307601965964,
          "validation_metric": 0.0005048008169978857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1716091759990377
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.008240004448452964,
          "train_loss": 0.006389837712049484,
          "validation_loss": 0.005709755700081587,
          "train_metric": 0.0006435610703192651,
          "validation_metric": 0.0005750657292082906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.188150340998618
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007031992237898521,
          "train_loss": 0.0061054527759552,
          "validation_loss": 0.005275707691907883,
          "train_metric": 0.000614918943028897,
          "validation_metric": 0.0005313499132171273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.204625161999502
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006232404222828336,
          "train_loss": 0.005883560515940189,
          "validation_loss": 0.004626329056918621,
          "train_metric": 0.0005925706936977804,
          "validation_metric": 0.00046594697050750256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2210073149999516
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006698730117932428,
          "train_loss": 0.00830888468772173,
          "validation_loss": 0.005257878918200731,
          "train_metric": 0.0008368405397050083,
          "validation_metric": 0.0005295543232932687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2375409369997215
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.007175867009209469,
          "train_loss": 0.005164492409676313,
          "validation_loss": 0.0041005248203873634,
          "train_metric": 0.0005201487801969051,
          "validation_metric": 0.00041298987343907356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.25865182599955
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005640656236209907,
          "train_loss": 0.005428474862128496,
          "validation_loss": 0.004622603766620159,
          "train_metric": 0.000546736060641706,
          "validation_metric": 0.00046557170571759343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3041251889990235
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.007278883465914987,
          "train_loss": 0.007279138546437025,
          "validation_loss": 0.005336048547178507,
          "train_metric": 0.0007331281085498631,
          "validation_metric": 0.0005374272586777806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.333465834999515
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006428858388972003,
          "train_loss": 0.005414268001914024,
          "validation_loss": 0.004378072917461395,
          "train_metric": 0.000545305258128792,
          "validation_metric": 0.0004409434332046658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3503695409999636
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.006930987336090766,
          "train_loss": 0.004939272999763489,
          "validation_loss": 0.004174148663878441,
          "train_metric": 0.0004974654875695705,
          "validation_metric": 0.0004204050055705011,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.367566610999347
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005648306665534619,
          "train_loss": 0.004681389778852463,
          "validation_loss": 0.004043219611048698,
          "train_metric": 0.00047149244346655905,
          "validation_metric": 0.0004072182346135378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.385166254998694
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005824902589665726,
          "train_loss": 0.005570245441049337,
          "validation_loss": 0.004077776335179806,
          "train_metric": 0.0005610146909020841,
          "validation_metric": 0.00041069870349019766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4021967129992845
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005732758989324793,
          "train_loss": 0.006048447918146849,
          "validation_loss": 0.0051933652721345425,
          "train_metric": 0.0006091775139793754,
          "validation_metric": 0.0005230566603131592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5184469439991517
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.00487511732353596,
          "train_loss": 0.004281728528439999,
          "validation_loss": 0.003500000573694706,
          "train_metric": 0.00043124001240357757,
          "validation_metric": 0.00035250725341029465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.536472458999924
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005490752446348779,
          "train_loss": 0.0075278449803590775,
          "validation_loss": 0.005237332079559565,
          "train_metric": 0.000758176960516721,
          "validation_metric": 0.0005274848663248122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5535832759996993
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.006032784658600576,
          "train_loss": 0.004147201310843229,
          "validation_loss": 0.003196843434125185,
          "train_metric": 0.00041769095696508884,
          "validation_metric": 0.00032197439577430487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5708482789996197
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0045539015118265525,
          "train_loss": 0.003808197332546115,
          "validation_loss": 0.0031398420687764883,
          "train_metric": 0.0003835476818494499,
          "validation_metric": 0.00031623340328224003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.588992982999116
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004446769387868699,
          "train_loss": 0.003838970558717847,
          "validation_loss": 0.00305731687694788,
          "train_metric": 0.00038664706517010927,
          "validation_metric": 0.0003079217567574233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6063390059989615
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004661549588490743,
          "train_loss": 0.003889218671247363,
          "validation_loss": 0.0030916959512978792,
          "train_metric": 0.0003917078720405698,
          "validation_metric": 0.0003113843558821827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6232684569986304
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004343237917055376,
          "train_loss": 0.003794271731749177,
          "validation_loss": 0.0030378333758562803,
          "train_metric": 0.0003821451682597399,
          "validation_metric": 0.000305959430988878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6413282379999146
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004010706965345889,
          "train_loss": 0.004695076495409012,
          "validation_loss": 0.0032074693590402603,
          "train_metric": 0.0004728709172923118,
          "validation_metric": 0.0003230445727240294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6586705459994846
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0048800584190757945,
          "train_loss": 0.004135572351515293,
          "validation_loss": 0.003357178997248411,
          "train_metric": 0.00041651964420452714,
          "validation_metric": 0.00033812274341471493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6751358549990982
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.004035678837681189,
          "train_loss": 0.003541690995916724,
          "validation_loss": 0.0029904022812843323,
          "train_metric": 0.00035670612123794854,
          "validation_metric": 0.0003011823573615402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6923814989986568
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004549244105874095,
          "train_loss": 0.007052480708807707,
          "validation_loss": 0.005183737725019455,
          "train_metric": 0.0007103000534698367,
          "validation_metric": 0.0005220869788900018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7092104089988425
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.008762747966102324,
          "train_loss": 0.004422488622367382,
          "validation_loss": 0.003422736655920744,
          "train_metric": 0.00044541689567267895,
          "validation_metric": 0.00034472550032660365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7256901589989866
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004970894042344298,
          "train_loss": 0.007190850097686052,
          "validation_loss": 0.004736239090561867,
          "train_metric": 0.0007242360734380782,
          "validation_metric": 0.00047701667062938213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.742250064999098
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.006154970556963235,
          "train_loss": 0.006614692974835634,
          "validation_loss": 0.006176328286528587,
          "train_metric": 0.0006662076339125633,
          "validation_metric": 0.0006220571231096983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.75904878899928
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.004848615717492066,
          "train_loss": 0.0038373428396880627,
          "validation_loss": 0.0031854603439569473,
          "train_metric": 0.0003864831232931465,
          "validation_metric": 0.00032082790858112276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.775720169998749
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0043659103830577806,
          "train_loss": 0.0037215915508568287,
          "validation_loss": 0.0033144704066216946,
          "train_metric": 0.00037482508923858404,
          "validation_metric": 0.0003338213136885315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.792124491999857
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0037703865782532375,
          "train_loss": 0.0033768422435969114,
          "validation_loss": 0.0028291926719248295,
          "train_metric": 0.00034010314266197383,
          "validation_metric": 0.00028494588332250714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8096051339998667
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003850408087600954,
          "train_loss": 0.0038566645234823227,
          "validation_loss": 0.002810301724821329,
          "train_metric": 0.0003884290927089751,
          "validation_metric": 0.00028304324951022863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8271638579990395
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0036429797328310087,
          "train_loss": 0.003893908578902483,
          "validation_loss": 0.0028391662053763866,
          "train_metric": 0.00039218022720888257,
          "validation_metric": 0.0002859504020307213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8443843139993987
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.003939842183172004,
          "train_loss": 0.004020857624709606,
          "validation_loss": 0.00285907881334424,
          "train_metric": 0.0004049660637974739,
          "validation_metric": 0.0002879559760913253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.861318372999449
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.006216302259417716,
          "train_loss": 0.004236565437167883,
          "validation_loss": 0.003314891830086708,
          "train_metric": 0.0004266913456376642,
          "validation_metric": 0.00033386374707333744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8776491209991946
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.004633348784409463,
          "train_loss": 0.0040680477395653725,
          "validation_loss": 0.0030227238312363625,
          "train_metric": 0.0004097189230378717,
          "validation_metric": 0.0003044376499019563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.894238131999373
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.00391630687227007,
          "train_loss": 0.006275310181081295,
          "validation_loss": 0.004126521293073893,
          "train_metric": 0.000632026232779026,
          "validation_metric": 0.00041560811223462224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9107500559985056
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004763353004818782,
          "train_loss": 0.0034501319751143456,
          "validation_loss": 0.002509691519662738,
          "train_metric": 0.0003474846016615629,
          "validation_metric": 0.00025276694213971496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9285787849985354
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.00421532570180716,
          "train_loss": 0.0038624031003564596,
          "validation_loss": 0.0031479850877076387,
          "train_metric": 0.00038900706567801535,
          "validation_metric": 0.00031705352012068033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.945572523998635
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0037217386852717027,
          "train_loss": 0.006393822841346264,
          "validation_loss": 0.005712313111871481,
          "train_metric": 0.0006439624121412635,
          "validation_metric": 0.0005753233563154936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.962359381999704
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.006156473507871851,
          "train_loss": 0.003115376690402627,
          "validation_loss": 0.0025999839417636395,
          "train_metric": 0.0003137692983727902,
          "validation_metric": 0.00026186087052337825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9794847889988887
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0036181848408887163,
          "train_loss": 0.0032176231034100056,
          "validation_loss": 0.0024235867895185947,
          "train_metric": 0.0003240671940147877,
          "validation_metric": 0.00024409478646703064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9973141979990032
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0032519254091312177,
          "train_loss": 0.0034723232965916395,
          "validation_loss": 0.003006942104548216,
          "train_metric": 0.00034971965942531824,
          "validation_metric": 0.00030284817330539227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0144456879988866
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0029091055839671753,
          "train_loss": 0.0030959751456975937,
          "validation_loss": 0.0021414398215711117,
          "train_metric": 0.0003118152671959251,
          "validation_metric": 0.0002156779810320586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0321029229999112
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0030300220460048877,
          "train_loss": 0.002905888482928276,
          "validation_loss": 0.0024696867913007736,
          "train_metric": 0.0002926704764831811,
          "validation_metric": 0.00024873780785128474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.04868377099956
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0030614418465120252,
          "train_loss": 0.003689597826451063,
          "validation_loss": 0.0024924471508711576,
          "train_metric": 0.00037160280044190586,
          "validation_metric": 0.0002510301419533789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.06505884499893
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0033793316761148162,
          "train_loss": 0.0029266411438584328,
          "validation_loss": 0.002395999850705266,
          "train_metric": 0.000294760538963601,
          "validation_metric": 0.00024131633108481765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.083147730998462
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0032884067622944713,
          "train_loss": 0.0027470323257148266,
          "validation_loss": 0.0022338894195854664,
          "train_metric": 0.0002766710240393877,
          "validation_metric": 0.00022498915495816618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0998379229986313
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0035691659795702435,
          "train_loss": 0.0031002990435808897,
          "validation_loss": 0.002106620464473963,
          "train_metric": 0.00031225071870721877,
          "validation_metric": 0.000212171085877344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.117104687999017
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.003656093656900339,
          "train_loss": 0.003222736297175288,
          "validation_loss": 0.0027210693806409836,
          "train_metric": 0.0003245821862947196,
          "validation_metric": 0.00027405619039200246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1337505809988215
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0036966746047255583,
          "train_loss": 0.002750445157289505,
          "validation_loss": 0.002295314334332943,
          "train_metric": 0.00027701474027708173,
          "validation_metric": 0.00023117563978303224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1508741049983655
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0033217461896128953,
          "train_loss": 0.0028786661569029093,
          "validation_loss": 0.0021965266205370426,
          "train_metric": 0.00028992866282351315,
          "validation_metric": 0.00022122611699160188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1682930079987273
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.00400566911775968,
          "train_loss": 0.0028464628849178553,
          "validation_loss": 0.0023582750000059605,
          "train_metric": 0.0002866853028535843,
          "validation_metric": 0.00023751685512252152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.184884501999477
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0029259902075864375,
          "train_loss": 0.002619984559714794,
          "validation_loss": 0.0021287843119353056,
          "train_metric": 0.0002638752630446106,
          "validation_metric": 0.00021440337877720594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.20155274499848
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0029659844767593313,
          "train_loss": 0.003342970972880721,
          "validation_loss": 0.0023593404330313206,
          "train_metric": 0.00033669182448647916,
          "validation_metric": 0.00023762413184158504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2180421550001483
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0025056919257622212,
          "train_loss": 0.0035709175281226635,
          "validation_loss": 0.0029357061721384525,
          "train_metric": 0.00035964971175417304,
          "validation_metric": 0.0002956735552288592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2344251109989273
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.004325161571614444,
          "train_loss": 0.004080065526068211,
          "validation_loss": 0.0034732180647552013,
          "train_metric": 0.0004109292640350759,
          "validation_metric": 0.0003498097648844123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.251486687999204
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0034356102732999716,
          "train_loss": 0.003065694123506546,
          "validation_loss": 0.002774572465568781,
          "train_metric": 0.00030876544769853354,
          "validation_metric": 0.0002794447646010667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2835596869990695
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0026468074865988456,
          "train_loss": 0.002433589193969965,
          "validation_loss": 0.00198219227604568,
          "train_metric": 0.00024510218645446,
          "validation_metric": 0.00019963915110565722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3015256479993695
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0026325581020500977,
          "train_loss": 0.0026633315719664097,
          "validation_loss": 0.0019586319103837013,
          "train_metric": 0.00026824098313227296,
          "validation_metric": 0.00019726621394511312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.31902827399972
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0032648108062858228,
          "train_loss": 0.005058703012764454,
          "validation_loss": 0.003942716401070356,
          "train_metric": 0.0005094940424896777,
          "validation_metric": 0.0003970960096921772,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.336301173998436
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.003347333404235542,
          "train_loss": 0.0037611848674714565,
          "validation_loss": 0.0030503952875733376,
          "train_metric": 0.00037881272146478295,
          "validation_metric": 0.0003072246618103236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3536024349996296
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004248867589922156,
          "train_loss": 0.004442738834768534,
          "validation_loss": 0.0032175909727811813,
          "train_metric": 0.00044745643390342593,
          "validation_metric": 0.0003240640216972679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3703835159994924
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.005000377132091671,
          "train_loss": 0.0031094548758119345,
          "validation_loss": 0.0022825312335044146,
          "train_metric": 0.0003131728444714099,
          "validation_metric": 0.00022988820273894817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.392485663998741
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.003343742628203472,
          "train_loss": 0.0022487544920295477,
          "validation_loss": 0.0018114466220140457,
          "train_metric": 0.00022648631420452148,
          "validation_metric": 0.00018244225066155195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.410069333998763
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0035823472535412293,
          "train_loss": 0.002789368387311697,
          "validation_loss": 0.0023474714253097773,
          "train_metric": 0.0002809349971357733,
          "validation_metric": 0.00023642870655748993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.426295602999744
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0032741659415478352,
          "train_loss": 0.0030390822794288397,
          "validation_loss": 0.002136779949069023,
          "train_metric": 0.000306085217744112,
          "validation_metric": 0.0002152086526621133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.442750841999441
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.004115030576940626,
          "train_loss": 0.0036924120504409075,
          "validation_loss": 0.0027729016728699207,
          "train_metric": 0.00037188621354289353,
          "validation_metric": 0.0002792764862533659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4595325699992827
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.00393645279109478,
          "train_loss": 0.008459090255200863,
          "validation_loss": 0.00747752794995904,
          "train_metric": 0.0008519685943610966,
          "validation_metric": 0.0007531092269346118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4763239309995697
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.006228268466657028,
          "train_loss": 0.009348686784505844,
          "validation_loss": 0.006419853772968054,
          "train_metric": 0.000941565609537065,
          "validation_metric": 0.000646584143396467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.493442575998415
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.004103624189156108,
          "train_loss": 0.003576316637918353,
          "validation_loss": 0.00284695066511631,
          "train_metric": 0.0003601935168262571,
          "validation_metric": 0.0002867344592232257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.517025337998348
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0034283755740034394,
          "train_loss": 0.005887740757316351,
          "validation_loss": 0.004913967568427324,
          "train_metric": 0.0005929917679168284,
          "validation_metric": 0.0004949168069288135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.559102658999109
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0034671610919758677,
          "train_loss": 0.002304855501279235,
          "validation_loss": 0.0016873340355232358,
          "train_metric": 0.0002321366046089679,
          "validation_metric": 0.0001699420972727239,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5839431770000374
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.002704069975152379,
          "train_loss": 0.004082697443664074,
          "validation_loss": 0.003579318756237626,
          "train_metric": 0.00041119434172287583,
          "validation_metric": 0.0003604958765208721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6008591419995355
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0031789580134500284,
          "train_loss": 0.0023760306648910046,
          "validation_loss": 0.0018197472672909498,
          "train_metric": 0.0002393051254330203,
          "validation_metric": 0.00018327828729525208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6175102759989386
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.003971660691604484,
          "train_loss": 0.0027882917784154415,
          "validation_loss": 0.0023017728235572577,
          "train_metric": 0.0002808265562634915,
          "validation_metric": 0.00023182611039374024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.634376992999023
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.003291083972726483,
          "train_loss": 0.0025270399637520313,
          "validation_loss": 0.0020886908750981092,
          "train_metric": 0.0002545142197050154,
          "validation_metric": 0.00021036529506091028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6508860679987265
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0025449169079365674,
          "train_loss": 0.0022497933823615313,
          "validation_loss": 0.001876316498965025,
          "train_metric": 0.0002265909279230982,
          "validation_metric": 0.00018897574045695364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.667672176999986
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0024972992650873493,
          "train_loss": 0.002153246197849512,
          "validation_loss": 0.001594153931364417,
          "train_metric": 0.0002168670907849446,
          "validation_metric": 0.00016055736341513693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.685711606998666
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.002246584597742185,
          "train_loss": 0.002830921672284603,
          "validation_loss": 0.002287196461111307,
          "train_metric": 0.0002851200697477907,
          "validation_metric": 0.00023035804042592645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.702498437000031
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.002593605038782698,
          "train_loss": 0.0021118426229804754,
          "validation_loss": 0.0016635372303426266,
          "train_metric": 0.0002126970503013581,
          "validation_metric": 0.00016754538228269666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7189917050000076
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0031510186217929004,
          "train_loss": 0.002624044194817543,
          "validation_loss": 0.0018387632444500923,
          "train_metric": 0.00026428408455103636,
          "validation_metric": 0.00018519350851420313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.735345477998635
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0027625510047073476,
          "train_loss": 0.0022697411477565765,
          "validation_loss": 0.001722378539852798,
          "train_metric": 0.00022860003809910268,
          "validation_metric": 0.0001734716643113643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7522880139986228
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0033160917519126087,
          "train_loss": 0.002653650473803282,
          "validation_loss": 0.002215032232925296,
          "train_metric": 0.000267265917500481,
          "validation_metric": 0.00022308994084596634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7691255239988095
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.002511357110051904,
          "train_loss": 0.0023278361186385155,
          "validation_loss": 0.0018823802238330245,
          "train_metric": 0.00023445113038178533,
          "validation_metric": 0.00018958644068334252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7857292979988415
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.003506247459881706,
          "train_loss": 0.0033895859960466623,
          "validation_loss": 0.0022423863410949707,
          "train_metric": 0.00034138665068894625,
          "validation_metric": 0.00022584492398891598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8024621280001156
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0029674919205717742,
          "train_loss": 0.008144916966557503,
          "validation_loss": 0.0066173565573990345,
          "train_metric": 0.0008203262113966048,
          "validation_metric": 0.0006664759130217135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8195928069999354
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0031998440390452743,
          "train_loss": 0.001968646189197898,
          "validation_loss": 0.0015453215455636382,
          "train_metric": 0.00019827483629342169,
          "validation_metric": 0.00015563909255433828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.837335834999976
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0021865172784600873,
          "train_loss": 0.0018129736417904496,
          "validation_loss": 0.0013912522699683905,
          "train_metric": 0.0001825960644055158,
          "validation_metric": 0.0001401218178216368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8564025629984826
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0024513482239854056,
          "train_loss": 0.002262311987578869,
          "validation_loss": 0.0019994315225631,
          "train_metric": 0.00022785179316997528,
          "validation_metric": 0.00020137542742304504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8742212309989554
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0024888959342206363,
          "train_loss": 0.0025285331066697836,
          "validation_loss": 0.0019790949299931526,
          "train_metric": 0.00025466459919698536,
          "validation_metric": 0.00019932720169890672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.891156593999767
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0022398457185772713,
          "train_loss": 0.002626333385705948,
          "validation_loss": 0.0022655182983726263,
          "train_metric": 0.0002645147033035755,
          "validation_metric": 0.0002281746856169775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9083527169987065
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.002452572782203788,
          "train_loss": 0.001836198614910245,
          "validation_loss": 0.0014690352836623788,
          "train_metric": 0.00018493521201889962,
          "validation_metric": 0.0001479558413848281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.925211308998769
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0022283794605755247,
          "train_loss": 0.002300068037584424,
          "validation_loss": 0.0016054699663072824,
          "train_metric": 0.00023165441234596074,
          "validation_metric": 0.00016169706941582263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.942206626999905
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.002103007213008823,
          "train_loss": 0.0028266573790460825,
          "validation_loss": 0.00233782222494483,
          "train_metric": 0.0002846905554179102,
          "validation_metric": 0.0002354569296585396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9590373129994987
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0024775326819508336,
          "train_loss": 0.0017721385229378939,
          "validation_loss": 0.0013533185701817274,
          "train_metric": 0.00017848331481218338,
          "validation_metric": 0.00013630131434183568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9775175859995215
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0019697380921570584,
          "train_loss": 0.0019243928836658597,
          "validation_loss": 0.0013920315541327,
          "train_metric": 0.00019381780293770134,
          "validation_metric": 0.0001402003108523786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9946894259992405
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0019699413060152438,
          "train_loss": 0.002462118398398161,
          "validation_loss": 0.0021635456942021847,
          "train_metric": 0.00024797554942779243,
          "validation_metric": 0.0002179044095100835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.011858935999044
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0026716859501902945,
          "train_loss": 0.002299377229064703,
          "validation_loss": 0.0017158512491732836,
          "train_metric": 0.0002315848832949996,
          "validation_metric": 0.00017281425243709236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0289965129995835
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0037436901147884782,
          "train_loss": 0.0029854823369532824,
          "validation_loss": 0.002388652181252837,
          "train_metric": 0.0003006868646480143,
          "validation_metric": 0.00024057630798779428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.046045448998484
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.006622522880206816,
          "train_loss": 0.00392886670306325,
          "validation_loss": 0.00383438216522336,
          "train_metric": 0.00039570106309838593,
          "validation_metric": 0.00038618495455011725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.062753937998423
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.003209460621292237,
          "train_loss": 0.003360713366419077,
          "validation_loss": 0.0022624803241342306,
          "train_metric": 0.0003384787414688617,
          "validation_metric": 0.0002278687315993011,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.081082384998808
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0035642458533402532,
          "train_loss": 0.005642406642436981,
          "validation_loss": 0.004157574847340584,
          "train_metric": 0.0005682825576514006,
          "validation_metric": 0.00041873575537465513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.097821487999681
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.002530596659198636,
          "train_loss": 0.002013087971135974,
          "validation_loss": 0.0017329694237560034,
          "train_metric": 0.00020275085989851505,
          "validation_metric": 0.0001745383342495188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.115163327000118
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0032096905233629514,
          "train_loss": 0.002447322476655245,
          "validation_loss": 0.0016677271341904998,
          "train_metric": 0.0002464853459969163,
          "validation_metric": 0.00016796737327240407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.132630528998561
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0020797251927433535,
          "train_loss": 0.0021595750004053116,
          "validation_loss": 0.0015930652152746916,
          "train_metric": 0.00021750450832769275,
          "validation_metric": 0.0001604477147338912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.149637803999212
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0021043486885901075,
          "train_loss": 0.0026218239217996597,
          "validation_loss": 0.0023277190048247576,
          "train_metric": 0.00026406050892546773,
          "validation_metric": 0.00023443932877853513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.1669523479995405
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0019778520909312647,
          "train_loss": 0.0016819211887195706,
          "validation_loss": 0.0013617869699373841,
          "train_metric": 0.0001693969388725236,
          "validation_metric": 0.00013715418754145503,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.183640455999921
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0019169856859662104,
          "train_loss": 0.0017264381749555469,
          "validation_loss": 0.0014076539082452655,
          "train_metric": 0.00017388054402545094,
          "validation_metric": 0.000141773751238361,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.200159997999435
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0017710090178297833,
          "train_loss": 0.005128386430442333,
          "validation_loss": 0.004145818296819925,
          "train_metric": 0.0005165123147889972,
          "validation_metric": 0.00041755163692869246,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.21673981099957
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.00447268095012987,
          "train_loss": 0.002005659742280841,
          "validation_loss": 0.0015149861574172974,
          "train_metric": 0.00020200270228087902,
          "validation_metric": 0.00015258387429639697,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.2339527619988075
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0027768107756855898,
          "train_loss": 0.0018368109595030546,
          "validation_loss": 0.0013070421991869807,
          "train_metric": 0.00018499685393180698,
          "validation_metric": 0.0001316404959652573,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.251160045998404
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0021696644289477263,
          "train_loss": 0.001552802394144237,
          "validation_loss": 0.0012080500600859523,
          "train_metric": 0.00015639257617294788,
          "validation_metric": 0.0001216703822137788,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.268734602999757
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0020757813726959284,
          "train_loss": 0.0017167623154819012,
          "validation_loss": 0.0013556681806221604,
          "train_metric": 0.00017290603136643767,
          "validation_metric": 0.00013653794303536415,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.285682017998624
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.002706244933506241,
          "train_loss": 0.001989746233448386,
          "validation_loss": 0.0015758444787934422,
          "train_metric": 0.0002003999543376267,
          "validation_metric": 0.0001587133010616526,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.302121130998785
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.001796101551008178,
          "train_loss": 0.002521284855902195,
          "validation_loss": 0.0020616191904991865,
          "train_metric": 0.0002539345878176391,
          "validation_metric": 0.00020763871725648642,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.318286830999568
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0017976562812691554,
          "train_loss": 0.0014691650867462158,
          "validation_loss": 0.0011078197276219726,
          "train_metric": 0.00014796895266044885,
          "validation_metric": 0.00011157555127283558,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.33493450899914
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0014899240850354545,
          "train_loss": 0.002110773231834173,
          "validation_loss": 0.001417491934262216,
          "train_metric": 0.0002125893661286682,
          "validation_metric": 0.0001427645911462605,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.452156162999017
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.00165997779549798,
          "train_loss": 0.0014537220122292638,
          "validation_loss": 0.0011650798842310905,
          "train_metric": 0.00014641357120126486,
          "validation_metric": 0.00011734259169315919,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.4694655099992815
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0015731148214399582,
          "train_loss": 0.00148296810220927,
          "validation_loss": 0.0011100837728008628,
          "train_metric": 0.00014935911167412996,
          "validation_metric": 0.00011180356523254886,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.485858834999817
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.001800666006602114,
          "train_loss": 0.0015802135458216071,
          "validation_loss": 0.0012421454302966595,
          "train_metric": 0.00015915330732241273,
          "validation_metric": 0.00012510432861745358,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.502583325998785
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0014803578415012453,
          "train_loss": 0.00138967449311167,
          "validation_loss": 0.0010469281114637852,
          "train_metric": 0.0001399629400111735,
          "validation_metric": 0.0001054427630151622,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.520600331999958
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.001903617216157727,
          "train_loss": 0.0018049192149192095,
          "validation_loss": 0.0014570783823728561,
          "train_metric": 0.00018178483878728002,
          "validation_metric": 0.00014675158308818936,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.537773874999402
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0015802379857632332,
          "train_loss": 0.0017137033864855766,
          "validation_loss": 0.0012339444365352392,
          "train_metric": 0.0001725979382172227,
          "validation_metric": 0.0001242783764610067,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.55946787099856
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0018203191630163928,
          "train_loss": 0.001959804678335786,
          "validation_loss": 0.0015370887704193592,
          "train_metric": 0.00019738433184102178,
          "validation_metric": 0.0001548099535284564,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.57652992399926
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0015734365697426256,
          "train_loss": 0.00144219771027565,
          "validation_loss": 0.0012105616042390466,
          "train_metric": 0.00014525288133881986,
          "validation_metric": 0.00012192332360427827,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.593515414999274
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0017951738664123695,
          "train_loss": 0.0018988532247021794,
          "validation_loss": 0.0015474151587113738,
          "train_metric": 0.0001912455481942743,
          "validation_metric": 0.000155849993461743,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.61028211699886
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0017054860581993125,
          "train_loss": 0.0016266872407868505,
          "validation_loss": 0.0011720348848029971,
          "train_metric": 0.00016383400361519307,
          "validation_metric": 0.00011804306268459186,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.626937369999723
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0016056856629802496,
          "train_loss": 0.0018218564800918102,
          "validation_loss": 0.0014231791719794273,
          "train_metric": 0.000183490730705671,
          "validation_metric": 0.0001433373981853947,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.643501511998693
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.002424423282718635,
          "train_loss": 0.0017286319052800536,
          "validation_loss": 0.001279994030483067,
          "train_metric": 0.000174101471202448,
          "validation_metric": 0.00012891633377876133,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.6601878619985655
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0019787890487350523,
          "train_loss": 0.0015384922735393047,
          "validation_loss": 0.0012566257501021028,
          "train_metric": 0.00015495131083298475,
          "validation_metric": 0.00012656275066547096,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.676717333999477
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.001645209356865962,
          "train_loss": 0.0014846232952550054,
          "validation_loss": 0.001063052099198103,
          "train_metric": 0.0001495258038630709,
          "validation_metric": 0.0001070667349267751,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.694250145999831
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0018308729449927341,
          "train_loss": 0.0014117571990936995,
          "validation_loss": 0.0011389418505132198,
          "train_metric": 0.00014218699652701616,
          "validation_metric": 0.00011471006291685626,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.711276282998369
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0014921554920874769,
          "train_loss": 0.0013240912230685353,
          "validation_loss": 0.0010723883751779795,
          "train_metric": 0.00013335762196220458,
          "validation_metric": 0.00010800702875712886,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.727861811999901
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0013716875764657743,
          "train_loss": 0.001337628229521215,
          "validation_loss": 0.0010845395736396313,
          "train_metric": 0.00013472102000378072,
          "validation_metric": 0.0001092308375518769,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.744141753999429
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0014140836156002479,
          "train_loss": 0.0013246799353510141,
          "validation_loss": 0.001088227960281074,
          "train_metric": 0.00013341692101676017,
          "validation_metric": 0.00010960233339574188,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.7602206299998215
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.001363284731269232,
          "train_loss": 0.0015442664735019207,
          "validation_loss": 0.001205280190333724,
          "train_metric": 0.0001555328635731712,
          "validation_metric": 0.00012139140744693577,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.776802679998582
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.001428811798177776,
          "train_loss": 0.00160607707221061,
          "validation_loss": 0.001294656191021204,
          "train_metric": 0.00016175821656361222,
          "validation_metric": 0.00013039303303230554,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.795595225999932
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.001566106620884966,
          "train_loss": 0.0016866166843101382,
          "validation_loss": 0.0013823385816067457,
          "train_metric": 0.00016986989066936076,
          "validation_metric": 0.00013922411017119884,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.812945566000053
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0014450199232669547,
          "train_loss": 0.0016379915177822113,
          "validation_loss": 0.0012143600033596158,
          "train_metric": 0.0001649725018069148,
          "validation_metric": 0.0001223058789037168,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.830545095999696
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0018245054798171623,
          "train_loss": 0.0020058900117874146,
          "validation_loss": 0.0014635028783231974,
          "train_metric": 0.00020202589803375304,
          "validation_metric": 0.00014739864855073392,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.847815194998475
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0015833011821086984,
          "train_loss": 0.0012881385628134012,
          "validation_loss": 0.0010504600359126925,
          "train_metric": 0.0001297365961363539,
          "validation_metric": 0.00010579849185887724,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 4.864516043999174
        }
      ],
      "best": {
        "epoch": 195,
        "updates": 3120
      },
      "best_validation_metric": 0.0001054427630151622,
      "config": {
        "label": "Validation plateau",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010268532661250204
    },
    "ed8dd0384733_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.2918598894029856,
          "train_loss": 0.08864232152700424,
          "validation_loss": 0.07799669355154037,
          "train_metric": 0.008927728049457073,
          "validation_metric": 0.007855541072785854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03228582800147706
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07088851742446423,
          "train_loss": 0.06481601297855377,
          "validation_loss": 0.0761234313249588,
          "train_metric": 0.006528032012283802,
          "validation_metric": 0.007666872814297676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07031641300090996
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.042878184583969414,
          "train_loss": 0.037330325692892075,
          "validation_loss": 0.03774002939462662,
          "train_metric": 0.003759773913770914,
          "validation_metric": 0.0038010375574231148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.123469775000558
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.035872515523806214,
          "train_loss": 0.03026277758181095,
          "validation_loss": 0.029498307034373283,
          "train_metric": 0.0030479563865810633,
          "validation_metric": 0.0029709613882005215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15014363800037245
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.03364978532772511,
          "train_loss": 0.04825786501169205,
          "validation_loss": 0.0387478843331337,
          "train_metric": 0.004860354587435722,
          "validation_metric": 0.0039025451987981796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17088644400064368
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.021768706443253905,
          "train_loss": 0.01536557637155056,
          "validation_loss": 0.012449638918042183,
          "train_metric": 0.0015475646359845996,
          "validation_metric": 0.001253881840966642,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19319873200038273
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.016361093265004456,
          "train_loss": 0.03012162074446678,
          "validation_loss": 0.022713040933012962,
          "train_metric": 0.003033739048987627,
          "validation_metric": 0.0022875743452459574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21446952000042074
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.015725061297416687,
          "train_loss": 0.02052845060825348,
          "validation_loss": 0.0129659753292799,
          "train_metric": 0.002067550551146269,
          "validation_metric": 0.0013058854965493083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23554145700109075
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.03412104261224158,
          "train_loss": 0.0281328447163105,
          "validation_loss": 0.02257581613957882,
          "train_metric": 0.0028334371745586395,
          "validation_metric": 0.0022737535182386637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2557168510011252
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.0245996072771959,
          "train_loss": 0.014161459170281887,
          "validation_loss": 0.01212075911462307,
          "train_metric": 0.001426290487870574,
          "validation_metric": 0.001220758305862546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27740376600013406
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.01971524575492367,
          "train_loss": 0.013194844126701355,
          "validation_loss": 0.01390079502016306,
          "train_metric": 0.0013289365451782942,
          "validation_metric": 0.0014000372029840946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29834652800127515
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.016061629517935216,
          "train_loss": 0.009087281301617622,
          "validation_loss": 0.00776623236015439,
          "train_metric": 0.0009152375860139728,
          "validation_metric": 0.000782186514697969,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3199487920010142
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.016754603464505635,
          "train_loss": 0.011239910498261452,
          "validation_loss": 0.008611350320279598,
          "train_metric": 0.0011320423800498247,
          "validation_metric": 0.0008673035772517323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34099723899998935
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.014856431516818702,
          "train_loss": 0.011729576624929905,
          "validation_loss": 0.00934813916683197,
          "train_metric": 0.0011813601013273,
          "validation_metric": 0.0009415103122591972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3620074010013923
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.022869698019349016,
          "train_loss": 0.03948310762643814,
          "validation_loss": 0.036755409091711044,
          "train_metric": 0.003976593725383282,
          "validation_metric": 0.003701870096847415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3832554180007719
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.021364092215662822,
          "train_loss": 0.018701668828725815,
          "validation_loss": 0.017138896510004997,
          "train_metric": 0.001883563818410039,
          "validation_metric": 0.0017261671600863338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4046956250003859
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.011553974603884853,
          "train_loss": 0.008492782711982727,
          "validation_loss": 0.006388961803168058,
          "train_metric": 0.0008553619845770299,
          "validation_metric": 0.000643473002128303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4268377030002739
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.011279565369477496,
          "train_loss": 0.01520159374922514,
          "validation_loss": 0.010612648911774158,
          "train_metric": 0.0015310489106923342,
          "validation_metric": 0.0010688671609386802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44811566899988975
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.013181203394196928,
          "train_loss": 0.0082817068323493,
          "validation_loss": 0.009010737761855125,
          "train_metric": 0.0008341032662428916,
          "validation_metric": 0.0009075284469872713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4686736370003928
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.008930652780691162,
          "train_loss": 0.01041437592357397,
          "validation_loss": 0.012514843605458736,
          "train_metric": 0.0010488976258784533,
          "validation_metric": 0.0012604491785168648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.490020874000038
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.007346003723796457,
          "train_loss": 0.005623728036880493,
          "validation_loss": 0.004434825852513313,
          "train_metric": 0.0005664012278430164,
          "validation_metric": 0.00044665936729870737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5126973260012164
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.01089118086383678,
          "train_loss": 0.013976730406284332,
          "validation_loss": 0.010552308522164822,
          "train_metric": 0.0014076848747208714,
          "validation_metric": 0.0010627898154780269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5336557880000328
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01286614901619032,
          "train_loss": 0.012942020781338215,
          "validation_loss": 0.01053493283689022,
          "train_metric": 0.001303472905419767,
          "validation_metric": 0.0010610398603603244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5541643420001492
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.01680598038365133,
          "train_loss": 0.007113293278962374,
          "validation_loss": 0.005545926280319691,
          "train_metric": 0.0007164248963817954,
          "validation_metric": 0.0005585654289461672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5739078990009148
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.009716997010400519,
          "train_loss": 0.007082410156726837,
          "validation_loss": 0.0067711095325648785,
          "train_metric": 0.0007133144536055624,
          "validation_metric": 0.0006819613045081496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5936396280012559
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.007690569633268751,
          "train_loss": 0.005337899550795555,
          "validation_loss": 0.004222970921546221,
          "train_metric": 0.0005376136396080256,
          "validation_metric": 0.0004253221850376576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6146855670012883
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015077051655680407,
          "train_loss": 0.024322792887687683,
          "validation_loss": 0.020289147272706032,
          "train_metric": 0.0024497027043253183,
          "validation_metric": 0.0020434486214071512,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6347034410009655
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01614713718299754,
          "train_loss": 0.010776746086776257,
          "validation_loss": 0.009896081872284412,
          "train_metric": 0.0010853944113478065,
          "validation_metric": 0.0009966972284018993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6548360360011429
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.009654254477936774,
          "train_loss": 0.01019070390611887,
          "validation_loss": 0.007867228239774704,
          "train_metric": 0.0010263704461976886,
          "validation_metric": 0.0007923583616502583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.675355663001028
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.009363326942548156,
          "train_loss": 0.007015187293291092,
          "validation_loss": 0.005033033899962902,
          "train_metric": 0.0007065440295264125,
          "validation_metric": 0.0005069086910225451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6955724130002636
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.01186647597933188,
          "train_loss": 0.036512408405542374,
          "validation_loss": 0.031734712421894073,
          "train_metric": 0.0036773961037397385,
          "validation_metric": 0.0031962040811777115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7157670010001311
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01852813793811947,
          "train_loss": 0.00820240005850792,
          "validation_loss": 0.007192524150013924,
          "train_metric": 0.0008261156617663801,
          "validation_metric": 0.0007244047010317445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7360401540008752
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.008491960179526359,
          "train_loss": 0.005819373764097691,
          "validation_loss": 0.004808956757187843,
          "train_metric": 0.0005861060344614089,
          "validation_metric": 0.00048434047494083643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7565444060001028
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.008550145721528679,
          "train_loss": 0.007180631626397371,
          "validation_loss": 0.00788583792746067,
          "train_metric": 0.0007232069037854671,
          "validation_metric": 0.000794232648331672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7764913790015271
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.009529343107715249,
          "train_loss": 0.013074944727122784,
          "validation_loss": 0.012273103930056095,
          "train_metric": 0.0013168605510145426,
          "validation_metric": 0.0012361020781099796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7968266790012422
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.026141693306271918,
          "train_loss": 0.037616487592458725,
          "validation_loss": 0.03295835480093956,
          "train_metric": 0.0037885946221649647,
          "validation_metric": 0.0033194448333233595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8170029770008114
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.01746085361810401,
          "train_loss": 0.013114810921251774,
          "validation_loss": 0.012586081400513649,
          "train_metric": 0.0013208757154643536,
          "validation_metric": 0.0012676240876317024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8500126330000057
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.010310552082955837,
          "train_loss": 0.006344502326101065,
          "validation_loss": 0.00621482590213418,
          "train_metric": 0.0006389950285665691,
          "validation_metric": 0.000625934568233788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9027827780009829
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.008191051034373231,
          "train_loss": 0.007911169901490211,
          "validation_loss": 0.006951791699975729,
          "train_metric": 0.0007967841811478138,
          "validation_metric": 0.0007001590565778315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9552587260004657
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.007784994799294509,
          "train_loss": 0.005339174065738916,
          "validation_loss": 0.005317959934473038,
          "train_metric": 0.0005377421039156616,
          "validation_metric": 0.0005356054753065109,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0093690730009257
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.004775989422341809,
          "train_loss": 0.004657363053411245,
          "validation_loss": 0.004882315639406443,
          "train_metric": 0.00046907257637940347,
          "validation_metric": 0.0004917289479635656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0626729550003802
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.006341782660456374,
          "train_loss": 0.008563660085201263,
          "validation_loss": 0.009967401623725891,
          "train_metric": 0.0008625004556961358,
          "validation_metric": 0.0010038800537586212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1086046760010504
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.00553731486434117,
          "train_loss": 0.004868757911026478,
          "validation_loss": 0.0032199930865317583,
          "train_metric": 0.0004903634544461966,
          "validation_metric": 0.0003243059036321938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2209962530014309
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.0040575927050667815,
          "train_loss": 0.002618771977722645,
          "validation_loss": 0.00262870523147285,
          "train_metric": 0.0002637531142681837,
          "validation_metric": 0.0002647535875439644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2996058280004945
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.008529115926648956,
          "train_loss": 0.008327375166118145,
          "validation_loss": 0.007731289137154818,
          "train_metric": 0.000838702660985291,
          "validation_metric": 0.0007786671631038189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.375926911001443
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.016755570097302552,
          "train_loss": 0.007430089637637138,
          "validation_loss": 0.005150164477527142,
          "train_metric": 0.0007483314257115126,
          "validation_metric": 0.0005187056958675385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4718903780012624
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.008577464468544349,
          "train_loss": 0.008080951869487762,
          "validation_loss": 0.005349951330572367,
          "train_metric": 0.000813883903902024,
          "validation_metric": 0.000538827502168715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5427294580003945
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0076643683278234676,
          "train_loss": 0.0079240333288908,
          "validation_loss": 0.007038249168545008,
          "train_metric": 0.0007980796508491039,
          "validation_metric": 0.0007088666898198426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6146606640013488
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00819114911428187,
          "train_loss": 0.005814986303448677,
          "validation_loss": 0.004741787910461426,
          "train_metric": 0.0005856640636920929,
          "validation_metric": 0.0004775755514856428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6860659009998926
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.005692969571100548,
          "train_loss": 0.0056963833048939705,
          "validation_loss": 0.0038059272337704897,
          "train_metric": 0.0005737188621424139,
          "validation_metric": 0.00038331907126121223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7588647280008445
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.006766240592696704,
          "train_loss": 0.004785281606018543,
          "validation_loss": 0.0033132622484117746,
          "train_metric": 0.00048195605631917715,
          "validation_metric": 0.00033369965967722237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8308895059999486
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0077474462013924494,
          "train_loss": 0.007025204598903656,
          "validation_loss": 0.005655215587466955,
          "train_metric": 0.0007075528847053647,
          "validation_metric": 0.0005695726140402257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9036207380013366
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.004530600315774791,
          "train_loss": 0.004373990930616856,
          "validation_loss": 0.004090127535164356,
          "train_metric": 0.0004405323415994644,
          "validation_metric": 0.00041194268851540983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9783862210006191
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.003984390397818061,
          "train_loss": 0.005259898025542498,
          "validation_loss": 0.003966358490288258,
          "train_metric": 0.0005297576426528394,
          "validation_metric": 0.00039947708137333393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0513068240015855
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008159240722307004,
          "train_loss": 0.012475830502808094,
          "validation_loss": 0.009567925706505775,
          "train_metric": 0.0012565198121592402,
          "validation_metric": 0.0009636465110816061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.125086830001237
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0086702328917454,
          "train_loss": 0.005426165647804737,
          "validation_loss": 0.0046656448394060135,
          "train_metric": 0.0005465035210363567,
          "validation_metric": 0.00046990669216029346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.200564843000393
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.005268391520075966,
          "train_loss": 0.0031222470570355654,
          "validation_loss": 0.0027326736599206924,
          "train_metric": 0.0003144612710457295,
          "validation_metric": 0.0002752248547039926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2751736890004395
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0030489538730762433,
          "train_loss": 0.0024803176056593657,
          "validation_loss": 0.001925042481161654,
          "train_metric": 0.00024980847956612706,
          "validation_metric": 0.0001938832429004833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.35390532800011
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.0023582019784953445,
          "train_loss": 0.0018314425833523273,
          "validation_loss": 0.0015462005976587534,
          "train_metric": 0.000184456177521497,
          "validation_metric": 0.0001557276409585029,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4299075880007877
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0020975926963728853,
          "train_loss": 0.0016878442838788033,
          "validation_loss": 0.001576113048940897,
          "train_metric": 0.0001699934946373105,
          "validation_metric": 0.00015874032396823168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.505655791001118
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.005168965901248157,
          "train_loss": 0.004011417273432016,
          "validation_loss": 0.0026179312262684107,
          "train_metric": 0.0004040152416564524,
          "validation_metric": 0.00026366845122538507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5947030970000924
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.003346200566738844,
          "train_loss": 0.007942265830934048,
          "validation_loss": 0.007680480368435383,
          "train_metric": 0.0007999158697202802,
          "validation_metric": 0.0007735498948022723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6766226370000368
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.007118845882359892,
          "train_loss": 0.004203805234283209,
          "validation_loss": 0.0035097904037684202,
          "train_metric": 0.0004233918443787843,
          "validation_metric": 0.0003534932329785079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.750508468001499
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.008065311383688822,
          "train_loss": 0.004836522974073887,
          "validation_loss": 0.005169498734176159,
          "train_metric": 0.00048711683484725654,
          "validation_metric": 0.0005206529167480767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8236119339999277
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0052472300885710865,
          "train_loss": 0.004214893560856581,
          "validation_loss": 0.003631205763667822,
          "train_metric": 0.00042450864566490054,
          "validation_metric": 0.0003657217603176832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8967439090010885
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.003399362372874748,
          "train_loss": 0.005308693740516901,
          "validation_loss": 0.005134794395416975,
          "train_metric": 0.0005346721736714244,
          "validation_metric": 0.0005171576631255448,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.969749784000669
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.003515738993883133,
          "train_loss": 0.00564270094037056,
          "validation_loss": 0.004527469165623188,
          "train_metric": 0.0005683121853508055,
          "validation_metric": 0.00045599014265462756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0427299840011983
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0031081880079000257,
          "train_loss": 0.003518682671710849,
          "validation_loss": 0.0036431862972676754,
          "train_metric": 0.00035438884515315294,
          "validation_metric": 0.00036692843423224986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1156338970013167
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.004939240978274029,
          "train_loss": 0.010351692326366901,
          "validation_loss": 0.008570339530706406,
          "train_metric": 0.0010425844229757786,
          "validation_metric": 0.0008631732780486345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1966952940001647
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006802876785513945,
          "train_loss": 0.0036110924556851387,
          "validation_loss": 0.003055431880056858,
          "train_metric": 0.0003636960464064032,
          "validation_metric": 0.00030773188336752355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2734885050012963
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.00389248228748329,
          "train_loss": 0.018107932060956955,
          "validation_loss": 0.0137685751542449,
          "train_metric": 0.0018237646436318755,
          "validation_metric": 0.0013867204543203115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.347912715000348
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.012503833830123767,
          "train_loss": 0.011462690308690071,
          "validation_loss": 0.010331950150430202,
          "train_metric": 0.0011544802691787481,
          "validation_metric": 0.0010405961656942964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4221535209999274
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.007826071305316873,
          "train_loss": 0.005536403972655535,
          "validation_loss": 0.005294628441333771,
          "train_metric": 0.000557606341317296,
          "validation_metric": 0.0005332555156201124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4948516479998943
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005640523639158346,
          "train_loss": 0.005221980158239603,
          "validation_loss": 0.0039605251513421535,
          "train_metric": 0.0005259387544356287,
          "validation_metric": 0.0003988895914517343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5693037979999644
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0042820946327992715,
          "train_loss": 0.002675869269296527,
          "validation_loss": 0.001717344275675714,
          "train_metric": 0.0002695037692319602,
          "validation_metric": 0.0001729646319290623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.642818497000917
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.008014508886844851,
          "train_loss": 0.007258978206664324,
          "validation_loss": 0.005191306117922068,
          "train_metric": 0.0007310977089218795,
          "validation_metric": 0.0005228493246249855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7652029740002035
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.009741475179907866,
          "train_loss": 0.012981198728084564,
          "validation_loss": 0.013232463039457798,
          "train_metric": 0.001307418686337769,
          "validation_metric": 0.0013327252818271518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.846809481001401
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.010571753096883185,
          "train_loss": 0.013879254460334778,
          "validation_loss": 0.01513265073299408,
          "train_metric": 0.0013978676870465279,
          "validation_metric": 0.0015241052024066448,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9220307499999763
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006359635735861957,
          "train_loss": 0.0031826281920075417,
          "validation_loss": 0.0024456093087792397,
          "train_metric": 0.0003205426037311554,
          "validation_metric": 0.0002463127893861383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9962272150005447
        }
      ],
      "best": {
        "epoch": 59,
        "updates": 944
      },
      "best_validation_metric": 0.0001557276409585029,
      "config": {
        "label": "LayerNorm",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012479088146114798
    },
    "ed8dd0384733_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.23888491606339812,
          "train_loss": 0.08166317641735077,
          "validation_loss": 0.06555181741714478,
          "train_metric": 0.00822481606155634,
          "validation_metric": 0.0066021401435136795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17380248900008155
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.061784587102010846,
          "train_loss": 0.054036371409893036,
          "validation_loss": 0.045189712196588516,
          "train_metric": 0.005442344583570957,
          "validation_metric": 0.004551341757178307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2525411290007469
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.04121296224184334,
          "train_loss": 0.03482135012745857,
          "validation_loss": 0.027817262336611748,
          "train_metric": 0.003507079090923071,
          "validation_metric": 0.0028016529977321625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33128448900060903
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.029260389623232186,
          "train_loss": 0.02433099038898945,
          "validation_loss": 0.01896427944302559,
          "train_metric": 0.002450528321787715,
          "validation_metric": 0.001910012448206544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40873076599928027
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.024199075298383832,
          "train_loss": 0.02122872695326805,
          "validation_loss": 0.021520046517252922,
          "train_metric": 0.0021380793768912554,
          "validation_metric": 0.0021674202289432287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5071424020006816
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.021335106168407947,
          "train_loss": 0.017451364547014236,
          "validation_loss": 0.01547304354608059,
          "train_metric": 0.0017576378304511309,
          "validation_metric": 0.0015583883505314589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.628528491999532
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.02976817957824096,
          "train_loss": 0.03284674137830734,
          "validation_loss": 0.034943778067827225,
          "train_metric": 0.0033082030713558197,
          "validation_metric": 0.0035194093361496925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7048871150000195
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.02878994884667918,
          "train_loss": 0.02708299085497856,
          "validation_loss": 0.019958991557359695,
          "train_metric": 0.002727699698880315,
          "validation_metric": 0.002010196214541793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8133077959992079
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.016897937923204154,
          "train_loss": 0.01903846114873886,
          "validation_loss": 0.01356436312198639,
          "train_metric": 0.0019174840999767184,
          "validation_metric": 0.0013661530101671815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8888565199995355
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.014107360038906336,
          "train_loss": 0.011092609725892544,
          "validation_loss": 0.009211700409650803,
          "train_metric": 0.001117207109928131,
          "validation_metric": 0.0009277687640860677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9635768820007797
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.011751949205063283,
          "train_loss": 0.008849341422319412,
          "validation_loss": 0.008775648660957813,
          "train_metric": 0.0008912733173929155,
          "validation_metric": 0.0008838512003421783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.040081907000058
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.010980227962136269,
          "train_loss": 0.025530699640512466,
          "validation_loss": 0.02131485566496849,
          "train_metric": 0.0025713585782796144,
          "validation_metric": 0.0021467541810125113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1124954919996526
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.02549975086003542,
          "train_loss": 0.01695833168923855,
          "validation_loss": 0.01574423722922802,
          "train_metric": 0.0017079809913411736,
          "validation_metric": 0.0015857020625844598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1847255699995003
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.019903491061995737,
          "train_loss": 0.013715945184230804,
          "validation_loss": 0.014348994940519333,
          "train_metric": 0.0013814197154715657,
          "validation_metric": 0.0014451781753450632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2705685250002716
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.013678477756911889,
          "train_loss": 0.019475849345326424,
          "validation_loss": 0.013384152203798294,
          "train_metric": 0.0019615362398326397,
          "validation_metric": 0.001348002697341144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3466772259998834
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.011750767356716096,
          "train_loss": 0.011588453315198421,
          "validation_loss": 0.008362028747797012,
          "train_metric": 0.0011671464890241623,
          "validation_metric": 0.0008421927923336625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4237922669999534
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.00787850258348044,
          "train_loss": 0.007302626967430115,
          "validation_loss": 0.0058487593196332455,
          "train_metric": 0.0007354938425123692,
          "validation_metric": 0.000589065661188215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.515732231999209
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.00889447663212195,
          "train_loss": 0.006209243554621935,
          "validation_loss": 0.004950473550707102,
          "train_metric": 0.0006253722822293639,
          "validation_metric": 0.0004985934938304126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5389745810007298
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.011325097439112142,
          "train_loss": 0.0055108158849179745,
          "validation_loss": 0.004341572523117065,
          "train_metric": 0.0005550291971303523,
          "validation_metric": 0.00043726732837967575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5676235699993413
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.007683465169975534,
          "train_loss": 0.004882046952843666,
          "validation_loss": 0.004007525276392698,
          "train_metric": 0.0004917018231935799,
          "validation_metric": 0.0004036233003716916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5834654129994306
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.008050136239035055,
          "train_loss": 0.005788047332316637,
          "validation_loss": 0.004216156434267759,
          "train_metric": 0.0005829509464092553,
          "validation_metric": 0.0004246358585078269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6059404659990832
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.01719522225903347,
          "train_loss": 0.023246873170137405,
          "validation_loss": 0.01762869581580162,
          "train_metric": 0.0023413393646478653,
          "validation_metric": 0.0017754975706338882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6205022060003103
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.02437688660575077,
          "train_loss": 0.009796607308089733,
          "validation_loss": 0.009899741038680077,
          "train_metric": 0.0009866784093901515,
          "validation_metric": 0.0009970655664801598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6383688839996466
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.01053882839914877,
          "train_loss": 0.006641744635999203,
          "validation_loss": 0.005747511517256498,
          "train_metric": 0.00066893215989694,
          "validation_metric": 0.0005788682610727847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6541411700000026
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.010213554996880703,
          "train_loss": 0.007118855603039265,
          "validation_loss": 0.007156596519052982,
          "train_metric": 0.0007169851451180875,
          "validation_metric": 0.0007207861635833979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.670058063000397
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.008563331473851576,
          "train_loss": 0.005285867024213076,
          "validation_loss": 0.004842318128794432,
          "train_metric": 0.0005323732038959861,
          "validation_metric": 0.0004877005412708968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6965123650006717
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.012568902275233995,
          "train_loss": 0.022204900160431862,
          "validation_loss": 0.018985990434885025,
          "train_metric": 0.0022363958414644003,
          "validation_metric": 0.0019121996592730284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.711626977999913
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01679351064376533,
          "train_loss": 0.006613657809793949,
          "validation_loss": 0.005368424113839865,
          "train_metric": 0.0006661033257842064,
          "validation_metric": 0.0005406880518421531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7260322559995984
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.007585612416733056,
          "train_loss": 0.007600780110806227,
          "validation_loss": 0.007061448413878679,
          "train_metric": 0.0007655228255316615,
          "validation_metric": 0.000711203261744231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7405371250006283
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.013993913438753225,
          "train_loss": 0.01605561003088951,
          "validation_loss": 0.010954132303595543,
          "train_metric": 0.001617062371224165,
          "validation_metric": 0.001103260088711977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7539102230002754
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008528002261300571,
          "train_loss": 0.006476330105215311,
          "validation_loss": 0.006094194017350674,
          "train_metric": 0.0006522722542285919,
          "validation_metric": 0.00061378488317132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7681694669990975
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.00967037225200329,
          "train_loss": 0.011177649721503258,
          "validation_loss": 0.010473490692675114,
          "train_metric": 0.0011257717851549387,
          "validation_metric": 0.00105485157109797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7839025510002102
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.00849347391340416,
          "train_loss": 0.020055709406733513,
          "validation_loss": 0.013946794904768467,
          "train_metric": 0.002019937615841627,
          "validation_metric": 0.001404670299962163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8006199409992405
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.030739209207240492,
          "train_loss": 0.027002345770597458,
          "validation_loss": 0.028418412432074547,
          "train_metric": 0.002719577169045806,
          "validation_metric": 0.0028621985111385584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8177813210004388
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.020379200344905257,
          "train_loss": 0.011778450571000576,
          "validation_loss": 0.009860259480774403,
          "train_metric": 0.0011862823739647865,
          "validation_metric": 0.000993089284747839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8341987330004486
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.012796510636690073,
          "train_loss": 0.009276814758777618,
          "validation_loss": 0.006051214877516031,
          "train_metric": 0.0009343268466182053,
          "validation_metric": 0.0006094561540521681,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8494463740007632
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.009710397309390828,
          "train_loss": 0.005910638254135847,
          "validation_loss": 0.006261457223445177,
          "train_metric": 0.0005952978390268981,
          "validation_metric": 0.0006306310533545911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8649084379994747
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.006474832509411499,
          "train_loss": 0.004515328444540501,
          "validation_loss": 0.003883256111294031,
          "train_metric": 0.00045476731611415744,
          "validation_metric": 0.00039110734360292554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8800202679994982
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.00824815449595917,
          "train_loss": 0.014396575279533863,
          "validation_loss": 0.011599487625062466,
          "train_metric": 0.0014499702956527472,
          "validation_metric": 0.001168257906101644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.896743499999502
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.014771834088605829,
          "train_loss": 0.013458059169352055,
          "validation_loss": 0.009508093819022179,
          "train_metric": 0.0013554464094340801,
          "validation_metric": 0.0009576204465702176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9115449559994886
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.008217255948693492,
          "train_loss": 0.004578107967972755,
          "validation_loss": 0.003996424376964569,
          "train_metric": 0.0004610902979038656,
          "validation_metric": 0.00040250521851703525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.926331972999833
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.00798505009879591,
          "train_loss": 0.0034405305050313473,
          "validation_loss": 0.0031360085122287273,
          "train_metric": 0.00034651762689463794,
          "validation_metric": 0.00031584728276357055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9415819219993864
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.01389157357334625,
          "train_loss": 0.007949343882501125,
          "validation_loss": 0.0055680060759186745,
          "train_metric": 0.0008006288553588092,
          "validation_metric": 0.0005607891362160444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.957335349999994
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.010077779748826288,
          "train_loss": 0.01106900442391634,
          "validation_loss": 0.01007043570280075,
          "train_metric": 0.0011148296762257814,
          "validation_metric": 0.0010142575483769178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9748463619998802
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.006897451472468674,
          "train_loss": 0.006545836105942726,
          "validation_loss": 0.006712344475090504,
          "train_metric": 0.000659272656776011,
          "validation_metric": 0.0006760427495464683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.003054888000406
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.005230337163084187,
          "train_loss": 0.003981802146881819,
          "validation_loss": 0.0030425931327044964,
          "train_metric": 0.00040103247738443315,
          "validation_metric": 0.0003064388583879918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0809228050002275
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.005196581478230655,
          "train_loss": 0.0028102793730795383,
          "validation_loss": 0.0024695247411727905,
          "train_metric": 0.00028304100851528347,
          "validation_metric": 0.00024872150970622897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.157994773999235
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.003838059456029441,
          "train_loss": 0.008629830554127693,
          "validation_loss": 0.006473086774349213,
          "train_metric": 0.0008691649418324232,
          "validation_metric": 0.0006519456510432065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.230817350000507
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.007640847092261538,
          "train_loss": 0.005728396121412516,
          "validation_loss": 0.004635212477296591,
          "train_metric": 0.0005769430426880717,
          "validation_metric": 0.0004668415931519121,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3049417969996284
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.006554884355864488,
          "train_loss": 0.002977886702865362,
          "validation_loss": 0.002342517487704754,
          "train_metric": 0.00029992181225679815,
          "validation_metric": 0.00023592979414388537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3804444239995064
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.004526794276898727,
          "train_loss": 0.006613605655729771,
          "validation_loss": 0.005400865338742733,
          "train_metric": 0.000666098203510046,
          "validation_metric": 0.0005439553642645478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6443058469994867
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0055393806542269886,
          "train_loss": 0.003952159080654383,
          "validation_loss": 0.002931263530626893,
          "train_metric": 0.0003980470064561814,
          "validation_metric": 0.00029522611293941736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.718625544999668
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.00907906961219851,
          "train_loss": 0.005236187018454075,
          "validation_loss": 0.003230944275856018,
          "train_metric": 0.0005273696733638644,
          "validation_metric": 0.00032540890970267355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7949384479998116
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.004642241474357434,
          "train_loss": 0.003771595424041152,
          "validation_loss": 0.003527394961565733,
          "train_metric": 0.00037986127426847816,
          "validation_metric": 0.000355266296537593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.867548063999493
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.0034833433564926963,
          "train_loss": 0.009802650660276413,
          "validation_loss": 0.00770798372104764,
          "train_metric": 0.0009872870286926627,
          "validation_metric": 0.0007763198809698224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.942381299000772
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0034438854672771413,
          "train_loss": 0.002093903487548232,
          "validation_loss": 0.001566996332257986,
          "train_metric": 0.0002108902990585193,
          "validation_metric": 0.00015782212722115219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.018826199999239
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.01263210390607128,
          "train_loss": 0.01482412125915289,
          "validation_loss": 0.017079167068004608,
          "train_metric": 0.0014930310426279902,
          "validation_metric": 0.001720151281915605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.113649502000044
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.014118471270194277,
          "train_loss": 0.021743040531873703,
          "validation_loss": 0.013790354132652283,
          "train_metric": 0.0021898793056607246,
          "validation_metric": 0.001388913718983531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.190955756999756
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.021053782664239407,
          "train_loss": 0.015445223078131676,
          "validation_loss": 0.01123218983411789,
          "train_metric": 0.0015555863501504064,
          "validation_metric": 0.0011312650749459863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.26684529799968
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.010374348625191487,
          "train_loss": 0.008734960108995438,
          "validation_loss": 0.008342346176505089,
          "train_metric": 0.0008797532063908875,
          "validation_metric": 0.0008402105886489153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.33592770299947
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.006581236171768978,
          "train_loss": 0.005572620779275894,
          "validation_loss": 0.003957104403525591,
          "train_metric": 0.0005612539243884385,
          "validation_metric": 0.00039854508941061795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3541983100003563
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.005923885270021856,
          "train_loss": 0.004924330394715071,
          "validation_loss": 0.0036511963699012995,
          "train_metric": 0.0004959605284966528,
          "validation_metric": 0.0003677351342048496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.370244621999518
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.016346898781193886,
          "train_loss": 0.04313359782099724,
          "validation_loss": 0.03367244824767113,
          "train_metric": 0.004344257991760969,
          "validation_metric": 0.0033913657534867525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3865345510002953
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.02645565348211676,
          "train_loss": 0.02120470628142357,
          "validation_loss": 0.014530458487570286,
          "train_metric": 0.0021356604993343353,
          "validation_metric": 0.0014634545659646392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4023510450006142
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.012640562345040962,
          "train_loss": 0.010627543553709984,
          "validation_loss": 0.010202363133430481,
          "train_metric": 0.001070367288775742,
          "validation_metric": 0.0010275447275489569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4180709699994622
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.00882364803692326,
          "train_loss": 0.006395716220140457,
          "validation_loss": 0.005956361070275307,
          "train_metric": 0.000644153100438416,
          "validation_metric": 0.0005999028217047453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.434026074999565
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.005252666207525181,
          "train_loss": 0.003907275386154652,
          "validation_loss": 0.003601963398978114,
          "train_metric": 0.0003935264830943197,
          "validation_metric": 0.00036277653998695314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4498385040005815
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00543970603030175,
          "train_loss": 0.005815334618091583,
          "validation_loss": 0.004776843823492527,
          "train_metric": 0.0005856992211192846,
          "validation_metric": 0.000481106253573671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4656637240004784
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0037035261702840216,
          "train_loss": 0.0028683566488325596,
          "validation_loss": 0.0025982754305005074,
          "train_metric": 0.0002888903836719692,
          "validation_metric": 0.00026168875047005713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4814078950003022
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.004881093012954807,
          "train_loss": 0.0034390990622341633,
          "validation_loss": 0.0029140969272702932,
          "train_metric": 0.0003463734174147248,
          "validation_metric": 0.00029349717078730464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.498534581000058
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.002359683541726554,
          "train_loss": 0.002168570877984166,
          "validation_loss": 0.001969740493223071,
          "train_metric": 0.00021841051056981087,
          "validation_metric": 0.00019838505249936134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.514454198000749
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00273386581466184,
          "train_loss": 0.008274690248072147,
          "validation_loss": 0.006647927686572075,
          "train_metric": 0.0008333965670317411,
          "validation_metric": 0.0006695549236610532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5306286539998837
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006147770400275476,
          "train_loss": 0.01349804736673832,
          "validation_loss": 0.009144273586571217,
          "train_metric": 0.0013594739139080048,
          "validation_metric": 0.0009209776762872934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.545675109999138
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.007349888299359009,
          "train_loss": 0.0067305658012628555,
          "validation_loss": 0.004965608473867178,
          "train_metric": 0.0006778779206797481,
          "validation_metric": 0.0005001179524697363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5644083229999524
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.005142083206010284,
          "train_loss": 0.0044827088713645935,
          "validation_loss": 0.003887706669047475,
          "train_metric": 0.0004514820175245404,
          "validation_metric": 0.00039155554259195924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5806202399999165
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.006381291153957136,
          "train_loss": 0.02036585658788681,
          "validation_loss": 0.014788160100579262,
          "train_metric": 0.002051174407824874,
          "validation_metric": 0.0014894092455506325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5998287099992012
        }
      ],
      "best": {
        "epoch": 56,
        "updates": 896
      },
      "best_validation_metric": 0.00015782212722115219,
      "config": {
        "label": "LayerNorm",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012562727698280823
    },
    "c160e9485e77_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.540454703848809,
          "train_loss": 0.22924257814884186,
          "validation_loss": 0.16837787628173828,
          "train_metric": 0.023088471964001656,
          "validation_metric": 0.01695840246975422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02422580299935362
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.18825210770592093,
          "train_loss": 0.1465892642736435,
          "validation_loss": 0.10817298293113708,
          "train_metric": 0.01476393360644579,
          "validation_metric": 0.01089478563517332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.053174462998867966
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12770230253227055,
          "train_loss": 0.10497328639030457,
          "validation_loss": 0.088044673204422,
          "train_metric": 0.010572525672614574,
          "validation_metric": 0.00886753760278225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08444798899836314
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09814765932969749,
          "train_loss": 0.08512938767671585,
          "validation_loss": 0.07042066007852554,
          "train_metric": 0.008573921397328377,
          "validation_metric": 0.007092510815709829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11170270999900822
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07805201376322657,
          "train_loss": 0.07071519643068314,
          "validation_loss": 0.05630437657237053,
          "train_metric": 0.0071221753023564816,
          "validation_metric": 0.00567077100276947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13916993899874797
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06752244534436613,
          "train_loss": 0.06566234678030014,
          "validation_loss": 0.05456046387553215,
          "train_metric": 0.006613271310925484,
          "validation_metric": 0.005495130084455013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1640757789991767
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.0646860096603632,
          "train_loss": 0.060111358761787415,
          "validation_loss": 0.049640312790870667,
          "train_metric": 0.006054195575416088,
          "validation_metric": 0.0049995905719697475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1885405159991933
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.0583970588631928,
          "train_loss": 0.056434664875268936,
          "validation_loss": 0.045372191816568375,
          "train_metric": 0.005683892872184515,
          "validation_metric": 0.004569720942527056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21670389700011583
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05694839800707996,
          "train_loss": 0.05342099070549011,
          "validation_loss": 0.044930510222911835,
          "train_metric": 0.005380365997552872,
          "validation_metric": 0.00452523585408926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23688217800008715
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.054414439713582397,
          "train_loss": 0.05020705237984657,
          "validation_loss": 0.041989028453826904,
          "train_metric": 0.005056669935584068,
          "validation_metric": 0.004228981677442789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24982457699843508
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0499857027316466,
          "train_loss": 0.04708756133913994,
          "validation_loss": 0.03956771269440651,
          "train_metric": 0.004742486868053675,
          "validation_metric": 0.003985114861279726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34066289399925154
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.0480853111948818,
          "train_loss": 0.04662153497338295,
          "validation_loss": 0.03896695375442505,
          "train_metric": 0.004695550072938204,
          "validation_metric": 0.003924609161913395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35348704799980624
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.04583198030013591,
          "train_loss": 0.04596062749624252,
          "validation_loss": 0.0399930477142334,
          "train_metric": 0.004628985188901424,
          "validation_metric": 0.0040279533714056015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3656206769992423
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.041924821329303086,
          "train_loss": 0.04063747450709343,
          "validation_loss": 0.0352192260324955,
          "train_metric": 0.004092857241630554,
          "validation_metric": 0.003547151107341051,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.379123945000174
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.041683239105623215,
          "train_loss": 0.03620542213320732,
          "validation_loss": 0.031074965372681618,
          "train_metric": 0.003646478056907654,
          "validation_metric": 0.0031297565437853336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3927157019988954
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.03565680829342455,
          "train_loss": 0.03456662967801094,
          "validation_loss": 0.03019879199564457,
          "train_metric": 0.0034814239479601383,
          "validation_metric": 0.0030415114015340805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4054806770000141
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03371906402753666,
          "train_loss": 0.03171124681830406,
          "validation_loss": 0.026567986235022545,
          "train_metric": 0.0031938403844833374,
          "validation_metric": 0.0026758299209177494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41817566299869213
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.03280753968283534,
          "train_loss": 0.027701739221811295,
          "validation_loss": 0.023050375282764435,
          "train_metric": 0.0027900179848074913,
          "validation_metric": 0.002321549691259861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4306800939993991
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.02811444055987522,
          "train_loss": 0.02731950394809246,
          "validation_loss": 0.02376052550971508,
          "train_metric": 0.0027515203692018986,
          "validation_metric": 0.002393072936683893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4426494149993232
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.024674008163856342,
          "train_loss": 0.022642597556114197,
          "validation_loss": 0.019582772627472878,
          "train_metric": 0.0022804795298725367,
          "validation_metric": 0.001972305355593562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.455098452999664
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.02404786489205435,
          "train_loss": 0.0208602137863636,
          "validation_loss": 0.017727727070450783,
          "train_metric": 0.0021009643096476793,
          "validation_metric": 0.0017854714533314109,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46761313599927234
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.020010972337331623,
          "train_loss": 0.01928883045911789,
          "validation_loss": 0.016313089057803154,
          "train_metric": 0.0019427004735916853,
          "validation_metric": 0.0016429946990683675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4802835359987512
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.018286925565917045,
          "train_loss": 0.019518526270985603,
          "validation_loss": 0.01710180938243866,
          "train_metric": 0.0019658340606838465,
          "validation_metric": 0.001722431741654873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49261528799979715
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.020361589835374616,
          "train_loss": 0.022084761410951614,
          "validation_loss": 0.01883300393819809,
          "train_metric": 0.0022242963314056396,
          "validation_metric": 0.0018967912765219808,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5041598449988669
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.02712810511002317,
          "train_loss": 0.021443234756588936,
          "validation_loss": 0.017770525068044662,
          "train_metric": 0.0021596839651465416,
          "validation_metric": 0.0017897821962833405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5162803889998031
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.018145493435440585,
          "train_loss": 0.018635449931025505,
          "validation_loss": 0.015294084325432777,
          "train_metric": 0.001876894268207252,
          "validation_metric": 0.0015403639990836382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5287721499989857
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01728524296777323,
          "train_loss": 0.013871625065803528,
          "validation_loss": 0.011865263804793358,
          "train_metric": 0.00139709934592247,
          "validation_metric": 0.0011950259795412421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5416866979994666
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01399575377581641,
          "train_loss": 0.012756173498928547,
          "validation_loss": 0.010706143453717232,
          "train_metric": 0.001284755184315145,
          "validation_metric": 0.0010782835306599736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5547530759995425
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.015203476272290573,
          "train_loss": 0.0134615832939744,
          "validation_loss": 0.011268102563917637,
          "train_metric": 0.0013558012433350086,
          "validation_metric": 0.0011348819825798273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5669181339999341
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.013635380513733253,
          "train_loss": 0.019247857853770256,
          "validation_loss": 0.015104220248758793,
          "train_metric": 0.001938573899678886,
          "validation_metric": 0.001521241501905024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5796329819986568
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.01627618321799673,
          "train_loss": 0.012054077349603176,
          "validation_loss": 0.010137046687304974,
          "train_metric": 0.0012140425387769938,
          "validation_metric": 0.0010209663305431604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5921065879992966
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.013218465974205174,
          "train_loss": 0.01171859260648489,
          "validation_loss": 0.010149524547159672,
          "train_metric": 0.0011802536901086569,
          "validation_metric": 0.001022223150357604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6051176879991544
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.014556173846358433,
          "train_loss": 0.01383502408862114,
          "validation_loss": 0.010857885703444481,
          "train_metric": 0.0013934130547568202,
          "validation_metric": 0.0010935665341094136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6171932950001064
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01456796434649732,
          "train_loss": 0.012345559895038605,
          "validation_loss": 0.010764048434793949,
          "train_metric": 0.0012433996889740229,
          "validation_metric": 0.0010841155890375376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6289992009988055
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.011629346437985078,
          "train_loss": 0.011931313201785088,
          "validation_loss": 0.009416318498551846,
          "train_metric": 0.001201678067445755,
          "validation_metric": 0.0009483771282248199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6413487489990075
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.010917649735347368,
          "train_loss": 0.009363003075122833,
          "validation_loss": 0.0077048675157129765,
          "train_metric": 0.0009430072968825698,
          "validation_metric": 0.0007760060834698379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6539876659990114
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.010771791494335048,
          "train_loss": 0.010960734449326992,
          "validation_loss": 0.008738063275814056,
          "train_metric": 0.0011039251694455743,
          "validation_metric": 0.0008800657233223319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6661598959999537
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.010923735419055447,
          "train_loss": 0.01105366088449955,
          "validation_loss": 0.010203905403614044,
          "train_metric": 0.0011132843792438507,
          "validation_metric": 0.0010276999091729522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6783314329986752
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01212187876808457,
          "train_loss": 0.009163382463157177,
          "validation_loss": 0.007841920480132103,
          "train_metric": 0.0009229023708030581,
          "validation_metric": 0.0007898096228018403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6904104969999025
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.008216747257392853,
          "train_loss": 0.007261296734213829,
          "validation_loss": 0.0059624332934618,
          "train_metric": 0.0007313311216421425,
          "validation_metric": 0.0006005145260132849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7032133149987203
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.008077889462583698,
          "train_loss": 0.007649097591638565,
          "validation_loss": 0.005851627327501774,
          "train_metric": 0.000770389218814671,
          "validation_metric": 0.0005893544293940067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7158903589988768
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.008498999843141064,
          "train_loss": 0.007307685445994139,
          "validation_loss": 0.00559674808755517,
          "train_metric": 0.0007360032759606838,
          "validation_metric": 0.000563683919608593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7284438049991877
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.007667303398193326,
          "train_loss": 0.011236198246479034,
          "validation_loss": 0.008767668157815933,
          "train_metric": 0.001131668803282082,
          "validation_metric": 0.0008830475271679461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7408899559995916
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.010174479757552035,
          "train_loss": 0.011961297132074833,
          "validation_loss": 0.010597237385809422,
          "train_metric": 0.0012046981137245893,
          "validation_metric": 0.0010673149954527617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.753096550999544
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.008994228963274509,
          "train_loss": 0.006712076719850302,
          "validation_loss": 0.005419877357780933,
          "train_metric": 0.0006760156829841435,
          "validation_metric": 0.0005458702798932791,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.765437733998624
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.006536780289025046,
          "train_loss": 0.005785043817013502,
          "validation_loss": 0.004688166081905365,
          "train_metric": 0.000582648441195488,
          "validation_metric": 0.0004721749573945999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7811901089989988
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0065881161208380945,
          "train_loss": 0.005793933756649494,
          "validation_loss": 0.00467495946213603,
          "train_metric": 0.0005835437914356589,
          "validation_metric": 0.0004708448250312358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7939569569989544
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.006135555697255768,
          "train_loss": 0.005960063077509403,
          "validation_loss": 0.0046024564653635025,
          "train_metric": 0.000600275699980557,
          "validation_metric": 0.0004635425575543195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8086112999990291
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.008340344073076267,
          "train_loss": 0.0053955321200191975,
          "validation_loss": 0.0042519099079072475,
          "train_metric": 0.0005434181657619774,
          "validation_metric": 0.00042823681724257767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8217218709996814
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.006701156497001648,
          "train_loss": 0.005240884143859148,
          "validation_loss": 0.004417655989527702,
          "train_metric": 0.0005278426688164473,
          "validation_metric": 0.00044493016321212053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8336725079989264
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.005232843104749918,
          "train_loss": 0.005497852806001902,
          "validation_loss": 0.004348119255155325,
          "train_metric": 0.0005537235410884023,
          "validation_metric": 0.0004379266465548426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8460099159983656
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.005574253547820263,
          "train_loss": 0.004958148114383221,
          "validation_loss": 0.003961426671594381,
          "train_metric": 0.0004993664915673435,
          "validation_metric": 0.0003989803954027593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8585442319999856
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.00596652805688791,
          "train_loss": 0.007669753395020962,
          "validation_loss": 0.005518784746527672,
          "train_metric": 0.0007724695024080575,
          "validation_metric": 0.0005558318225666881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8753680829995574
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.007040478718408849,
          "train_loss": 0.0045489477925002575,
          "validation_loss": 0.0034997821785509586,
          "train_metric": 0.00045815337216481566,
          "validation_metric": 0.00035248525091446936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8928928389996145
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.004746855513076298,
          "train_loss": 0.004951787646859884,
          "validation_loss": 0.003824794664978981,
          "train_metric": 0.0004987259744666517,
          "validation_metric": 0.0003852193185593933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9060990589987341
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.004682888953539077,
          "train_loss": 0.004173777066171169,
          "validation_loss": 0.003389031393453479,
          "train_metric": 0.00042036757804453373,
          "validation_metric": 0.0003413308004382998,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9190600149995589
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.004865808194153942,
          "train_loss": 0.005138661712408066,
          "validation_loss": 0.004393394570797682,
          "train_metric": 0.0005175471887923777,
          "validation_metric": 0.00044248663471080363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9315275209992251
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.004322871180193033,
          "train_loss": 0.0038354084827005863,
          "validation_loss": 0.0030910433270037174,
          "train_metric": 0.00038628833135589957,
          "validation_metric": 0.0003113185230176896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9450390099991637
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.005110518111905549,
          "train_loss": 0.0039623798802495,
          "validation_loss": 0.0033197978045791388,
          "train_metric": 0.0003990764089394361,
          "validation_metric": 0.00033435781369917095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9581005370000639
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.004062076404807158,
          "train_loss": 0.0038646855391561985,
          "validation_loss": 0.002952179405838251,
          "train_metric": 0.0003892369568347931,
          "validation_metric": 0.00029733270639553666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9733313500000804
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.003836657095234841,
          "train_loss": 0.003577067982405424,
          "validation_loss": 0.002734485315158963,
          "train_metric": 0.0003602691867854446,
          "validation_metric": 0.0002754073648247868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9885696889996325
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.00509686558143585,
          "train_loss": 0.004282197915017605,
          "validation_loss": 0.003710122313350439,
          "train_metric": 0.0004312872770242393,
          "validation_metric": 0.0003736699582077563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0012351989989838
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.0041459641652181745,
          "train_loss": 0.003551909700036049,
          "validation_loss": 0.002971809357404709,
          "train_metric": 0.0003577353782020509,
          "validation_metric": 0.00029930975870229304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.01339944099891
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.004384167586977128,
          "train_loss": 0.003314019413664937,
          "validation_loss": 0.002620602725073695,
          "train_metric": 0.000333775911713019,
          "validation_metric": 0.0002639375161379576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0267000759995426
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.005648268532240763,
          "train_loss": 0.005830828566104174,
          "validation_loss": 0.005836748518049717,
          "train_metric": 0.0005872596520930529,
          "validation_metric": 0.0005878559313714504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.039625878998777
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005174226003873628,
          "train_loss": 0.003936780616641045,
          "validation_loss": 0.003049918683245778,
          "train_metric": 0.00039649810059927404,
          "validation_metric": 0.00030717661138623953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0552784969986533
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.00453178755560657,
          "train_loss": 0.0036793455947190523,
          "validation_loss": 0.0026932216715067625,
          "train_metric": 0.0003705702256411314,
          "validation_metric": 0.0002712513960432261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0699297129995102
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.004205399258353282,
          "train_loss": 0.0038899113424122334,
          "validation_loss": 0.0028871390968561172,
          "train_metric": 0.0003917776048183441,
          "validation_metric": 0.0002907820453401655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0863554339994153
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0038728301879018545,
          "train_loss": 0.003499331884086132,
          "validation_loss": 0.003215200500562787,
          "train_metric": 0.0003524399362504482,
          "validation_metric": 0.0003238232166040689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.098221864000152
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0036473638683673926,
          "train_loss": 0.003771416377276182,
          "validation_loss": 0.0027857832610607147,
          "train_metric": 0.00037984325899742544,
          "validation_metric": 0.00028057387680746615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1121328469998844
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.003458400802628603,
          "train_loss": 0.003879622323438525,
          "validation_loss": 0.0033926162868738174,
          "train_metric": 0.000390741421142593,
          "validation_metric": 0.00034169189166277647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1251328949983872
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.005329517705831677,
          "train_loss": 0.0032775886356830597,
          "validation_loss": 0.0024773955810815096,
          "train_metric": 0.0003301067918073386,
          "validation_metric": 0.00024951421073637903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1381226799985598
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.003316688191262074,
          "train_loss": 0.003785735694691539,
          "validation_loss": 0.0031544978264719248,
          "train_metric": 0.0003812854120042175,
          "validation_metric": 0.00031770949135534465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.150178521998896
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.00453625380032463,
          "train_loss": 0.004079543985426426,
          "validation_loss": 0.0032002145890146494,
          "train_metric": 0.0004108767316211015,
          "validation_metric": 0.00032231389195658267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1638900379984989
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0037663868024537805,
          "train_loss": 0.004027614835649729,
          "validation_loss": 0.003966418094933033,
          "train_metric": 0.0004056465986650437,
          "validation_metric": 0.0003994831058662385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1767592460000742
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0034443380991433514,
          "train_loss": 0.003962175454944372,
          "validation_loss": 0.0028440284077078104,
          "train_metric": 0.00039905577432364225,
          "validation_metric": 0.0002864401030819863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1917054449986608
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.005247549022897147,
          "train_loss": 0.004941886756569147,
          "validation_loss": 0.0034212388563901186,
          "train_metric": 0.0004977287608198822,
          "validation_metric": 0.0003445746551733464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2053943390001223
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.0038986103536444716,
          "train_loss": 0.004505482967942953,
          "validation_loss": 0.004252587910741568,
          "train_metric": 0.0004537757486104965,
          "validation_metric": 0.00042830509482882917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2175079609987733
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.002914044765930157,
          "train_loss": 0.0029864690732210875,
          "validation_loss": 0.002093313029035926,
          "train_metric": 0.00030078619602136314,
          "validation_metric": 0.00021083082538098097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.233156346999749
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.003816168988123536,
          "train_loss": 0.006137951742857695,
          "validation_loss": 0.004789342638105154,
          "train_metric": 0.0006181921344250441,
          "validation_metric": 0.000482365139760077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2463782949998858
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.004099367077287752,
          "train_loss": 0.0053718434646725655,
          "validation_loss": 0.00436248118057847,
          "train_metric": 0.0005410324083641171,
          "validation_metric": 0.00043937310692854226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3436774490000971
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0035126487346133217,
          "train_loss": 0.004578305408358574,
          "validation_loss": 0.0038876014295965433,
          "train_metric": 0.000461110204923898,
          "validation_metric": 0.000391544948797673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.356895019998774
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.004955210104526486,
          "train_loss": 0.004151687026023865,
          "validation_loss": 0.003414468839764595,
          "train_metric": 0.0004181427357252687,
          "validation_metric": 0.00034389275242574513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3693331809990923
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0037477743317140266,
          "train_loss": 0.002746446058154106,
          "validation_loss": 0.0021930004004389048,
          "train_metric": 0.000276611972367391,
          "validation_metric": 0.00022087099205236882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3834314649993757
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.003487267626042012,
          "train_loss": 0.0022940083872526884,
          "validation_loss": 0.001710016862489283,
          "train_metric": 0.00023104414867702872,
          "validation_metric": 0.00017222663154825568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3981265549991804
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.0031680403280915925,
          "train_loss": 0.0025812857784330845,
          "validation_loss": 0.0019492690917104483,
          "train_metric": 0.00025997761986218393,
          "validation_metric": 0.00019632326439023018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4140223239992338
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0032549872012168635,
          "train_loss": 0.003036903915926814,
          "validation_loss": 0.0020862831734120846,
          "train_metric": 0.00030586583307012916,
          "validation_metric": 0.00021012278739362955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4299714579992724
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.003344183242006693,
          "train_loss": 0.0035849774722009897,
          "validation_loss": 0.0024253600277006626,
          "train_metric": 0.00036106581683270633,
          "validation_metric": 0.00024427336757071316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4433604259993444
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004090118804015219,
          "train_loss": 0.007177236024290323,
          "validation_loss": 0.0058961305767297745,
          "train_metric": 0.0007228649919852614,
          "validation_metric": 0.0005938367103226483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4581614109993097
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.005558350021601655,
          "train_loss": 0.005319507792592049,
          "validation_loss": 0.004802244249731302,
          "train_metric": 0.0005357612972147763,
          "validation_metric": 0.0004836644802708179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.471121761998802
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0056469647970516235,
          "train_loss": 0.0035494505427777767,
          "validation_loss": 0.0027745466213673353,
          "train_metric": 0.00035748767550103366,
          "validation_metric": 0.0002794422034639865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4835435529985261
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.003224120449885959,
          "train_loss": 0.004026519600301981,
          "validation_loss": 0.0034287115558981895,
          "train_metric": 0.00040553638245910406,
          "validation_metric": 0.0003453272511251271,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5007073910001054
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0036395309107319918,
          "train_loss": 0.007384158670902252,
          "validation_loss": 0.005439894273877144,
          "train_metric": 0.0007437054300680757,
          "validation_metric": 0.0005478861276060343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5188059899992368
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0073529550681996625,
          "train_loss": 0.0036527393385767937,
          "validation_loss": 0.0027631015982478857,
          "train_metric": 0.0003678905195556581,
          "validation_metric": 0.00027828948805108666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5323427819985227
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004833844868699089,
          "train_loss": 0.003450297750532627,
          "validation_loss": 0.0025516068562865257,
          "train_metric": 0.000347501365467906,
          "validation_metric": 0.00025698854005895555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5463612549992831
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0034988251063623466,
          "train_loss": 0.002902941545471549,
          "validation_loss": 0.0022173181641846895,
          "train_metric": 0.0002923736465163529,
          "validation_metric": 0.00022332016669679433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5599092379998183
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0031514863076154143,
          "train_loss": 0.002242566319182515,
          "validation_loss": 0.001698821666650474,
          "train_metric": 0.00022586305567529052,
          "validation_metric": 0.00017109912005253136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5741515800000343
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0027485364298627246,
          "train_loss": 0.004670264665037394,
          "validation_loss": 0.003991364501416683,
          "train_metric": 0.0004703719459939748,
          "validation_metric": 0.0004019955813419074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5880250399986835
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.003541326201229822,
          "train_loss": 0.002916081342846155,
          "validation_loss": 0.0020876191556453705,
          "train_metric": 0.0002936969976872206,
          "validation_metric": 0.00021025733440183103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.600846661000105
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0027197273884667084,
          "train_loss": 0.004372874740511179,
          "validation_loss": 0.002934122923761606,
          "train_metric": 0.0004404199426062405,
          "validation_metric": 0.00029551409534178674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6151932979992125
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.003504258409520844,
          "train_loss": 0.002982980804517865,
          "validation_loss": 0.00237321387976408,
          "train_metric": 0.00030043491278775036,
          "validation_metric": 0.00023902140674181283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.63025647899849
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.002345476181289996,
          "train_loss": 0.0021544736810028553,
          "validation_loss": 0.0016405207570642233,
          "train_metric": 0.00021699069475289434,
          "validation_metric": 0.0001652272476349026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6477633679987775
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.002726215847360436,
          "train_loss": 0.00498184422031045,
          "validation_loss": 0.0036196422297507524,
          "train_metric": 0.0005017531220801175,
          "validation_metric": 0.0003645571123342961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6638081959990814
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004399049619678408,
          "train_loss": 0.0054013985209167,
          "validation_loss": 0.004183716606348753,
          "train_metric": 0.0005440091481432319,
          "validation_metric": 0.0004213686042930931,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6969302259985852
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004836963496927638,
          "train_loss": 0.011842809617519379,
          "validation_loss": 0.008214925415813923,
          "train_metric": 0.001192764495499432,
          "validation_metric": 0.0008273771381936967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7147577699997782
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.005187090078834444,
          "train_loss": 0.0028289908077567816,
          "validation_loss": 0.0025834289845079184,
          "train_metric": 0.00028492556884884834,
          "validation_metric": 0.00026019354118034244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7395179719987937
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.003248647124564741,
          "train_loss": 0.005783838219940662,
          "validation_loss": 0.004615449346601963,
          "train_metric": 0.0005825269035995007,
          "validation_metric": 0.0004648511530831456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.755916703999901
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.004937880214129109,
          "train_loss": 0.005871355067938566,
          "validation_loss": 0.00485972547903657,
          "train_metric": 0.0005913414061069489,
          "validation_metric": 0.0004894537851214409,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7688228959996195
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.004334301305789268,
          "train_loss": 0.003120627487078309,
          "validation_loss": 0.0021392914932221174,
          "train_metric": 0.00031429817317985,
          "validation_metric": 0.00021546159405261278,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7888706359990465
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.003540068275469821,
          "train_loss": 0.004014411475509405,
          "validation_loss": 0.0031976248137652874,
          "train_metric": 0.0004043167573399842,
          "validation_metric": 0.0003220530925318599,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8042337429997133
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.0032229784519586246,
          "train_loss": 0.0032652674708515406,
          "validation_loss": 0.002350008115172386,
          "train_metric": 0.00032886574626900256,
          "validation_metric": 0.00023668419453315437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8305531409987452
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004296511193388142,
          "train_loss": 0.003373754909262061,
          "validation_loss": 0.002181832678616047,
          "train_metric": 0.0003397921973373741,
          "validation_metric": 0.00021974615810904652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8518049539998174
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004272726309864083,
          "train_loss": 0.012255322188138962,
          "validation_loss": 0.008851749822497368,
          "train_metric": 0.0012343112612143159,
          "validation_metric": 0.0008915159269236028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.864677875999405
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.01044464748702012,
          "train_loss": 0.004407211672514677,
          "validation_loss": 0.003683132119476795,
          "train_metric": 0.00044387823436409235,
          "validation_metric": 0.000370951573131606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8778969699997106
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004382519124192186,
          "train_loss": 0.0030288584530353546,
          "validation_loss": 0.002407840918749571,
          "train_metric": 0.0003050555242225528,
          "validation_metric": 0.0002425089624011889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8907793139987916
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0023513078813266475,
          "train_loss": 0.0025506787933409214,
          "validation_loss": 0.0018231121357530355,
          "train_metric": 0.0002568950585555285,
          "validation_metric": 0.00018361720140092075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.903460814999562
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.001825830151574337,
          "train_loss": 0.0019922577776014805,
          "validation_loss": 0.0015229035634547472,
          "train_metric": 0.0002006529102800414,
          "validation_metric": 0.00015338126104325056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9171726399999898
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0022063254018576117,
          "train_loss": 0.0019316565012559295,
          "validation_loss": 0.0015387659659609199,
          "train_metric": 0.00019454937137197703,
          "validation_metric": 0.0001549788867123425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9305015739992086
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.002154480076569598,
          "train_loss": 0.0024113664403557777,
          "validation_loss": 0.0017903544940054417,
          "train_metric": 0.00024286401458084583,
          "validation_metric": 0.0001803179766284302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9438628669995524
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0019091311369265895,
          "train_loss": 0.0017380539793521166,
          "validation_loss": 0.0013068488333374262,
          "train_metric": 0.00017505043069832027,
          "validation_metric": 0.0001316210109507665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9576427029987826
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0019350768552612863,
          "train_loss": 0.001794988289475441,
          "validation_loss": 0.0015393062494695187,
          "train_metric": 0.00018078464199788868,
          "validation_metric": 0.00015503328177146614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.970716347999769
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0023044914141792106,
          "train_loss": 0.001893677283078432,
          "validation_loss": 0.0013487932737916708,
          "train_metric": 0.00019072422583121806,
          "validation_metric": 0.000135845533804968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9836982439992425
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0019876232072419953,
          "train_loss": 0.001550995628349483,
          "validation_loss": 0.0012883511371910572,
          "train_metric": 0.00015621058992110193,
          "validation_metric": 0.00012975800200365484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9968650409991824
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.00222699336882215,
          "train_loss": 0.001985098933801055,
          "validation_loss": 0.0014261873438954353,
          "train_metric": 0.00019993192108813673,
          "validation_metric": 0.00014364035450853407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.010145301999728
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0018074178769893479,
          "train_loss": 0.0021582795307040215,
          "validation_loss": 0.001751840696670115,
          "train_metric": 0.00021737402130383998,
          "validation_metric": 0.00017643898900132626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0234348039994074
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0027196591181564145,
          "train_loss": 0.002093837596476078,
          "validation_loss": 0.0015529922675341368,
          "train_metric": 0.00021088364883325994,
          "validation_metric": 0.00015641169738955796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.036669739998615
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0018084968251059763,
          "train_loss": 0.0017936417134478688,
          "validation_loss": 0.0012959603918716311,
          "train_metric": 0.0001806490181479603,
          "validation_metric": 0.00013052437861915678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.049865218999912
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0017670158340479247,
          "train_loss": 0.0019063361687585711,
          "validation_loss": 0.0014254538109526038,
          "train_metric": 0.00019199919188395143,
          "validation_metric": 0.0001435665035387501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0628523729992594
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.001559081616505864,
          "train_loss": 0.002851765602827072,
          "validation_loss": 0.0019219876267015934,
          "train_metric": 0.00028721941635012627,
          "validation_metric": 0.0001935755426529795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.075596330998451
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0030184237475623377,
          "train_loss": 0.002037081168964505,
          "validation_loss": 0.0017491793259978294,
          "train_metric": 0.00020516735094133765,
          "validation_metric": 0.00017617092817090452,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.088249738999366
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0024248214285762515,
          "train_loss": 0.001506276079453528,
          "validation_loss": 0.0011315372539684176,
          "train_metric": 0.0001517066266387701,
          "validation_metric": 0.0001139642990892753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.100800153999444
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0015072305268404307,
          "train_loss": 0.001806273590773344,
          "validation_loss": 0.0012655986938625574,
          "train_metric": 0.00018192124844063073,
          "validation_metric": 0.00012746648280881345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.113044497998999
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0016914744865061948,
          "train_loss": 0.001737129408866167,
          "validation_loss": 0.0012283026007935405,
          "train_metric": 0.00017495731299277395,
          "validation_metric": 0.0001237101387232542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1252033689997916
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0027076987653344986,
          "train_loss": 0.002088127424940467,
          "validation_loss": 0.0017315553268417716,
          "train_metric": 0.00021030854259151965,
          "validation_metric": 0.00017439590010326356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1374843809990125
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.003391030611965107,
          "train_loss": 0.0017252409597858787,
          "validation_loss": 0.0012662751832976937,
          "train_metric": 0.00017375995230395347,
          "validation_metric": 0.00012753461487591267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1494514219994016
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.004514789230597671,
          "train_loss": 0.00251501495949924,
          "validation_loss": 0.00191604089923203,
          "train_metric": 0.00025330306380055845,
          "validation_metric": 0.00019297664402984083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.161267964998842
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.004851010533457156,
          "train_loss": 0.005317860282957554,
          "validation_loss": 0.004701729863882065,
          "train_metric": 0.0005355954053811729,
          "validation_metric": 0.00047354106209240854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1728222249985265
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0036616259058064315,
          "train_loss": 0.002335464349016547,
          "validation_loss": 0.002163161290809512,
          "train_metric": 0.00023521942785009742,
          "validation_metric": 0.00021786567231174558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1840528469983838
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.004035086156363832,
          "train_loss": 0.002103046979755163,
          "validation_loss": 0.0015019297134131193,
          "train_metric": 0.00021181117335800081,
          "validation_metric": 0.00015126884682103992,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.195826245999342
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.003658001693111146,
          "train_loss": 0.002724127843976021,
          "validation_loss": 0.0018821307457983494,
          "train_metric": 0.0002743641671258956,
          "validation_metric": 0.00018956130952574313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.207481610999821
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.004450508458830882,
          "train_loss": 0.004592586308717728,
          "validation_loss": 0.00393714988604188,
          "train_metric": 0.00046254845801740885,
          "validation_metric": 0.0003965353243984282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2189931419998175
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0030336353738675825,
          "train_loss": 0.010151272639632225,
          "validation_loss": 0.007980483584105968,
          "train_metric": 0.0010223989374935627,
          "validation_metric": 0.0008037651423364878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2304723509987525
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.006176056922413409,
          "train_loss": 0.007812047377228737,
          "validation_loss": 0.0053138891234993935,
          "train_metric": 0.0007868007523939013,
          "validation_metric": 0.0005351954605430365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2422084829995583
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0033567054888408165,
          "train_loss": 0.0015956431161612272,
          "validation_loss": 0.001269079279154539,
          "train_metric": 0.00016070732090156525,
          "validation_metric": 0.0001278170384466648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2542773919994943
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.002378961977228755,
          "train_loss": 0.003984562121331692,
          "validation_loss": 0.0036227814853191376,
          "train_metric": 0.00040131047717295587,
          "validation_metric": 0.00036487323814071715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.265965402999427
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0017419954056094866,
          "train_loss": 0.0019975474569946527,
          "validation_loss": 0.0014747733948752284,
          "train_metric": 0.00020118568500038236,
          "validation_metric": 0.0001485337852500379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2778125730001193
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.002009236726735253,
          "train_loss": 0.0032222780864685774,
          "validation_loss": 0.002430833177641034,
          "train_metric": 0.00032453605672344565,
          "validation_metric": 0.0002448245941195637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2902121909992275
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0021178481038077734,
          "train_loss": 0.0018424798035994172,
          "validation_loss": 0.0013109984574839473,
          "train_metric": 0.00018556779832579195,
          "validation_metric": 0.00013203897105995566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.301817678999214
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.003351371229655342,
          "train_loss": 0.0015642755897715688,
          "validation_loss": 0.0012831274652853608,
          "train_metric": 0.00015754811465740204,
          "validation_metric": 0.00012923189206048846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3135599229990476
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.003843556118226843,
          "train_loss": 0.0019408680964261293,
          "validation_loss": 0.0016805661143735051,
          "train_metric": 0.00019547712872736156,
          "validation_metric": 0.00016926045645959675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3254605869988154
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.002168361314033973,
          "train_loss": 0.001711748423986137,
          "validation_loss": 0.0014651279198005795,
          "train_metric": 0.00017240102170035243,
          "validation_metric": 0.0001475623284932226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.337116265000077
        }
      ],
      "best": {
        "epoch": 131,
        "updates": 2096
      },
      "best_validation_metric": 0.0001139642990892753,
      "config": {
        "label": "Residual",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": true,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010675406272797082
    },
    "5c044ad193b4_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.725183792412281,
          "train_loss": 0.271520733833313,
          "validation_loss": 0.22441114485263824,
          "train_metric": 0.027346575632691383,
          "validation_metric": 0.022601868957281113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04613069299921335
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.19251352921128273,
          "train_loss": 0.13974319398403168,
          "validation_loss": 0.1179056391119957,
          "train_metric": 0.014074422419071198,
          "validation_metric": 0.011875025928020477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06791702800001076
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12273477180860937,
          "train_loss": 0.10032490640878677,
          "validation_loss": 0.08398673683404922,
          "train_metric": 0.010104356333613396,
          "validation_metric": 0.008458836935460567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08689118999973289
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09152746060863137,
          "train_loss": 0.08048801124095917,
          "validation_loss": 0.06612011045217514,
          "train_metric": 0.008106458932161331,
          "validation_metric": 0.006659375037997961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10561180399963632
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07133611291646957,
          "train_loss": 0.06516025960445404,
          "validation_loss": 0.05265787988901138,
          "train_metric": 0.0065627028234303,
          "validation_metric": 0.005303508602082729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12535853999906976
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06069809605833143,
          "train_loss": 0.05711313337087631,
          "validation_loss": 0.047087281942367554,
          "train_metric": 0.005752225872129202,
          "validation_metric": 0.004742457997053862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14536120999946434
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05503121088258922,
          "train_loss": 0.05024784803390503,
          "validation_loss": 0.041228801012039185,
          "train_metric": 0.005060779396444559,
          "validation_metric": 0.004152413923293352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16510700099934184
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04905405070167035,
          "train_loss": 0.04574151337146759,
          "validation_loss": 0.03793175891041756,
          "train_metric": 0.004606917500495911,
          "validation_metric": 0.0038203480653464794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18577693499901216
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04519221524242312,
          "train_loss": 0.05024033412337303,
          "validation_loss": 0.04192633181810379,
          "train_metric": 0.005060022231191397,
          "validation_metric": 0.004222666844725609,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20488474999910977
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.049359321128576994,
          "train_loss": 0.046037573367357254,
          "validation_loss": 0.039142146706581116,
          "train_metric": 0.004636736121028662,
          "validation_metric": 0.003942254465073347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22325832599926798
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04093247506534681,
          "train_loss": 0.03829994797706604,
          "validation_loss": 0.03261260688304901,
          "train_metric": 0.0038574300706386566,
          "validation_metric": 0.003284622449427843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.242351561999385
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.036824317183345556,
          "train_loss": 0.032820843160152435,
          "validation_loss": 0.028341077268123627,
          "train_metric": 0.003305594902485609,
          "validation_metric": 0.002854409161955118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26254227599929436
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03441062942147255,
          "train_loss": 0.03700372204184532,
          "validation_loss": 0.02941253036260605,
          "train_metric": 0.0037268793676048517,
          "validation_metric": 0.0029623224399983883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28168109899888805
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.032163326046429574,
          "train_loss": 0.02908608876168728,
          "validation_loss": 0.024541635066270828,
          "train_metric": 0.0029294448904693127,
          "validation_metric": 0.0024717431515455246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41240718599874526
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03053868602728471,
          "train_loss": 0.02759496122598648,
          "validation_loss": 0.023318752646446228,
          "train_metric": 0.0027792633045464754,
          "validation_metric": 0.0023485796991735697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4312095629993564
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.026482005312573165,
          "train_loss": 0.027983058243989944,
          "validation_loss": 0.023301327601075172,
          "train_metric": 0.002818351611495018,
          "validation_metric": 0.002346824621781707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44980179099911766
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.02441761939553544,
          "train_loss": 0.023983746767044067,
          "validation_loss": 0.02047683112323284,
          "train_metric": 0.002415555063635111,
          "validation_metric": 0.0020623516757041216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46800199300014356
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.023378114216029644,
          "train_loss": 0.02086450532078743,
          "validation_loss": 0.017464669421315193,
          "train_metric": 0.002101396443322301,
          "validation_metric": 0.0017589774215593934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48578859100052796
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.02140462800161913,
          "train_loss": 0.02289021760225296,
          "validation_loss": 0.018739325925707817,
          "train_metric": 0.002305419184267521,
          "validation_metric": 0.001887356280349195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5028034240003763
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.02563406975241378,
          "train_loss": 0.02189042419195175,
          "validation_loss": 0.017806872725486755,
          "train_metric": 0.0022047236561775208,
          "validation_metric": 0.001793443108908832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.519443715998932
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.023129807494115084,
          "train_loss": 0.020913323387503624,
          "validation_loss": 0.017478393390774727,
          "train_metric": 0.0021063131280243397,
          "validation_metric": 0.00176035996992141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5358352589992137
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.02696609584381804,
          "train_loss": 0.026205450296401978,
          "validation_loss": 0.021345825865864754,
          "train_metric": 0.0026393167208880186,
          "validation_metric": 0.002149873413145542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5528051639994374
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.019721148069947958,
          "train_loss": 0.01606700010597706,
          "validation_loss": 0.01406087540090084,
          "train_metric": 0.0016182095278054476,
          "validation_metric": 0.0014161599101498723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5705452450001758
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.017273633362492546,
          "train_loss": 0.014825965277850628,
          "validation_loss": 0.012481079436838627,
          "train_metric": 0.0014932170743122697,
          "validation_metric": 0.0012570485705509782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5883865079995303
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019757675006985664,
          "train_loss": 0.01432263944298029,
          "validation_loss": 0.012306571938097477,
          "train_metric": 0.001442523906007409,
          "validation_metric": 0.0012394727673381567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6060891319993971
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.017510230012703687,
          "train_loss": 0.014158128760755062,
          "validation_loss": 0.01238947082310915,
          "train_metric": 0.0014259549789130688,
          "validation_metric": 0.0012478221906349063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6230047270000796
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015726784709841013,
          "train_loss": 0.013463691808283329,
          "validation_loss": 0.01134172547608614,
          "train_metric": 0.0013560139341279864,
          "validation_metric": 0.001142297056503594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.658575497000129
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.013707112200791016,
          "train_loss": 0.012812243774533272,
          "validation_loss": 0.01123434491455555,
          "train_metric": 0.001290402258746326,
          "validation_metric": 0.0011314821895211935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7064270520004357
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016051386279286817,
          "train_loss": 0.02502411976456642,
          "validation_loss": 0.020168717950582504,
          "train_metric": 0.0025203379336744547,
          "validation_metric": 0.002031319309026003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7491778759995213
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01542237150715664,
          "train_loss": 0.014315691776573658,
          "validation_loss": 0.011095796711742878,
          "train_metric": 0.001441823784261942,
          "validation_metric": 0.001117528066970408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7936106389988709
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.015414124500239268,
          "train_loss": 0.021503446623682976,
          "validation_loss": 0.018202824518084526,
          "train_metric": 0.0021657487377524376,
          "validation_metric": 0.0018333217594772577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.834469965999233
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.020806275395443663,
          "train_loss": 0.017160432413220406,
          "validation_loss": 0.012870152480900288,
          "train_metric": 0.0017283360939472914,
          "validation_metric": 0.0012962345499545336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.865355902000374
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01630209616268985,
          "train_loss": 0.011724606156349182,
          "validation_loss": 0.009741163812577724,
          "train_metric": 0.0011808593990281224,
          "validation_metric": 0.0009810944320634007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8889529899988702
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013805997223244049,
          "train_loss": 0.023595713078975677,
          "validation_loss": 0.019702864810824394,
          "train_metric": 0.0023764735087752342,
          "validation_metric": 0.0019844004418700933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9200211799998215
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.015536964187049307,
          "train_loss": 0.011063741520047188,
          "validation_loss": 0.00977888610213995,
          "train_metric": 0.0011142995208501816,
          "validation_metric": 0.000984893529675901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9508316989995365
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.012316050342633389,
          "train_loss": 0.011551490053534508,
          "validation_loss": 0.0093732550740242,
          "train_metric": 0.0011634236434474587,
          "validation_metric": 0.0009440399007871747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0029045920000499
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.01322234922554344,
          "train_loss": 0.01286236196756363,
          "validation_loss": 0.010114856995642185,
          "train_metric": 0.0012954497942700982,
          "validation_metric": 0.001018731389194727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.023873703999925
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011155385203892365,
          "train_loss": 0.011489743366837502,
          "validation_loss": 0.010611145757138729,
          "train_metric": 0.0011572049697861075,
          "validation_metric": 0.0010687157046049833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0544678810001642
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.011601212609093636,
          "train_loss": 0.014726401306688786,
          "validation_loss": 0.010290934704244137,
          "train_metric": 0.0014831890584900975,
          "validation_metric": 0.0010364654008299112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0855778000004648
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012412100812071003,
          "train_loss": 0.012658714316785336,
          "validation_loss": 0.01127684861421585,
          "train_metric": 0.0012749392772093415,
          "validation_metric": 0.00113576278090477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1070867679991352
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011675719724735245,
          "train_loss": 0.009901653975248337,
          "validation_loss": 0.007692788261920214,
          "train_metric": 0.0009972583502531052,
          "validation_metric": 0.0007747894851490855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.134580077999999
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010013210485340096,
          "train_loss": 0.008797609247267246,
          "validation_loss": 0.007866949774324894,
          "train_metric": 0.0008860630914568901,
          "validation_metric": 0.0007923304219730198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1570314099990355
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009386818463099189,
          "train_loss": 0.013002975843846798,
          "validation_loss": 0.009492841549217701,
          "train_metric": 0.001309612300246954,
          "validation_metric": 0.0009560841717757285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1874150719995669
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01224605915194843,
          "train_loss": 0.009541220031678677,
          "validation_loss": 0.0073880041018128395,
          "train_metric": 0.0009609567932784557,
          "validation_metric": 0.0007440927438437939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2105030219991022
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.01069485695916228,
          "train_loss": 0.01067455392330885,
          "validation_loss": 0.009969848208129406,
          "train_metric": 0.0010751018999144435,
          "validation_metric": 0.0010041267378255725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2429527139993297
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010661405161954463,
          "train_loss": 0.009650606662034988,
          "validation_loss": 0.008344125933945179,
          "train_metric": 0.0009719737572595477,
          "validation_metric": 0.0008403898100368679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.263677937000466
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011100387433543801,
          "train_loss": 0.015353727154433727,
          "validation_loss": 0.010953432880342007,
          "train_metric": 0.0015463712625205517,
          "validation_metric": 0.0011031896574422717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2799213039997994
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.010565859702182934,
          "train_loss": 0.008216824382543564,
          "validation_loss": 0.007050653453916311,
          "train_metric": 0.0008275684667751193,
          "validation_metric": 0.0007101160590536892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2972017449992563
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00831423222552985,
          "train_loss": 0.007751990109682083,
          "validation_loss": 0.007047897204756737,
          "train_metric": 0.0007807521033100784,
          "validation_metric": 0.000709838408511132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.314483926000321
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.008484421006869525,
          "train_loss": 0.008210123516619205,
          "validation_loss": 0.006662710104137659,
          "train_metric": 0.0008268934907391667,
          "validation_metric": 0.0006710437010042369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3317760459995043
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009121570168645121,
          "train_loss": 0.008138127624988556,
          "validation_loss": 0.006662833970040083,
          "train_metric": 0.0008196425042115152,
          "validation_metric": 0.0006710562738589942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3480439209997712
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.012653169003897347,
          "train_loss": 0.008039749227464199,
          "validation_loss": 0.006858522072434425,
          "train_metric": 0.0008097341051325202,
          "validation_metric": 0.0006907652714289725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3643681659996219
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008644956105854362,
          "train_loss": 0.008325216360390186,
          "validation_loss": 0.00712231732904911,
          "train_metric": 0.0008384853135794401,
          "validation_metric": 0.0007173336343839765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4023625779991562
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.00779974230681546,
          "train_loss": 0.008252358995378017,
          "validation_loss": 0.007306789513677359,
          "train_metric": 0.0008311473648063838,
          "validation_metric": 0.0007359131122939289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4194533219997538
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.00925846655445639,
          "train_loss": 0.00652270158752799,
          "validation_loss": 0.0057301875203847885,
          "train_metric": 0.0006569424876943231,
          "validation_metric": 0.0005771234282292426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4364527119996637
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.012231946529936977,
          "train_loss": 0.00858000572770834,
          "validation_loss": 0.00622341874986887,
          "train_metric": 0.0008641468011774123,
          "validation_metric": 0.0006267998833209276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4545662599994102
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008446793217444792,
          "train_loss": 0.00689212791621685,
          "validation_loss": 0.006349670700728893,
          "train_metric": 0.0006941498722881079,
          "validation_metric": 0.0006395156960934401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4726297180004622
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.008316159102832898,
          "train_loss": 0.007118558511137962,
          "validation_loss": 0.006148797459900379,
          "train_metric": 0.0007169551681727171,
          "validation_metric": 0.0006192844011820853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.489795122999567
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007200873791589402,
          "train_loss": 0.006566303316503763,
          "validation_loss": 0.005335439462214708,
          "train_metric": 0.0006613340810872614,
          "validation_metric": 0.0005373659078031778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5082377239996276
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006584036775166169,
          "train_loss": 0.006239462643861771,
          "validation_loss": 0.005712360609322786,
          "train_metric": 0.0006284158444032073,
          "validation_metric": 0.0005753280129283667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.525946974999897
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.006660603481577709,
          "train_loss": 0.006314736790955067,
          "validation_loss": 0.005067717283964157,
          "train_metric": 0.0006359972176142037,
          "validation_metric": 0.0005104019655846059,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5471406569995452
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.006349309740471654,
          "train_loss": 0.005507147405296564,
          "validation_loss": 0.004698514938354492,
          "train_metric": 0.0005546596948988736,
          "validation_metric": 0.00047321722377091646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5723836659999506
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.008569663259549998,
          "train_loss": 0.015207894146442413,
          "validation_loss": 0.0099509721621871,
          "train_metric": 0.001531683374196291,
          "validation_metric": 0.0010022253263741732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5914630129991565
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.010702520652557723,
          "train_loss": 0.0073378500528633595,
          "validation_loss": 0.007015361916273832,
          "train_metric": 0.0007390414248220623,
          "validation_metric": 0.0007065616082400084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.610082216999217
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.007163544389186427,
          "train_loss": 0.0053948210552334785,
          "validation_loss": 0.0044744606129825115,
          "train_metric": 0.0005433466285467148,
          "validation_metric": 0.00045065133599564433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6305230340003618
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005669868871336803,
          "train_loss": 0.005300572142004967,
          "validation_loss": 0.0045133717358112335,
          "train_metric": 0.0005338541814126074,
          "validation_metric": 0.00045457022497430444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6490556739991007
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0069226891282596625,
          "train_loss": 0.007420714013278484,
          "validation_loss": 0.006919790990650654,
          "train_metric": 0.0007473871810361743,
          "validation_metric": 0.000696935981977731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6675467000004573
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00731872973847203,
          "train_loss": 0.006452545523643494,
          "validation_loss": 0.0051038870587944984,
          "train_metric": 0.00064987683435902,
          "validation_metric": 0.0005140448338352144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6867432399994868
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.009972294181352481,
          "train_loss": 0.0056060985662043095,
          "validation_loss": 0.004278527572751045,
          "train_metric": 0.0005646257195621729,
          "validation_metric": 0.00043091760016977787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7053536379989964
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.007715228144661523,
          "train_loss": 0.00778129231184721,
          "validation_loss": 0.007210019044578075,
          "train_metric": 0.0007837032317183912,
          "validation_metric": 0.0007261668215505779,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7233165349989576
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006288037599006202,
          "train_loss": 0.005706281866878271,
          "validation_loss": 0.00420799758285284,
          "train_metric": 0.0005747157265432179,
          "validation_metric": 0.00042381405364722013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7422459019999224
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00576221395022003,
          "train_loss": 0.005137464962899685,
          "validation_loss": 0.004070205148309469,
          "train_metric": 0.000517426582518965,
          "validation_metric": 0.0004099361540284008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7600796940005239
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.005247840948868543,
          "train_loss": 0.006295354105532169,
          "validation_loss": 0.005427327938377857,
          "train_metric": 0.0006340449908748269,
          "validation_metric": 0.0005466206348501146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7767507709995698
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005847955937497318,
          "train_loss": 0.0056092459708452225,
          "validation_loss": 0.004303757566958666,
          "train_metric": 0.0005649426602758467,
          "validation_metric": 0.0004334586556069553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7933424989987543
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00761535503261257,
          "train_loss": 0.0046490756794810295,
          "validation_loss": 0.004038749262690544,
          "train_metric": 0.00046823782031424344,
          "validation_metric": 0.00040676796925254166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8105332739996811
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0055943846455193125,
          "train_loss": 0.005045777186751366,
          "validation_loss": 0.003924061544239521,
          "train_metric": 0.000508192228153348,
          "validation_metric": 0.0003952171537093818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.828887797999414
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.00631263930699788,
          "train_loss": 0.005989368073642254,
          "validation_loss": 0.005381503142416477,
          "train_metric": 0.0006032272358424962,
          "validation_metric": 0.0005420053494162858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8460242450000806
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.005129424927872606,
          "train_loss": 0.004593792837113142,
          "validation_loss": 0.0038168870378285646,
          "train_metric": 0.00046266999561339617,
          "validation_metric": 0.0003844228631351143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8638475829993695
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.00618596728600096,
          "train_loss": 0.004681388847529888,
          "validation_loss": 0.004123073071241379,
          "train_metric": 0.0004714923270512372,
          "validation_metric": 0.00041526081622578204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8801999400002387
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004919681974570267,
          "train_loss": 0.004786643199622631,
          "validation_loss": 0.0038449480198323727,
          "train_metric": 0.00048209322267211974,
          "validation_metric": 0.00038724910700693727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8965845740003715
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006906665556016378,
          "train_loss": 0.012297742068767548,
          "validation_loss": 0.009444266557693481,
          "train_metric": 0.0012385835871100426,
          "validation_metric": 0.0009511919342912734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.91313860199989
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.010686725392588414,
          "train_loss": 0.0064812321215868,
          "validation_loss": 0.0057849641889333725,
          "train_metric": 0.00065276597160846,
          "validation_metric": 0.0005826404667459428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.929836115999933
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.007700789232330862,
          "train_loss": 0.004105992149561644,
          "validation_loss": 0.003409368684515357,
          "train_metric": 0.0004135404305998236,
          "validation_metric": 0.0003433791280258447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9472767890001705
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.004603917019267101,
          "train_loss": 0.004845749586820602,
          "validation_loss": 0.0035455594770610332,
          "train_metric": 0.0004880461492575705,
          "validation_metric": 0.00035709579242393374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9642451820000133
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.007054458052152768,
          "train_loss": 0.015696968883275986,
          "validation_loss": 0.012972714379429817,
          "train_metric": 0.0015809413744136691,
          "validation_metric": 0.0013065641978755593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9811879779990704
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.011451014113845304,
          "train_loss": 0.020557669922709465,
          "validation_loss": 0.013468781486153603,
          "train_metric": 0.0020704930648207664,
          "validation_metric": 0.0013565262779593468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9976727009998285
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.015485930867725983,
          "train_loss": 0.00581009965389967,
          "validation_loss": 0.004421717952936888,
          "train_metric": 0.0005851719761267304,
          "validation_metric": 0.0004453392466530204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.015418167999087
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.006737215720931999,
          "train_loss": 0.0061674462631344795,
          "validation_loss": 0.0055349646136164665,
          "train_metric": 0.0006211625295691192,
          "validation_metric": 0.0005574613460339606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.036291214999437
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005034035624703392,
          "train_loss": 0.0039005510043352842,
          "validation_loss": 0.0032008232083171606,
          "train_metric": 0.00039284920785576105,
          "validation_metric": 0.000322375213727355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.059158578000279
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004518714900768828,
          "train_loss": 0.0038000443018972874,
          "validation_loss": 0.0031561877112835646,
          "train_metric": 0.0003827265463769436,
          "validation_metric": 0.0003178796614520252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.090031880999959
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.006248384044738486,
          "train_loss": 0.005066498648375273,
          "validation_loss": 0.004222129005938768,
          "train_metric": 0.0005102792056277394,
          "validation_metric": 0.00042523740557953715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.120562078000148
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.005300535114656668,
          "train_loss": 0.004341588355600834,
          "validation_loss": 0.0038378548342734575,
          "train_metric": 0.0004372688417788595,
          "validation_metric": 0.00038653466617688537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1544731369995134
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004943848747643642,
          "train_loss": 0.0077809556387364864,
          "validation_loss": 0.005293975584208965,
          "train_metric": 0.0007836693548597395,
          "validation_metric": 0.0005331897991709411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2008124940002745
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004287212868803181,
          "train_loss": 0.008774102665483952,
          "validation_loss": 0.007212408818304539,
          "train_metric": 0.0008836955530568957,
          "validation_metric": 0.0007264073938131332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2211285779994796
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0035014182722079568,
          "train_loss": 0.004472360014915466,
          "validation_loss": 0.004246356897056103,
          "train_metric": 0.00045043969294056296,
          "validation_metric": 0.0004276775289326906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.240846276999946
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004747023289382923,
          "train_loss": 0.00342956674285233,
          "validation_loss": 0.0028597929049283266,
          "train_metric": 0.000345413398463279,
          "validation_metric": 0.00028802783344872296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2616200279990153
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005751226031861734,
          "train_loss": 0.006062327418476343,
          "validation_loss": 0.004158920142799616,
          "train_metric": 0.0006105754291638732,
          "validation_metric": 0.0004188711754977703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.282315757998731
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.006527146659209393,
          "train_loss": 0.004064810462296009,
          "validation_loss": 0.003415434155613184,
          "train_metric": 0.000409392814617604,
          "validation_metric": 0.00034399001742713153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3023292580001
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.005003856465918943,
          "train_loss": 0.0043891118839383125,
          "validation_loss": 0.0036584429908543825,
          "train_metric": 0.00044205531594343483,
          "validation_metric": 0.000368465029168874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3218675009993603
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.005998564207402524,
          "train_loss": 0.004563015885651112,
          "validation_loss": 0.004187696613371372,
          "train_metric": 0.00045957029215060174,
          "validation_metric": 0.0004217694513499737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.341236537999066
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005680662710801698,
          "train_loss": 0.004552732687443495,
          "validation_loss": 0.0037168257404118776,
          "train_metric": 0.0004585345450323075,
          "validation_metric": 0.0003743450506590307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.361719106998862
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.006389895494066877,
          "train_loss": 0.004864814691245556,
          "validation_loss": 0.004019264131784439,
          "train_metric": 0.000489966303575784,
          "validation_metric": 0.00040480552706867456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.381684685999062
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.007791414565872401,
          "train_loss": 0.005660008639097214,
          "validation_loss": 0.004888178315013647,
          "train_metric": 0.000570055388379842,
          "validation_metric": 0.0004923194646835327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.401906240000244
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0048040308320196345,
          "train_loss": 0.0035663405433297157,
          "validation_loss": 0.0027851785998791456,
          "train_metric": 0.0003591887070797384,
          "validation_metric": 0.00028051293338648975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.423358054000346
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004540373971394729,
          "train_loss": 0.004792868625372648,
          "validation_loss": 0.003553792368620634,
          "train_metric": 0.0004827201773878187,
          "validation_metric": 0.0003579249605536461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.442733583999143
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004516240820521489,
          "train_loss": 0.0034971204586327076,
          "validation_loss": 0.00261951913125813,
          "train_metric": 0.00035221720463596284,
          "validation_metric": 0.0002638284058775753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.461071820000143
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0035208531917305663,
          "train_loss": 0.00888960249722004,
          "validation_loss": 0.0073181455954909325,
          "train_metric": 0.0008953282376751304,
          "validation_metric": 0.0007370567764155567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.478605443999186
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.004127585325477412,
          "train_loss": 0.005249641835689545,
          "validation_loss": 0.0033487235195934772,
          "train_metric": 0.0005287246312946081,
          "validation_metric": 0.00033727110712789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.496475850999559
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.004842965594434645,
          "train_loss": 0.002905305940657854,
          "validation_loss": 0.0024432791396975517,
          "train_metric": 0.00029261180316098034,
          "validation_metric": 0.0002460781252011657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.514701430000059
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.005225695929766516,
          "train_loss": 0.0038044615648686886,
          "validation_loss": 0.0030250484123826027,
          "train_metric": 0.0003831714275293052,
          "validation_metric": 0.0003046718193218112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5323124259994074
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004379622194392141,
          "train_loss": 0.004060350824147463,
          "validation_loss": 0.0038772965781390667,
          "train_metric": 0.00040894359699450433,
          "validation_metric": 0.00039050710620358586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.551822841998728
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004134063194214832,
          "train_loss": 0.004081107676029205,
          "validation_loss": 0.0038114036433398724,
          "train_metric": 0.00041103424155153334,
          "validation_metric": 0.0003838706179521978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.569269023999368
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0034449776394467335,
          "train_loss": 0.002791373757645488,
          "validation_loss": 0.00257500889711082,
          "train_metric": 0.0002811369486153126,
          "validation_metric": 0.00025934548466466367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.585618629998862
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.002757111069513485,
          "train_loss": 0.004488422069698572,
          "validation_loss": 0.0030822553671896458,
          "train_metric": 0.0004520574293565005,
          "validation_metric": 0.0003104334173258394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6024540799990064
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004882394205196761,
          "train_loss": 0.005478053819388151,
          "validation_loss": 0.004331367090344429,
          "train_metric": 0.0005517295212484896,
          "validation_metric": 0.0004362394392956048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6195696869999665
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0031214265327434987,
          "train_loss": 0.0033513247035443783,
          "validation_loss": 0.002528879791498184,
          "train_metric": 0.0003375331871211529,
          "validation_metric": 0.00025469952379353344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6362601179989724
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003535403222485911,
          "train_loss": 0.003856767201796174,
          "validation_loss": 0.0025257794186472893,
          "train_metric": 0.0003884395118802786,
          "validation_metric": 0.0002543872396927327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6527885460000107
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003580433942261152,
          "train_loss": 0.0028860922902822495,
          "validation_loss": 0.0023965039290487766,
          "train_metric": 0.0002906766312662512,
          "validation_metric": 0.00024136710271704942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6702557709995745
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0032227757619693875,
          "train_loss": 0.0024330851156264544,
          "validation_loss": 0.001969155389815569,
          "train_metric": 0.000245051400270313,
          "validation_metric": 0.00019832611724268645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.687457058000291
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.00312216387828812,
          "train_loss": 0.002400434808805585,
          "validation_loss": 0.001938469591550529,
          "train_metric": 0.00024176300212275237,
          "validation_metric": 0.00019523555238265544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7047967219987186
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.005049306331784464,
          "train_loss": 0.00326095730997622,
          "validation_loss": 0.0026451763696968555,
          "train_metric": 0.0003284316626377404,
          "validation_metric": 0.00026641247677616775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7215938249992178
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0034608831701916642,
          "train_loss": 0.0028723659925162792,
          "validation_loss": 0.0023677286226302385,
          "train_metric": 0.00028929419931955636,
          "validation_metric": 0.0002384689578320831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.738576286999887
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0028234652345418,
          "train_loss": 0.0027469599153846502,
          "validation_loss": 0.0020657279528677464,
          "train_metric": 0.0002766637480817735,
          "validation_metric": 0.00020805251551792026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.75588547299958
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0026031358720501885,
          "train_loss": 0.002991296350955963,
          "validation_loss": 0.0021256254985928535,
          "train_metric": 0.00030127246282063425,
          "validation_metric": 0.00021408521570265293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7729884999989736
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0029182495054556057,
          "train_loss": 0.002263699658215046,
          "validation_loss": 0.0020225406624376774,
          "train_metric": 0.00022799152065999806,
          "validation_metric": 0.00020370287529658526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7897188309998455
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.00309807939993334,
          "train_loss": 0.002908478956669569,
          "validation_loss": 0.0021945068147033453,
          "train_metric": 0.0002929313632193953,
          "validation_metric": 0.00022102269576862454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8061999599995033
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.00266599341557594,
          "train_loss": 0.0024597709998488426,
          "validation_loss": 0.0018571208929643035,
          "train_metric": 0.0002477391099091619,
          "validation_metric": 0.00018704244575928897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8229348530003335
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.004796148758032359,
          "train_loss": 0.004223702475428581,
          "validation_loss": 0.0028237407095730305,
          "train_metric": 0.0004253957886248827,
          "validation_metric": 0.00028439678135327995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8394350329999725
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.009019669574627187,
          "train_loss": 0.006652167998254299,
          "validation_loss": 0.005630786530673504,
          "train_metric": 0.0006699820514768362,
          "validation_metric": 0.0005671122344210744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.855812457999491
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.010253318818286061,
          "train_loss": 0.0058790575712919235,
          "validation_loss": 0.0038958757650107145,
          "train_metric": 0.0005921171396039426,
          "validation_metric": 0.00039237833698280156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8724371179996524
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0057141728029819205,
          "train_loss": 0.007645321078598499,
          "validation_loss": 0.00648478651419282,
          "train_metric": 0.0007700088317506015,
          "validation_metric": 0.0006531239487230778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8890299710001273
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.008052829813095741,
          "train_loss": 0.016303688287734985,
          "validation_loss": 0.011124902404844761,
          "train_metric": 0.001642047893255949,
          "validation_metric": 0.0011204594047740102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9058788170004846
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.008326188559294678,
          "train_loss": 0.004383569583296776,
          "validation_loss": 0.004234254360198975,
          "train_metric": 0.00044149704626761377,
          "validation_metric": 0.00042645863140933216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9223441039994213
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0033992926000792067,
          "train_loss": 0.0024024329613894224,
          "validation_loss": 0.0019044072832912207,
          "train_metric": 0.0002419642696622759,
          "validation_metric": 0.0001918049092637375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9388947219995316
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.002910113355028443,
          "train_loss": 0.005612793378531933,
          "validation_loss": 0.0055975476279854774,
          "train_metric": 0.0005652999971061945,
          "validation_metric": 0.0005637644208036363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9562087939993944
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.00336274285291438,
          "train_loss": 0.0024300217628479004,
          "validation_loss": 0.0018143948400393128,
          "train_metric": 0.0002447428705636412,
          "validation_metric": 0.00018273918249178678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9738263159997587
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003300657535874052,
          "train_loss": 0.00337297678925097,
          "validation_loss": 0.002545601222664118,
          "train_metric": 0.00033971384982578456,
          "validation_metric": 0.00025638361694291234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9902282309994916
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0036947657208656892,
          "train_loss": 0.002643207088112831,
          "validation_loss": 0.0021333610638976097,
          "train_metric": 0.00026621416327543557,
          "validation_metric": 0.00021486429614014924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.006156084000395
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0027905169190489687,
          "train_loss": 0.0048367613926529884,
          "validation_loss": 0.003739910200238228,
          "train_metric": 0.0004871408746112138,
          "validation_metric": 0.0003766701265703887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0225954449997516
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.005829701800394105,
          "train_loss": 0.010636413469910622,
          "validation_loss": 0.009382311254739761,
          "train_metric": 0.0010712605435401201,
          "validation_metric": 0.0009449520148336887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0396624419990985
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.010631502969772555,
          "train_loss": 0.008176790550351143,
          "validation_loss": 0.0050178635865449905,
          "train_metric": 0.0008235364221036434,
          "validation_metric": 0.0005053807981312275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.056452671999068
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.013285969442222267,
          "train_loss": 0.005460844840854406,
          "validation_loss": 0.004616655874997377,
          "train_metric": 0.0005499962717294693,
          "validation_metric": 0.0004649726615753025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0730434300003253
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.004386250668176217,
          "train_loss": 0.003142002271488309,
          "validation_loss": 0.002617886755615473,
          "train_metric": 0.0003164509544149041,
          "validation_metric": 0.00026366396923549473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0898031939996145
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0034448182450432796,
          "train_loss": 0.0038700609002262354,
          "validation_loss": 0.003437807783484459,
          "train_metric": 0.0003897783753927797,
          "validation_metric": 0.0003462434688117355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1068621199992776
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.004726875435153488,
          "train_loss": 0.0074235061183571815,
          "validation_loss": 0.0051461271941661835,
          "train_metric": 0.0007476683822460473,
          "validation_metric": 0.000518299057148397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.123833559999184
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0032514285885554273,
          "train_loss": 0.0020597921684384346,
          "validation_loss": 0.0017143039731308818,
          "train_metric": 0.00020745473739225417,
          "validation_metric": 0.00017265844508074224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1416522649997205
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0028847064895671792,
          "train_loss": 0.00222614873200655,
          "validation_loss": 0.0017028801376000047,
          "train_metric": 0.00022420956520363688,
          "validation_metric": 0.00017150785424746573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.160392874999161
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0026161493515246548,
          "train_loss": 0.002388717606663704,
          "validation_loss": 0.0022249927278608084,
          "train_metric": 0.0002405828854534775,
          "validation_metric": 0.0002240930771222338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.183982210999602
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0024451421122648753,
          "train_loss": 0.002188307698816061,
          "validation_loss": 0.0019238610984757543,
          "train_metric": 0.00022039833129383624,
          "validation_metric": 0.00019376425188966095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.20514689599986
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.002076177905109944,
          "train_loss": 0.002661527367308736,
          "validation_loss": 0.0017820227658376098,
          "train_metric": 0.000268059317022562,
          "validation_metric": 0.00017947881133295596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.232087391999812
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0022713253238180187,
          "train_loss": 0.0021308159921318293,
          "validation_loss": 0.0016065847594290972,
          "train_metric": 0.000214607993257232,
          "validation_metric": 0.00016180932288989425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.257206963000499
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.002485668381268624,
          "train_loss": 0.0022483482025563717,
          "validation_loss": 0.0015288699651136994,
          "train_metric": 0.00022644540877081454,
          "validation_metric": 0.00015398218238260597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2833199739998236
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0044211139174876735,
          "train_loss": 0.0026093204505741596,
          "validation_loss": 0.0019262847490608692,
          "train_metric": 0.00026280118618160486,
          "validation_metric": 0.00019400837481953204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.306783922998875
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0023705231069470756,
          "train_loss": 0.0022000048775225878,
          "validation_loss": 0.001563854981213808,
          "train_metric": 0.00022157642524689436,
          "validation_metric": 0.00015750573948025703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.325772700000016
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0022128868331492413,
          "train_loss": 0.0028951175045222044,
          "validation_loss": 0.002665025880560279,
          "train_metric": 0.00029158563120290637,
          "validation_metric": 0.00026841164799407125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.343418472999474
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0028586407934199087,
          "train_loss": 0.002053749980404973,
          "validation_loss": 0.0015130533138290048,
          "train_metric": 0.0002068461908493191,
          "validation_metric": 0.0001523892133263871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.365803982998841
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0024523594947822858,
          "train_loss": 0.0020305104553699493,
          "validation_loss": 0.0018025942845270038,
          "train_metric": 0.00020450558804441243,
          "validation_metric": 0.00018155068391934037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.389070946999709
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.002135887203621678,
          "train_loss": 0.001838091528043151,
          "validation_loss": 0.001428671763278544,
          "train_metric": 0.000185125827556476,
          "validation_metric": 0.0001438906037947163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.422877213999527
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0023210191720863804,
          "train_loss": 0.0025264176074415445,
          "validation_loss": 0.0016575446352362633,
          "train_metric": 0.0002544515300542116,
          "validation_metric": 0.00016694182704668492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.440735580999899
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0023631388357898686,
          "train_loss": 0.0028521050699055195,
          "validation_loss": 0.002574087819084525,
          "train_metric": 0.000287253555143252,
          "validation_metric": 0.0002592527016531676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.457603494000068
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.002485522913048044,
          "train_loss": 0.002309029456228018,
          "validation_loss": 0.001501815509982407,
          "train_metric": 0.00023255700943991542,
          "validation_metric": 0.0001512573508080095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.474767167999744
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.002716261169553036,
          "train_loss": 0.002138255164027214,
          "validation_loss": 0.0016314181266352534,
          "train_metric": 0.00021535724226851016,
          "validation_metric": 0.00016431046242360026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4912735409998277
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.00289134705599281,
          "train_loss": 0.0017966300947591662,
          "validation_loss": 0.0015389397740364075,
          "train_metric": 0.0001809499808587134,
          "validation_metric": 0.000154996378114447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.518895135999628
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0024329685511474963,
          "train_loss": 0.0017711227992549539,
          "validation_loss": 0.0013358104042708874,
          "train_metric": 0.00017838101484812796,
          "validation_metric": 0.00013453791325446218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5544434059993364
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0024775342462817207,
          "train_loss": 0.007562030106782913,
          "validation_loss": 0.004347797017544508,
          "train_metric": 0.0007616200600750744,
          "validation_metric": 0.0004378942248877138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5711258289993566
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.003393010872969171,
          "train_loss": 0.004969136323779821,
          "validation_loss": 0.003458495484665036,
          "train_metric": 0.0005004731356166303,
          "validation_metric": 0.0003483269829303026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5877467749996867
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0035988839917990845,
          "train_loss": 0.0024098376743495464,
          "validation_loss": 0.002166891237720847,
          "train_metric": 0.00024271004076581448,
          "validation_metric": 0.00021824135910719633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6041335579993756
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0026283382467227057,
          "train_loss": 0.0037163584493100643,
          "validation_loss": 0.002459761220961809,
          "train_metric": 0.00037429804797284305,
          "validation_metric": 0.0002477381203789264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.620310893998976
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.003228494020731887,
          "train_loss": 0.002183516975492239,
          "validation_loss": 0.002094058785587549,
          "train_metric": 0.0002199158479925245,
          "validation_metric": 0.00021090592781547457,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6368823169996176
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0027687562760547735,
          "train_loss": 0.00206978851929307,
          "validation_loss": 0.002118093427270651,
          "train_metric": 0.00020846155530307442,
          "validation_metric": 0.00021332658070605248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6532479830002558
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.00244362272133003,
          "train_loss": 0.0024799888487905264,
          "validation_loss": 0.0014960987027734518,
          "train_metric": 0.00024977544671855867,
          "validation_metric": 0.00015068158973008394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6700643570002285
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0024938411006587557,
          "train_loss": 0.0029078428633511066,
          "validation_loss": 0.0021193339489400387,
          "train_metric": 0.00029286730568856,
          "validation_metric": 0.00021345156710594893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.686768038000082
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.002222596980573144,
          "train_loss": 0.0017096850788220763,
          "validation_loss": 0.0011506980517879128,
          "train_metric": 0.0001721931912470609,
          "validation_metric": 0.00011589410132728517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7040685529991606
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0020034787467011483,
          "train_loss": 0.003791782772168517,
          "validation_loss": 0.0024487022310495377,
          "train_metric": 0.00038189446786418557,
          "validation_metric": 0.0002466243167873472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7210549039991747
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.002604292611067649,
          "train_loss": 0.001992400037124753,
          "validation_loss": 0.0013801869936287403,
          "train_metric": 0.00020066724391654134,
          "validation_metric": 0.00013900737394578755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.737965862999772
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0015919984834908973,
          "train_loss": 0.002042272360995412,
          "validation_loss": 0.0015085138147696853,
          "train_metric": 0.00020569018670357764,
          "validation_metric": 0.0001519319775979966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.754686643000241
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.002792971406961442,
          "train_loss": 0.0023457002826035023,
          "validation_loss": 0.0019018009770661592,
          "train_metric": 0.00023625030007679015,
          "validation_metric": 0.00019154239271301776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.77113975499924
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0028879059718747158,
          "train_loss": 0.002785990247502923,
          "validation_loss": 0.0017061191610991955,
          "train_metric": 0.00028059474425390363,
          "validation_metric": 0.00017183407908305526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.814025395000499
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0039992633974179626,
          "train_loss": 0.0028155019972473383,
          "validation_loss": 0.001958691980689764,
          "train_metric": 0.00028356711845844984,
          "validation_metric": 0.00019727229664567858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8571521829999256
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0038601312116952613,
          "train_loss": 0.0020218396093696356,
          "validation_loss": 0.0013067623367533088,
          "train_metric": 0.0002036323130596429,
          "validation_metric": 0.00013161230890545994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.889182749999236
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002202932177169714,
          "train_loss": 0.0015829962212592363,
          "validation_loss": 0.001153468736447394,
          "train_metric": 0.00015943360631354153,
          "validation_metric": 0.00011617315612966195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9056843309990654
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.002359665710173431,
          "train_loss": 0.0038902978412806988,
          "validation_loss": 0.0025316239334642887,
          "train_metric": 0.00039181651663966477,
          "validation_metric": 0.0002549758937675506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.92257467699892
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0027126112690893933,
          "train_loss": 0.0029516371432691813,
          "validation_loss": 0.0022234974894672632,
          "train_metric": 0.0002972780494019389,
          "validation_metric": 0.00022394253755919635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9393676669988054
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.002648346364367171,
          "train_loss": 0.0014680414460599422,
          "validation_loss": 0.0011522223940119147,
          "train_metric": 0.00014785575331188738,
          "validation_metric": 0.00011604760948102921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9556247130003612
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0015444501805177424,
          "train_loss": 0.0013326776679605246,
          "validation_loss": 0.0011286141816526651,
          "train_metric": 0.0001342224277323112,
          "validation_metric": 0.00011366987746441737,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.9727667970000766
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0014438526286539854,
          "train_loss": 0.0028183930553495884,
          "validation_loss": 0.002481012139469385,
          "train_metric": 0.0002838582149706781,
          "validation_metric": 0.00024987844517454505,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.989574535999054
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.002805521486152429,
          "train_loss": 0.0014091484481468797,
          "validation_loss": 0.0012372239725664258,
          "train_metric": 0.00014192429080139846,
          "validation_metric": 0.00012460867583286017,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.005991161999191
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.002527379860111978,
          "train_loss": 0.0018518472788855433,
          "validation_loss": 0.001227227388881147,
          "train_metric": 0.0001865113154053688,
          "validation_metric": 0.0001236018433701247,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.022228043999348
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0015583820713800378,
          "train_loss": 0.0030256714671850204,
          "validation_loss": 0.0019375901902094483,
          "train_metric": 0.0003047345671802759,
          "validation_metric": 0.00019514700397849083,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.038394794999476
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.004495949146075873,
          "train_loss": 0.0018739859806373715,
          "validation_loss": 0.0016684001311659813,
          "train_metric": 0.0001887410180643201,
          "validation_metric": 0.0001680351560935378,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.055030415998772
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0023916953687148634,
          "train_loss": 0.002186657628044486,
          "validation_loss": 0.002166250953450799,
          "train_metric": 0.0002202321484219283,
          "validation_metric": 0.0002181768650189042,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.07186080899919
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.0014535716745740501,
          "train_loss": 0.0012643536319956183,
          "validation_loss": 0.001038054353557527,
          "train_metric": 0.0001273410744033754,
          "validation_metric": 0.00010454904258949682,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.08979128899955
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0023627783884876408,
          "train_loss": 0.0023189205676317215,
          "validation_loss": 0.0022536322940140963,
          "train_metric": 0.00023355320445261896,
          "validation_metric": 0.00022697758686263114,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.148156944998846
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.001900272851344198,
          "train_loss": 0.0012602136703208089,
          "validation_loss": 0.0009810935007408261,
          "train_metric": 0.00012692410382442176,
          "validation_metric": 9.881213918561116e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.193751119999433
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0013291564864630345,
          "train_loss": 0.0016141049563884735,
          "validation_loss": 0.0011184168979525566,
          "train_metric": 0.00016256676462944597,
          "validation_metric": 0.00011264283239142969,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.220179312000255
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.001902655986668833,
          "train_loss": 0.001627777935937047,
          "validation_loss": 0.0010533332824707031,
          "train_metric": 0.00016394384147133678,
          "validation_metric": 0.00010608786396915093,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.237430060000406
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.002589630797956488,
          "train_loss": 0.002890179632231593,
          "validation_loss": 0.002744512865319848,
          "train_metric": 0.0002910883049480617,
          "validation_metric": 0.00027641726774163544,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.254542999999103
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.001925155322169303,
          "train_loss": 0.0013105501420795918,
          "validation_loss": 0.0009345832513645291,
          "train_metric": 0.0001319938019150868,
          "validation_metric": 9.412780491402373e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.2721815040004
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0012750103614962427,
          "train_loss": 0.0012210578424856067,
          "validation_loss": 0.0008948380127549171,
          "train_metric": 0.00012298047658987343,
          "validation_metric": 9.012482041725889e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.290479055000105
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0013941083689132938,
          "train_loss": 0.0011597024276852608,
          "validation_loss": 0.0009230952709913254,
          "train_metric": 0.00011680097668431699,
          "validation_metric": 9.297077485825866e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.308422078998774
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0013753145558439428,
          "train_loss": 0.0012393876677379012,
          "validation_loss": 0.0010221584234386683,
          "train_metric": 0.0001248266053153202,
          "validation_metric": 0.00010294805542798713,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.326074308999523
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0015152628820942482,
          "train_loss": 0.0012557429727166891,
          "validation_loss": 0.0008953476208262146,
          "train_metric": 0.00012647383846342564,
          "validation_metric": 9.017613774631172e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.350818019000144
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0019305098958284361,
          "train_loss": 0.0024097333662211895,
          "validation_loss": 0.0014671796234324574,
          "train_metric": 0.00024269950517918915,
          "validation_metric": 0.0001477689656894654,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.367718872999831
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.001991205252124928,
          "train_loss": 0.0023183785378932953,
          "validation_loss": 0.001654172781854868,
          "train_metric": 0.0002334985911147669,
          "validation_metric": 0.00016660219989717007,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.384730772999319
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0014049673663976137,
          "train_loss": 0.001289384439587593,
          "validation_loss": 0.0011746175587177277,
          "train_metric": 0.0001298620627494529,
          "validation_metric": 0.00011830317816929892,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.402044075999584
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0014139943814370781,
          "train_loss": 0.0012069448130205274,
          "validation_loss": 0.0009933608816936612,
          "train_metric": 0.0001215590673382394,
          "validation_metric": 0.00010004767682403326,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.4192993879987625
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0012138614074501675,
          "train_loss": 0.0012098526349291205,
          "validation_loss": 0.0008651012321934104,
          "train_metric": 0.0001218519319081679,
          "validation_metric": 8.712982526049018e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.437756692999756
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.001513936296760221,
          "train_loss": 0.0018569240346550941,
          "validation_loss": 0.0016664834693074226,
          "train_metric": 0.0001870226114988327,
          "validation_metric": 0.00016784209583420306,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.4551884079992305
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.00115161936628283,
          "train_loss": 0.0010892606806010008,
          "validation_loss": 0.0008254727581515908,
          "train_metric": 0.00010970633593387902,
          "validation_metric": 8.313858415931463e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.473604032998992
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0012215049337100936,
          "train_loss": 0.0010771305533125997,
          "validation_loss": 0.0008148029446601868,
          "train_metric": 0.00010848463716683909,
          "validation_metric": 8.206396887544543e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.491162987998905
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.001339199468930019,
          "train_loss": 0.0010478756157681346,
          "validation_loss": 0.0008257753797806799,
          "train_metric": 0.00010553818719927222,
          "validation_metric": 8.316907042171806e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.508390801000132
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0012278878912184155,
          "train_loss": 0.002119696931913495,
          "validation_loss": 0.0015634380979463458,
          "train_metric": 0.00021348809241317213,
          "validation_metric": 0.00015746377175673842,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.533288869999524
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0021536827880481724,
          "train_loss": 0.0021692703012377024,
          "validation_loss": 0.001296084956265986,
          "train_metric": 0.00021848097094334662,
          "validation_metric": 0.0001305369078181684,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.556240093999804
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.002478535403497517,
          "train_loss": 0.0022397900465875864,
          "validation_loss": 0.002206685720011592,
          "train_metric": 0.00022558342607226223,
          "validation_metric": 0.0002222493349108845,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.715273550000347
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.002960709261969896,
          "train_loss": 0.0018048516940325499,
          "validation_loss": 0.0015788677847012877,
          "train_metric": 0.00018177808669861406,
          "validation_metric": 0.00015901778533589095,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.735417933999997
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0022949368503759615,
          "train_loss": 0.00134421291295439,
          "validation_loss": 0.0009245164692401886,
          "train_metric": 0.00013538419443648309,
          "validation_metric": 9.311392204836011e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.755537401999391
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0013899773839511909,
          "train_loss": 0.0011031446047127247,
          "validation_loss": 0.0009181465720757842,
          "train_metric": 0.0001111046876758337,
          "validation_metric": 9.247235720977187e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.77400755599956
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0022210860679479083,
          "train_loss": 0.002029487630352378,
          "validation_loss": 0.00113514659460634,
          "train_metric": 0.00020440257503651083,
          "validation_metric": 0.00011432779865572229,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.791090011000051
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0015579893333779182,
          "train_loss": 0.0010632624616846442,
          "validation_loss": 0.000812578247860074,
          "train_metric": 0.00010708789341151714,
          "validation_metric": 8.183989848475903e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.810351296999215
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.001354978785457206,
          "train_loss": 0.0012964514316990972,
          "validation_loss": 0.001090256730094552,
          "train_metric": 0.00013057385513093323,
          "validation_metric": 0.00010980666411342099,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.829289201999927
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0011793288194894558,
          "train_loss": 0.0010471523273736238,
          "validation_loss": 0.0008010684396140277,
          "train_metric": 0.00010546534031163901,
          "validation_metric": 8.06806783657521e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.848752453999623
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.001163515980806551,
          "train_loss": 0.001036358647979796,
          "validation_loss": 0.0007641048869118094,
          "train_metric": 0.00010437823948450387,
          "validation_metric": 7.695784006500617e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.866137200999219
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.001226809513354965,
          "train_loss": 0.001550791203044355,
          "validation_loss": 0.0012130910763517022,
          "train_metric": 0.0001561899989610538,
          "validation_metric": 0.0001221780985360965,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.883371015999728
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.001968821215996286,
          "train_loss": 0.002181445714086294,
          "validation_loss": 0.0013640549732372165,
          "train_metric": 0.0002197072608396411,
          "validation_metric": 0.0001373826089547947,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.900067623999348
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.002001340628339676,
          "train_loss": 0.001255254726856947,
          "validation_loss": 0.0010116316843777895,
          "train_metric": 0.00012642465298995376,
          "validation_metric": 0.00010188783198827878,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.9166371439987415
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0013726306169701274,
          "train_loss": 0.0018920153379440308,
          "validation_loss": 0.0016072666039690375,
          "train_metric": 0.0001905568497022614,
          "validation_metric": 0.00016187800792977214,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.93334721499923
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0030681827029184205,
          "train_loss": 0.004806465469300747,
          "validation_loss": 0.0036986994091421366,
          "train_metric": 0.0004840895999222994,
          "validation_metric": 0.0003725194255821407,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.950311462998798
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0036063121806364506,
          "train_loss": 0.0030873222276568413,
          "validation_loss": 0.0029117788653820753,
          "train_metric": 0.00031094381120055914,
          "validation_metric": 0.0002932636998593807,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.967574771999352
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.002066498218482593,
          "train_loss": 0.0011575818061828613,
          "validation_loss": 0.0008597797714173794,
          "train_metric": 0.00011658739822451025,
          "validation_metric": 8.659387094667181e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 4.9848307899992506
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0015108050574781373,
          "train_loss": 0.0025991727598011494,
          "validation_loss": 0.002036432735621929,
          "train_metric": 0.00026177914696745574,
          "validation_metric": 0.00020510204194579273,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 5.002084120000291
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.0017257255494769197,
          "train_loss": 0.0010526946280151606,
          "validation_loss": 0.0008209642255678773,
          "train_metric": 0.00010602353722788393,
          "validation_metric": 8.268452802440152e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 5.01935444099945
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0014195994444889948,
          "train_loss": 0.0010000827023759484,
          "validation_loss": 0.0008208623039536178,
          "train_metric": 0.00010072467557620257,
          "validation_metric": 8.267424709629267e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 5.036491793000096
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.0015853269487706712,
          "train_loss": 0.0010731968795880675,
          "validation_loss": 0.0008083997527137399,
          "train_metric": 0.00010808845399878919,
          "validation_metric": 8.141905709635466e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 5.05375521499991
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.001302438426137087,
          "train_loss": 0.0010881817433983088,
          "validation_loss": 0.0007518643978983164,
          "train_metric": 0.00010959768405882642,
          "validation_metric": 7.572502363473177e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 5.0718990080004005
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.001248696868060506,
          "train_loss": 0.000973308109678328,
          "validation_loss": 0.0007874683942645788,
          "train_metric": 9.802803833736107e-05,
          "validation_metric": 7.931092113722116e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 5.089138154000466
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.0010679568640625803,
          "train_loss": 0.0017124403966590762,
          "validation_loss": 0.0010568912839516997,
          "train_metric": 0.00017247071082238108,
          "validation_metric": 0.00010644621215760708,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 5.106205952999517
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.00156240647265804,
          "train_loss": 0.0012235227040946484,
          "validation_loss": 0.0007952682208269835,
          "train_metric": 0.00012322873226366937,
          "validation_metric": 8.009650628082454e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 5.12329097699876
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.0012300029593461659,
          "train_loss": 0.001304383622482419,
          "validation_loss": 0.0008879990200512111,
          "train_metric": 0.00013137274072505534,
          "validation_metric": 8.943600550992414e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 5.140532305998931
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.0011370622196409386,
          "train_loss": 0.0014189599314704537,
          "validation_loss": 0.0009623033693060279,
          "train_metric": 0.00014291246770881116,
          "validation_metric": 9.691966261016205e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 5.157677597999282
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.001111720637709368,
          "train_loss": 0.0009378037066198885,
          "validation_loss": 0.0007226705783978105,
          "train_metric": 9.44521525525488e-05,
          "validation_metric": 7.278472912730649e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 5.175287444999412
        }
      ],
      "best": {
        "epoch": 240,
        "updates": 3840
      },
      "best_validation_metric": 7.278472912730649e-05,
      "config": {
        "label": "Validation plateau",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.008531396669204081
    },
    "9967999fa0f8_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.1679138047620654,
          "train_loss": 0.5831692218780518,
          "validation_loss": 0.4311050772666931,
          "train_metric": 0.058734677731990814,
          "validation_metric": 0.043419331312179565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.022690865000186022
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07002143678255379,
          "train_loss": 0.1620592623949051,
          "validation_loss": 0.13151726126670837,
          "train_metric": 0.016322016716003418,
          "validation_metric": 0.013245937414467335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.044700528000248596
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.05611764686182141,
          "train_loss": 0.04965560510754585,
          "validation_loss": 0.036141712218523026,
          "train_metric": 0.0050011309795081615,
          "validation_metric": 0.003640060778707266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06723935599984543
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.048124873894266784,
          "train_loss": 0.028714342042803764,
          "validation_loss": 0.022267421707510948,
          "train_metric": 0.0028920036274939775,
          "validation_metric": 0.0022426932118833065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0899434149996523
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.06642384477891028,
          "train_loss": 0.03015810251235962,
          "validation_loss": 0.024469908326864243,
          "train_metric": 0.0030374135822057724,
          "validation_metric": 0.0024645195808261633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11103851000007126
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.04668538738042116,
          "train_loss": 0.023140406236052513,
          "validation_loss": 0.017209015786647797,
          "train_metric": 0.0023306168150156736,
          "validation_metric": 0.0017332291463389993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13316910100002133
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.044543810945469886,
          "train_loss": 0.024159453809261322,
          "validation_loss": 0.018866777420043945,
          "train_metric": 0.002433251589536667,
          "validation_metric": 0.0019001929322257638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1548571270013781
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.049829973198939115,
          "train_loss": 0.030344858765602112,
          "validation_loss": 0.024414196610450745,
          "train_metric": 0.0030562228057533503,
          "validation_metric": 0.0024589085951447487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17557536400090612
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.03958173515275121,
          "train_loss": 0.02075006067752838,
          "validation_loss": 0.016649581491947174,
          "train_metric": 0.002089870162308216,
          "validation_metric": 0.0016768850618973374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21152052800061938
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.048497957177460194,
          "train_loss": 0.02191157639026642,
          "validation_loss": 0.01727931760251522,
          "train_metric": 0.0022068535909056664,
          "validation_metric": 0.0017403094097971916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26615029599997797
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04584456601878628,
          "train_loss": 0.030903693288564682,
          "validation_loss": 0.026309259235858917,
          "train_metric": 0.0031125068198889494,
          "validation_metric": 0.002649772446602583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3212331570011884
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.043977227876894176,
          "train_loss": 0.020122937858104706,
          "validation_loss": 0.015020892024040222,
          "train_metric": 0.002026708796620369,
          "validation_metric": 0.0015128492377698421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3657941470009973
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.02477082674158737,
          "train_loss": 0.018304593861103058,
          "validation_loss": 0.015012234449386597,
          "train_metric": 0.0018435717793181539,
          "validation_metric": 0.0015119772870093584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3881804020002164
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.04737904708599672,
          "train_loss": 0.05622466653585434,
          "validation_loss": 0.040056660771369934,
          "train_metric": 0.0056627425365149975,
          "validation_metric": 0.004034359939396381,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4098793740013207
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.061866504955105484,
          "train_loss": 0.0848754271864891,
          "validation_loss": 0.06052950397133827,
          "train_metric": 0.008548342622816563,
          "validation_metric": 0.006096309050917625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43059081800129206
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.049686715356074274,
          "train_loss": 0.017621345818042755,
          "validation_loss": 0.01429538894444704,
          "train_metric": 0.001774757169187069,
          "validation_metric": 0.0014397792983800173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45214559600026405
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03551947826053947,
          "train_loss": 0.02655225805938244,
          "validation_loss": 0.02205834537744522,
          "train_metric": 0.002674246206879616,
          "validation_metric": 0.0022216355428099632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4727134630011278
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.04077337629860267,
          "train_loss": 0.02229268290102482,
          "validation_loss": 0.018882783129811287,
          "train_metric": 0.0022452373523265123,
          "validation_metric": 0.0019018049351871014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4933449260006455
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.04355371929705143,
          "train_loss": 0.0166607778519392,
          "validation_loss": 0.013903462328016758,
          "train_metric": 0.0016780123114585876,
          "validation_metric": 0.0014003058895468712,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5249253600013617
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.04053163173375651,
          "train_loss": 0.01519621629267931,
          "validation_loss": 0.01270003616809845,
          "train_metric": 0.0015305073466151953,
          "validation_metric": 0.0012791010085493326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5768126000002667
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.033880555420182645,
          "train_loss": 0.015688393265008926,
          "validation_loss": 0.012057787738740444,
          "train_metric": 0.0015800775727257133,
          "validation_metric": 0.0012144162319600582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5995703250009683
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.033668921678327024,
          "train_loss": 0.042417749762535095,
          "validation_loss": 0.03647554665803909,
          "train_metric": 0.0042721605859696865,
          "validation_metric": 0.0036736836191266775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6198986020008306
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.04606936406344175,
          "train_loss": 0.014555183239281178,
          "validation_loss": 0.010964584536850452,
          "train_metric": 0.0014659448061138391,
          "validation_metric": 0.001104312832467258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6410107680003421
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.04876400780631229,
          "train_loss": 0.030287690460681915,
          "validation_loss": 0.020612508058547974,
          "train_metric": 0.0030504653695970774,
          "validation_metric": 0.0020760162733495235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6612388230005308
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.03669262275798246,
          "train_loss": 0.015138848684728146,
          "validation_loss": 0.009774107486009598,
          "train_metric": 0.0015247296541929245,
          "validation_metric": 0.0009844122687354684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.683120470999711
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.03052059543551877,
          "train_loss": 0.013456888496875763,
          "validation_loss": 0.009625928476452827,
          "train_metric": 0.0013553284807130694,
          "validation_metric": 0.0009694883483462036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7053805680006917
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.037601628922857344,
          "train_loss": 0.017850065603852272,
          "validation_loss": 0.0130152003839612,
          "train_metric": 0.001797793316654861,
          "validation_metric": 0.001310843275859952,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7261827929996798
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.024997590691782534,
          "train_loss": 0.016174279153347015,
          "validation_loss": 0.01239065919071436,
          "train_metric": 0.0016290141502395272,
          "validation_metric": 0.0012479416327551007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7471604270012904
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.02897122738067992,
          "train_loss": 0.053334832191467285,
          "validation_loss": 0.04158470779657364,
          "train_metric": 0.005371689330786467,
          "validation_metric": 0.004188260063529015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7685358730013832
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.03891480795573443,
          "train_loss": 0.02195945754647255,
          "validation_loss": 0.016714870929718018,
          "train_metric": 0.002211676212027669,
          "validation_metric": 0.00168346066493541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7896630390005157
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.035517099138814956,
          "train_loss": 0.012664569541811943,
          "validation_loss": 0.010139083489775658,
          "train_metric": 0.0012755290372297168,
          "validation_metric": 0.0010211713379248977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8104138620001322
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.03585810569347814,
          "train_loss": 0.014967112801969051,
          "validation_loss": 0.01061791367828846,
          "train_metric": 0.0015074328985065222,
          "validation_metric": 0.0010693974327296019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8309332200005883
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.028600875986739993,
          "train_loss": 0.012109045870602131,
          "validation_loss": 0.009384501725435257,
          "train_metric": 0.0012195787858217955,
          "validation_metric": 0.0009451725636608899,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8530061010005738
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.0324279204942286,
          "train_loss": 0.04353649914264679,
          "validation_loss": 0.028165917843580246,
          "train_metric": 0.0043848371133208275,
          "validation_metric": 0.002836768049746752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.873929744000634
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.04115924039797392,
          "train_loss": 0.019086651504039764,
          "validation_loss": 0.014985820278525352,
          "train_metric": 0.0019223374547436833,
          "validation_metric": 0.0015093170804902911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8953147560005164
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.04461798630654812,
          "train_loss": 0.027782723307609558,
          "validation_loss": 0.02101128175854683,
          "train_metric": 0.00279817427508533,
          "validation_metric": 0.0021161793265491724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9159410000011121
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.040300877357367426,
          "train_loss": 0.011074727401137352,
          "validation_loss": 0.008553712628781796,
          "train_metric": 0.0011154059320688248,
          "validation_metric": 0.0008614985854364932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9375399790005758
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.035434633609838784,
          "train_loss": 0.02648579701781273,
          "validation_loss": 0.023315703496336937,
          "train_metric": 0.0026675525587052107,
          "validation_metric": 0.002348272129893303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9584629250002763
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.05380376271205023,
          "train_loss": 0.01574905775487423,
          "validation_loss": 0.0107577508315444,
          "train_metric": 0.0015861872816458344,
          "validation_metric": 0.0010834812419489026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9786364419996971
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.03492005093721673,
          "train_loss": 0.016294078901410103,
          "validation_loss": 0.013399814255535603,
          "train_metric": 0.001641079899854958,
          "validation_metric": 0.0013495800085365772,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1199264300012146
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.03413265885319561,
          "train_loss": 0.013220510445535183,
          "validation_loss": 0.010732715018093586,
          "train_metric": 0.0013315214309841394,
          "validation_metric": 0.001080959802493453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1636393680000765
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.044461335928644985,
          "train_loss": 0.023716086521744728,
          "validation_loss": 0.021218806505203247,
          "train_metric": 0.002388597000390291,
          "validation_metric": 0.0021370803005993366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1844211450006696
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.031767549633514136,
          "train_loss": 0.018811456859111786,
          "validation_loss": 0.016377173364162445,
          "train_metric": 0.0018946212949231267,
          "validation_metric": 0.0016494488809257746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.205737649999719
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.033554797468241304,
          "train_loss": 0.009173491969704628,
          "validation_loss": 0.006926172412931919,
          "train_metric": 0.0009239205392077565,
          "validation_metric": 0.0006975787109695375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.227398824001284
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.031966720314812846,
          "train_loss": 0.017387688159942627,
          "validation_loss": 0.014267228543758392,
          "train_metric": 0.0017512240447103977,
          "validation_metric": 0.0014369429554790258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2485105490013666
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.03790816734544933,
          "train_loss": 0.00975831225514412,
          "validation_loss": 0.00724437739700079,
          "train_metric": 0.0009828215697780252,
          "validation_metric": 0.0007296271505765617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.269673128001159
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0315881401184015,
          "train_loss": 0.010181362740695477,
          "validation_loss": 0.00767604960128665,
          "train_metric": 0.0010254295775666833,
          "validation_metric": 0.0007731036166660488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.290586238999822
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.043640270858304575,
          "train_loss": 0.013429512269794941,
          "validation_loss": 0.010790303349494934,
          "train_metric": 0.0013525714166462421,
          "validation_metric": 0.0010867598466575146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.312146563999704
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.024616379261715338,
          "train_loss": 0.013638966716825962,
          "validation_loss": 0.011869535781443119,
          "train_metric": 0.0013736665714532137,
          "validation_metric": 0.0011954562505707145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3330613550006092
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.0278749217686709,
          "train_loss": 0.008291922509670258,
          "validation_loss": 0.006158225238323212,
          "train_metric": 0.0008351320284418762,
          "validation_metric": 0.0006202338263392448,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3551488290013367
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.02732852633926086,
          "train_loss": 0.011660993099212646,
          "validation_loss": 0.010281147435307503,
          "train_metric": 0.001174452598206699,
          "validation_metric": 0.001035479479469359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3768877900001826
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.025268875469919294,
          "train_loss": 0.011895044706761837,
          "validation_loss": 0.009626802988350391,
          "train_metric": 0.001198025420308113,
          "validation_metric": 0.0009695764165371656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.397828909000964
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.040222966752480716,
          "train_loss": 0.012431349605321884,
          "validation_loss": 0.008838512934744358,
          "train_metric": 0.001252040034160018,
          "validation_metric": 0.0008901826804503798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4200376349999715
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.024146173062035814,
          "train_loss": 0.009605194441974163,
          "validation_loss": 0.007469425443559885,
          "train_metric": 0.0009673999738879502,
          "validation_metric": 0.0007522932137362659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4420175140003266
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.034781657508574426,
          "train_loss": 0.009690068662166595,
          "validation_loss": 0.007669575046747923,
          "train_metric": 0.0009759481763467193,
          "validation_metric": 0.0007724514580331743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4646526830001676
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.03246967325685546,
          "train_loss": 0.008762728422880173,
          "validation_loss": 0.006584405433386564,
          "train_metric": 0.0008825499098747969,
          "validation_metric": 0.0006631572032347322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4873105560000113
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.026239555096253753,
          "train_loss": 0.008539458736777306,
          "validation_loss": 0.0061043365858495235,
          "train_metric": 0.0008600629516877234,
          "validation_metric": 0.0006148063694126904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5106469220008876
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.04131140420213342,
          "train_loss": 0.015017633326351643,
          "validation_loss": 0.013301929458975792,
          "train_metric": 0.001512521062977612,
          "validation_metric": 0.001339721609838307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.53196369699981
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.02920761168934405,
          "train_loss": 0.011533601209521294,
          "validation_loss": 0.007978016510605812,
          "train_metric": 0.001161621999926865,
          "validation_metric": 0.0008035165956243873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5537545019997197
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.02359271905152127,
          "train_loss": 0.012317276559770107,
          "validation_loss": 0.009677289985120296,
          "train_metric": 0.0012405510060489178,
          "validation_metric": 0.0009746613213792443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5750605270004598
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.02675862197065726,
          "train_loss": 0.007876105606555939,
          "validation_loss": 0.006131825968623161,
          "train_metric": 0.0007932524313218892,
          "validation_metric": 0.0006175750750117004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5961926609998045
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.025390354334376752,
          "train_loss": 0.009267187677323818,
          "validation_loss": 0.006700033321976662,
          "train_metric": 0.0009333571651950479,
          "validation_metric": 0.0006748028099536896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6170333709997067
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.03439182380679995,
          "train_loss": 0.015124981291592121,
          "validation_loss": 0.012224029749631882,
          "train_metric": 0.001523332903161645,
          "validation_metric": 0.0012311593163758516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6384517630012851
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.027365016110707074,
          "train_loss": 0.011141716502606869,
          "validation_loss": 0.008401593193411827,
          "train_metric": 0.0011221527820453048,
          "validation_metric": 0.0008461777470074594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6594119540004613
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.021839445980731398,
          "train_loss": 0.008064255118370056,
          "validation_loss": 0.006157503463327885,
          "train_metric": 0.0008122024009935558,
          "validation_metric": 0.0006201611831784248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6802812710011494
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.021816835331264883,
          "train_loss": 0.012621586211025715,
          "validation_loss": 0.009771988727152348,
          "train_metric": 0.0012711999006569386,
          "validation_metric": 0.0009841988794505596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7012274120006623
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.03300009804661386,
          "train_loss": 0.007725333329290152,
          "validation_loss": 0.005427040159702301,
          "train_metric": 0.0007780672749504447,
          "validation_metric": 0.0005465915892273188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7338696429997071
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.021670665650162846,
          "train_loss": 0.011505154892802238,
          "validation_loss": 0.008886481635272503,
          "train_metric": 0.0011587570188567042,
          "validation_metric": 0.0008950139163061976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7566264610013604
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.03034562958055176,
          "train_loss": 0.013439375907182693,
          "validation_loss": 0.00992676056921482,
          "train_metric": 0.0013535649050027132,
          "validation_metric": 0.000999787007458508,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7778897200005304
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.04452735168160871,
          "train_loss": 0.010864109732210636,
          "validation_loss": 0.008415358141064644,
          "train_metric": 0.0010941933142021298,
          "validation_metric": 0.0008475641370750964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7991107680009009
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.025061569991521537,
          "train_loss": 0.008418632671236992,
          "validation_loss": 0.0063396031036973,
          "train_metric": 0.0008478938252665102,
          "validation_metric": 0.0006385016022250056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8207013280007232
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.03396414627786726,
          "train_loss": 0.008620799519121647,
          "validation_loss": 0.006843735929578543,
          "train_metric": 0.0008682551560923457,
          "validation_metric": 0.0006892759702168405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.84193997300099
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.03721781843341887,
          "train_loss": 0.012987720780074596,
          "validation_loss": 0.010280090384185314,
          "train_metric": 0.0013080756179988384,
          "validation_metric": 0.001035373075865209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9208253050001076
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0308240779559128,
          "train_loss": 0.009648801758885384,
          "validation_loss": 0.008086887188255787,
          "train_metric": 0.000971791916526854,
          "validation_metric": 0.0008144816383719444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9990297870008362
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.027479368494823575,
          "train_loss": 0.008916768245398998,
          "validation_loss": 0.006634051911532879,
          "train_metric": 0.0008980642887763679,
          "validation_metric": 0.0006681574159301817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.07689507699979
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.03556239866884425,
          "train_loss": 0.011050021275877953,
          "validation_loss": 0.008209588006138802,
          "train_metric": 0.0011129177873954177,
          "validation_metric": 0.0008268396486528218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4534668260002945
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.03389069592230953,
          "train_loss": 0.00955924391746521,
          "validation_loss": 0.008126853033900261,
          "train_metric": 0.0009627719991840422,
          "validation_metric": 0.0008185067563317716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.560999985000308
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.029371395416092128,
          "train_loss": 0.008399340324103832,
          "validation_loss": 0.00665842043235898,
          "train_metric": 0.0008459508535452187,
          "validation_metric": 0.0006706116837449372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6580806100009795
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.026820576516911387,
          "train_loss": 0.008390912786126137,
          "validation_loss": 0.006347248330712318,
          "train_metric": 0.0008451020112261176,
          "validation_metric": 0.000639271573163569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.736409277000348
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.025236848858185112,
          "train_loss": 0.008142550475895405,
          "validation_loss": 0.006563831120729446,
          "train_metric": 0.0008200877928175032,
          "validation_metric": 0.0006610850105062127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.813765771001272
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.03616610309109092,
          "train_loss": 0.007977547124028206,
          "validation_loss": 0.005879503674805164,
          "train_metric": 0.0008034693892113864,
          "validation_metric": 0.0005921621341258287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8964050699996733
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.028099152812501416,
          "train_loss": 0.010970138013362885,
          "validation_loss": 0.008743153885006905,
          "train_metric": 0.0011048719752579927,
          "validation_metric": 0.0008805784746073186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9737902600008965
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.04241425730288029,
          "train_loss": 0.023524338379502296,
          "validation_loss": 0.015465587377548218,
          "train_metric": 0.0023692853283137083,
          "validation_metric": 0.0015576372388750315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0519350619997567
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.03716611984418705,
          "train_loss": 0.014319912530481815,
          "validation_loss": 0.012140016071498394,
          "train_metric": 0.0014422491658478975,
          "validation_metric": 0.0012226979015395045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.13015667900072
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.030298508005216718,
          "train_loss": 0.007873867638409138,
          "validation_loss": 0.0060581485740840435,
          "train_metric": 0.0007930270512588322,
          "validation_metric": 0.0006101545295678079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2081196180006373
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.030128406244330108,
          "train_loss": 0.00831285398453474,
          "validation_loss": 0.006724537815898657,
          "train_metric": 0.0008372401352971792,
          "validation_metric": 0.0006772708729840815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2893188960006228
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.02935460559092462,
          "train_loss": 0.017227191478013992,
          "validation_loss": 0.013011032715439796,
          "train_metric": 0.0017350594280287623,
          "validation_metric": 0.001310423482209444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.372987071999887
        }
      ],
      "best": {
        "epoch": 67,
        "updates": 1072
      },
      "best_validation_metric": 0.0005465915892273188,
      "config": {
        "label": "BatchNorm",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "batch",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.02337929830485335
    },
    "c160e9485e77_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.26856481935828924,
          "train_loss": 0.1564040631055832,
          "validation_loss": 0.12997667491436005,
          "train_metric": 0.01575244404375553,
          "validation_metric": 0.013090775348246098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01367428000048676
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.12438978999853134,
          "train_loss": 0.09684181958436966,
          "validation_loss": 0.07960475236177444,
          "train_metric": 0.009753554128110409,
          "validation_metric": 0.00801749899983406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.026724892000856926
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.08788533369079232,
          "train_loss": 0.08004368096590042,
          "validation_loss": 0.06172280013561249,
          "train_metric": 0.008061707019805908,
          "validation_metric": 0.006216494366526604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03975142700073775
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.07180167478509247,
          "train_loss": 0.06588040292263031,
          "validation_loss": 0.05277034267783165,
          "train_metric": 0.006635232828557491,
          "validation_metric": 0.005314836278557777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05303521899986663
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.06279360281769186,
          "train_loss": 0.06062968075275421,
          "validation_loss": 0.04795650765299797,
          "train_metric": 0.006106399931013584,
          "validation_metric": 0.004830004181712866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06600744500065048
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.05910049087833613,
          "train_loss": 0.057395197451114655,
          "validation_loss": 0.04686475545167923,
          "train_metric": 0.0057806349359452724,
          "validation_metric": 0.0047200461849570274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07904163400053221
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05682261206675321,
          "train_loss": 0.05291968584060669,
          "validation_loss": 0.04275095835328102,
          "train_metric": 0.005329877138137817,
          "validation_metric": 0.004305719863623381,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09231628900124633
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.05156957160215825,
          "train_loss": 0.047955647110939026,
          "validation_loss": 0.03872738406062126,
          "train_metric": 0.004829916637390852,
          "validation_metric": 0.003900480456650257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10511792600118497
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04769943910650909,
          "train_loss": 0.052331093698740005,
          "validation_loss": 0.04272550716996193,
          "train_metric": 0.005270596593618393,
          "validation_metric": 0.0043031563982367516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11744902500140597
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05117311293724924,
          "train_loss": 0.044052157551050186,
          "validation_loss": 0.03668962046504021,
          "train_metric": 0.004436771385371685,
          "validation_metric": 0.003695244435220957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1301008290010941
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04260836902540177,
          "train_loss": 0.037946879863739014,
          "validation_loss": 0.030791016295552254,
          "train_metric": 0.0038218707777559757,
          "validation_metric": 0.00310115865431726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1433525299999019
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.03768860793206841,
          "train_loss": 0.034922730177640915,
          "validation_loss": 0.028478503227233887,
          "train_metric": 0.0035172896459698677,
          "validation_metric": 0.0028682504780590534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1560571110003366
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.034489742189180106,
          "train_loss": 0.035768602043390274,
          "validation_loss": 0.02817763388156891,
          "train_metric": 0.0036024826113134623,
          "validation_metric": 0.0028379480354487896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.169424716001231
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03205141890794039,
          "train_loss": 0.02890515699982643,
          "validation_loss": 0.023330947384238243,
          "train_metric": 0.002911221468821168,
          "validation_metric": 0.0023498076479882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.182704670000021
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.029147319321054965,
          "train_loss": 0.025692367926239967,
          "validation_loss": 0.020649176090955734,
          "train_metric": 0.0025876411236822605,
          "validation_metric": 0.0020797092001885176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1958621360008692
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.025168986758217216,
          "train_loss": 0.024655364453792572,
          "validation_loss": 0.019782746210694313,
          "train_metric": 0.002483198419213295,
          "validation_metric": 0.001992445671930909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21038170199972228
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.023386888089589775,
          "train_loss": 0.023372037336230278,
          "validation_loss": 0.01934860274195671,
          "train_metric": 0.0023539457470178604,
          "validation_metric": 0.001948720309883356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22314609100067173
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.023384401574730873,
          "train_loss": 0.02144448645412922,
          "validation_loss": 0.016917847096920013,
          "train_metric": 0.0021598103921860456,
          "validation_metric": 0.0017039034282788634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23586010500002885
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.021844984672497958,
          "train_loss": 0.02267218753695488,
          "validation_loss": 0.01874099299311638,
          "train_metric": 0.00228345999494195,
          "validation_metric": 0.0018875241512432694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24828041900036624
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.027959421277046204,
          "train_loss": 0.020730987191200256,
          "validation_loss": 0.017180578783154488,
          "train_metric": 0.002087949076667428,
          "validation_metric": 0.0017303648637607694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2604373840003973
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.022898548166267574,
          "train_loss": 0.01903417333960533,
          "validation_loss": 0.015614808537065983,
          "train_metric": 0.0019170523155480623,
          "validation_metric": 0.001572666340507567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27259217699975125
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.024446852505207062,
          "train_loss": 0.01983354426920414,
          "validation_loss": 0.016570474952459335,
          "train_metric": 0.001997562125325203,
          "validation_metric": 0.0016689177136868238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28516623600080493
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.023444161575753242,
          "train_loss": 0.01750955916941166,
          "validation_loss": 0.014578438363969326,
          "train_metric": 0.001763498643413186,
          "validation_metric": 0.0014682868495583534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2982123700003285
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.02048990296316333,
          "train_loss": 0.017068009823560715,
          "validation_loss": 0.013973916880786419,
          "train_metric": 0.0017190274083986878,
          "validation_metric": 0.0014074018690735102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3113274099996488
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.021343321888707578,
          "train_loss": 0.01674187183380127,
          "validation_loss": 0.013968662358820438,
          "train_metric": 0.0016861800104379654,
          "validation_metric": 0.0014068724121898413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3245231450000574
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.019522768445312977,
          "train_loss": 0.015722976997494698,
          "validation_loss": 0.012727980501949787,
          "train_metric": 0.0015835606027394533,
          "validation_metric": 0.0012819155817851424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3381643150005402
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.017787194927223027,
          "train_loss": 0.01496818009763956,
          "validation_loss": 0.012338947504758835,
          "train_metric": 0.0015075402334332466,
          "validation_metric": 0.0012427335605025291,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3519348860008904
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015520317858317867,
          "train_loss": 0.014398195780813694,
          "validation_loss": 0.011873327195644379,
          "train_metric": 0.0014501335099339485,
          "validation_metric": 0.0011958378599956632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36446159000115586
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.01736800718936138,
          "train_loss": 0.02331288531422615,
          "validation_loss": 0.019839972257614136,
          "train_metric": 0.002347988309338689,
          "validation_metric": 0.0019982093945145607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37658530400040036
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.016681191860698164,
          "train_loss": 0.0166134312748909,
          "validation_loss": 0.013425116427242756,
          "train_metric": 0.0016732439398765564,
          "validation_metric": 0.0013521285727620125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38897704600094585
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.016136521328007802,
          "train_loss": 0.022352367639541626,
          "validation_loss": 0.019709279760718346,
          "train_metric": 0.0022512483410537243,
          "validation_metric": 0.001985046546906233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40113031700093416
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.02121598564554006,
          "train_loss": 0.021724993363022804,
          "validation_loss": 0.0183981005102396,
          "train_metric": 0.0021880618296563625,
          "validation_metric": 0.0018529893131926656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41360551599973405
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.018153683486161754,
          "train_loss": 0.013500056229531765,
          "validation_loss": 0.011093583889305592,
          "train_metric": 0.0013596763601526618,
          "validation_metric": 0.0011173051316291094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4263744689997111
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013416525995125994,
          "train_loss": 0.02082075923681259,
          "validation_loss": 0.01799742504954338,
          "train_metric": 0.0020969905890524387,
          "validation_metric": 0.0018126347567886114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43834896000043955
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.015258144558174536,
          "train_loss": 0.011574605479836464,
          "validation_loss": 0.009676911868155003,
          "train_metric": 0.0011657518334686756,
          "validation_metric": 0.0009746231371536851,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4509791390009923
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.01262954126286786,
          "train_loss": 0.01135251298546791,
          "validation_loss": 0.009591642767190933,
          "train_metric": 0.001143383327871561,
          "validation_metric": 0.0009660351788625121,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46414523600105895
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013332895614439622,
          "train_loss": 0.012162867933511734,
          "validation_loss": 0.009434236213564873,
          "train_metric": 0.0012249995488673449,
          "validation_metric": 0.0009501817403361201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.477073719001055
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011938160372665152,
          "train_loss": 0.010830353014171124,
          "validation_loss": 0.009401759132742882,
          "train_metric": 0.0010907934047281742,
          "validation_metric": 0.0009469108190387487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49033666200011794
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.012070931261405349,
          "train_loss": 0.015150758437812328,
          "validation_loss": 0.01039953250437975,
          "train_metric": 0.0015259289648383856,
          "validation_metric": 0.0010474028531461954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.502750483001364
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.013272067866637371,
          "train_loss": 0.014177961274981499,
          "validation_loss": 0.013008009642362595,
          "train_metric": 0.0014279524330049753,
          "validation_metric": 0.0013101191725581884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5151775020003697
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.011496021936181933,
          "train_loss": 0.010944796726107597,
          "validation_loss": 0.008532698266208172,
          "train_metric": 0.001102319685742259,
          "validation_metric": 0.0008593820966780186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5280897749998985
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010176989526371472,
          "train_loss": 0.009800639934837818,
          "validation_loss": 0.008510386571288109,
          "train_metric": 0.0009870845824480057,
          "validation_metric": 0.0008571348735131323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5413189500013686
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009940968375303783,
          "train_loss": 0.01564965210855007,
          "validation_loss": 0.012279627844691277,
          "train_metric": 0.0015761756803840399,
          "validation_metric": 0.0012367591261863708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5536770360013179
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.011834226257633418,
          "train_loss": 0.009067519567906857,
          "validation_loss": 0.007470233365893364,
          "train_metric": 0.0009132472914643586,
          "validation_metric": 0.000752374529838562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5660542080004234
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.009457082094741054,
          "train_loss": 0.009713280014693737,
          "validation_loss": 0.008271859027445316,
          "train_metric": 0.0009782857960090041,
          "validation_metric": 0.0008331112912856042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5784517230003985
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.009470705583225936,
          "train_loss": 0.010028257966041565,
          "validation_loss": 0.008909277617931366,
          "train_metric": 0.0010100093204528093,
          "validation_metric": 0.0008973099174909294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5907422279997263
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011381873351638205,
          "train_loss": 0.01868240162730217,
          "validation_loss": 0.013789458200335503,
          "train_metric": 0.0018816230585798621,
          "validation_metric": 0.0013888237299397588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6029039650002233
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.011791408498538658,
          "train_loss": 0.00844588689506054,
          "validation_loss": 0.007234173361212015,
          "train_metric": 0.0008506389567628503,
          "validation_metric": 0.0007285995525307953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6156898219996947
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.008761773438891396,
          "train_loss": 0.007501610554754734,
          "validation_loss": 0.005921256262809038,
          "train_metric": 0.0007555347983725369,
          "validation_metric": 0.0005963671719655395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6286955080013286
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.008461370060103945,
          "train_loss": 0.00771644152700901,
          "validation_loss": 0.005776180885732174,
          "train_metric": 0.0007771716918796301,
          "validation_metric": 0.00058175582671538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6414573080000991
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.00915710430854233,
          "train_loss": 0.007970030419528484,
          "validation_loss": 0.006049149204045534,
          "train_metric": 0.000802712282165885,
          "validation_metric": 0.0006092481780797243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6677015820005181
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.013394176115980372,
          "train_loss": 0.008955839090049267,
          "validation_loss": 0.007901805453002453,
          "train_metric": 0.0009019994176924229,
          "validation_metric": 0.0007958408677950501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6991119610011083
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008483909216010943,
          "train_loss": 0.00696143414825201,
          "validation_loss": 0.006234260741621256,
          "train_metric": 0.0007011301931925118,
          "validation_metric": 0.0006278919172473252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7240378030001011
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.007608028157847002,
          "train_loss": 0.00789773277938366,
          "validation_loss": 0.007182017900049686,
          "train_metric": 0.000795430620200932,
          "validation_metric": 0.0007233464857563376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7529138070003683
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.009676984424004331,
          "train_loss": 0.006707190070301294,
          "validation_loss": 0.005720959044992924,
          "train_metric": 0.000675523653626442,
          "validation_metric": 0.0005761940265074372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7859767900008592
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.013093463698169217,
          "train_loss": 0.009094911627471447,
          "validation_loss": 0.0067943693138659,
          "train_metric": 0.000916006276383996,
          "validation_metric": 0.000684303988236934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8150356060014019
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008355949074029922,
          "train_loss": 0.006546635180711746,
          "validation_loss": 0.005608290433883667,
          "train_metric": 0.0006593531579710543,
          "validation_metric": 0.0005648464430123568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9263905120005802
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.008330995930009522,
          "train_loss": 0.007372529245913029,
          "validation_loss": 0.006412808783352375,
          "train_metric": 0.0007425341755151749,
          "validation_metric": 0.0006458745338022709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9386230560012336
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.0071738916740287095,
          "train_loss": 0.006414851639419794,
          "validation_loss": 0.004838612861931324,
          "train_metric": 0.0006460804725065827,
          "validation_metric": 0.000487327401060611,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9520773439999175
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006638482183916494,
          "train_loss": 0.006604726426303387,
          "validation_loss": 0.00576860411092639,
          "train_metric": 0.0006652038428001106,
          "validation_metric": 0.0005809926660731435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.964057089000562
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0066618486162042245,
          "train_loss": 0.005912338849157095,
          "validation_loss": 0.004710252862423658,
          "train_metric": 0.0005954691441729665,
          "validation_metric": 0.00047439939226023853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9763327870005014
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.006402640232408885,
          "train_loss": 0.005535063799470663,
          "validation_loss": 0.004463979974389076,
          "train_metric": 0.0005574714159592986,
          "validation_metric": 0.00044959576916880906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9886853600000904
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.008135063471854664,
          "train_loss": 0.013005940243601799,
          "validation_loss": 0.008761103264987469,
          "train_metric": 0.0013099105563014746,
          "validation_metric": 0.0008823861717246473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0006337699996948
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.009909371248795651,
          "train_loss": 0.0066342223435640335,
          "validation_loss": 0.005919354967772961,
          "train_metric": 0.0006681746453978121,
          "validation_metric": 0.000596175785176456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.01253048900071
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.006464159785537049,
          "train_loss": 0.006166266743093729,
          "validation_loss": 0.005600084085017443,
          "train_metric": 0.0006210438441485167,
          "validation_metric": 0.0005640199524350464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0243558160000248
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.006254389823880047,
          "train_loss": 0.005562745966017246,
          "validation_loss": 0.004616580903530121,
          "train_metric": 0.000560259388294071,
          "validation_metric": 0.0004649651818908751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.03605451400108
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007393788968329318,
          "train_loss": 0.007429416757076979,
          "validation_loss": 0.0066094896756112576,
          "train_metric": 0.0007482636137865484,
          "validation_metric": 0.0006656835321336985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.048049376000563
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.007559565987321548,
          "train_loss": 0.0062525589019060135,
          "validation_loss": 0.004652619361877441,
          "train_metric": 0.0006297348299995065,
          "validation_metric": 0.0004685947496909648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0596301610003138
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.00991035062179435,
          "train_loss": 0.005598491057753563,
          "validation_loss": 0.004576039966195822,
          "train_metric": 0.000563859473913908,
          "validation_metric": 0.0004608819435816258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0719300050004676
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0077747701507178135,
          "train_loss": 0.008254184387624264,
          "validation_loss": 0.007471392396837473,
          "train_metric": 0.0008313310681842268,
          "validation_metric": 0.0007524911779910326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.083656750000955
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0068277411701274104,
          "train_loss": 0.005832648370414972,
          "validation_loss": 0.004067858215421438,
          "train_metric": 0.0005874428898096085,
          "validation_metric": 0.0004096997145097703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0958436359996995
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.005976636792183854,
          "train_loss": 0.004917902406305075,
          "validation_loss": 0.003659688401967287,
          "train_metric": 0.0004953131428919733,
          "validation_metric": 0.00036859040847048163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1080891720012005
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.005211313546169549,
          "train_loss": 0.006019643507897854,
          "validation_loss": 0.005548641551285982,
          "train_metric": 0.0006062763859517872,
          "validation_metric": 0.0005588388885371387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.120005521001076
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0058156218292424455,
          "train_loss": 0.005504276137799025,
          "validation_loss": 0.004302618093788624,
          "train_metric": 0.0005543704610317945,
          "validation_metric": 0.00043334392830729485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1316695460009214
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.007377389483735897,
          "train_loss": 0.00461285375058651,
          "validation_loss": 0.003916803281754255,
          "train_metric": 0.0004645896842703223,
          "validation_metric": 0.00039448606548830867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1434358349997638
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.005634413566440344,
          "train_loss": 0.004895028658211231,
          "validation_loss": 0.0038695107214152813,
          "train_metric": 0.0004930093418806791,
          "validation_metric": 0.00038972293259575963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1558733239999128
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.006400102196494117,
          "train_loss": 0.005606420803815126,
          "validation_loss": 0.0048522925935685635,
          "train_metric": 0.0005646581994369626,
          "validation_metric": 0.000488705059979111,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1681766600013361
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.005273940318147652,
          "train_loss": 0.004276078660041094,
          "validation_loss": 0.0036198454909026623,
          "train_metric": 0.0004306709743104875,
          "validation_metric": 0.00036457754322327673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.180629268001212
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006028086645528674,
          "train_loss": 0.004306692164391279,
          "validation_loss": 0.0035555297508835793,
          "train_metric": 0.0004337542341090739,
          "validation_metric": 0.0003580999036785215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.193389236999792
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0047607290325686336,
          "train_loss": 0.004225960932672024,
          "validation_loss": 0.003378460183739662,
          "train_metric": 0.000425623293267563,
          "validation_metric": 0.000340266153216362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.206115631001012
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006771579668566119,
          "train_loss": 0.009409100748598576,
          "validation_loss": 0.008010607212781906,
          "train_metric": 0.0009476501727476716,
          "validation_metric": 0.0008067990420386195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2180737020007655
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.008914235106203705,
          "train_loss": 0.006813282147049904,
          "validation_loss": 0.005993482656776905,
          "train_metric": 0.0006862088339403272,
          "validation_metric": 0.0006036416161805391,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.229658621001363
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.007516566343838349,
          "train_loss": 0.004129443783313036,
          "validation_loss": 0.0033012607600539923,
          "train_metric": 0.0004159024974796921,
          "validation_metric": 0.00033249089028686285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2424922980007977
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.004796326044015586,
          "train_loss": 0.0052872709929943085,
          "validation_loss": 0.003630508203059435,
          "train_metric": 0.0005325145903043449,
          "validation_metric": 0.0003656514745671302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2547038970005815
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.007441916277457494,
          "train_loss": 0.01576286181807518,
          "validation_loss": 0.013864773325622082,
          "train_metric": 0.001587577979080379,
          "validation_metric": 0.0013964091194793582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2694479689998843
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.011092678498243913,
          "train_loss": 0.020933406427502632,
          "validation_loss": 0.014481311663985252,
          "train_metric": 0.002108336193487048,
          "validation_metric": 0.001458504586480558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.284598508000272
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.01372989741503261,
          "train_loss": 0.005516444332897663,
          "validation_loss": 0.004566176328808069,
          "train_metric": 0.0005555959651246667,
          "validation_metric": 0.00045988854253664613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2964706570001
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0059394005802460015,
          "train_loss": 0.004217673093080521,
          "validation_loss": 0.0034367325715720654,
          "train_metric": 0.00042478859541006386,
          "validation_metric": 0.0003461351152509451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3088236439998582
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004478430419112556,
          "train_loss": 0.004131665453314781,
          "validation_loss": 0.003314823843538761,
          "train_metric": 0.000416126218624413,
          "validation_metric": 0.0003338569076731801,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3220288749998872
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004198886701487936,
          "train_loss": 0.0036892308853566647,
          "validation_loss": 0.002885697642341256,
          "train_metric": 0.00037156580947339535,
          "validation_metric": 0.00029063684633001685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3350758859996859
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.005015779821405886,
          "train_loss": 0.0042351605370640755,
          "validation_loss": 0.0035789627581834793,
          "train_metric": 0.00042654984281398356,
          "validation_metric": 0.0003604600206017494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3472643149998476
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004449703123100335,
          "train_loss": 0.00422236742451787,
          "validation_loss": 0.003265146864578128,
          "train_metric": 0.00042526141623966396,
          "validation_metric": 0.00032885369728319347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3597442980008054
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004666383872972801,
          "train_loss": 0.006725810002535582,
          "validation_loss": 0.004612378776073456,
          "train_metric": 0.0006773988716304302,
          "validation_metric": 0.00046454189578071237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.372057226000834
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0042376437049824744,
          "train_loss": 0.008611996658146381,
          "validation_loss": 0.007313916459679604,
          "train_metric": 0.0008673688862472773,
          "validation_metric": 0.0007366309291683137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.384529898999972
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.003718590465723537,
          "train_loss": 0.0046224151737987995,
          "validation_loss": 0.004048390779644251,
          "train_metric": 0.00046555273002013564,
          "validation_metric": 0.0004077390767633915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3968417050000426
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.004934152733767405,
          "train_loss": 0.003956385888159275,
          "validation_loss": 0.0029742075130343437,
          "train_metric": 0.0003984726790804416,
          "validation_metric": 0.00029955129139125347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4100827680013026
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005800496721349191,
          "train_loss": 0.0057286652736365795,
          "validation_loss": 0.004312220495194197,
          "train_metric": 0.0005769702838733792,
          "validation_metric": 0.00043431101948954165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.422845885001152
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.005798119978862815,
          "train_loss": 0.003874324494972825,
          "validation_loss": 0.003203080501407385,
          "train_metric": 0.0003902077442035079,
          "validation_metric": 0.0003226025146432221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4355387249997875
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.004693603797932155,
          "train_loss": 0.0036567864008247852,
          "validation_loss": 0.0030667304527014494,
          "train_metric": 0.00036829817690886557,
          "validation_metric": 0.0003088698722422123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4480018280009972
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.006078236154280603,
          "train_loss": 0.0039098323322832584,
          "validation_loss": 0.003250532317906618,
          "train_metric": 0.00039378402289003134,
          "validation_metric": 0.00032738171285018325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4606359400004294
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.006170019289129414,
          "train_loss": 0.004382819402962923,
          "validation_loss": 0.003927698824554682,
          "train_metric": 0.00044142152182757854,
          "validation_metric": 0.00039558339631184936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4735742310003843
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.006724856426444603,
          "train_loss": 0.006056437734514475,
          "validation_loss": 0.003893089946359396,
          "train_metric": 0.000609982234891504,
          "validation_metric": 0.00039209777605719864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4854510300010588
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.008048473362578079,
          "train_loss": 0.00635053776204586,
          "validation_loss": 0.005864571779966354,
          "train_metric": 0.0006396028911694884,
          "validation_metric": 0.0005906581645831466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4975926509996498
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004812011764443014,
          "train_loss": 0.004753896035254002,
          "validation_loss": 0.002990598324686289,
          "train_metric": 0.00047879506018944085,
          "validation_metric": 0.0003012021188624203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5097034160007752
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.006023943002219312,
          "train_loss": 0.004237508401274681,
          "validation_loss": 0.0034708583261817694,
          "train_metric": 0.00042678636964410543,
          "validation_metric": 0.00034957219031639397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5218301559998508
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004566476432955824,
          "train_loss": 0.004845383111387491,
          "validation_loss": 0.003289751475676894,
          "train_metric": 0.00048800918739289045,
          "validation_metric": 0.00033133174292743206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5338094770013413
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004862325105932541,
          "train_loss": 0.015290547162294388,
          "validation_loss": 0.012459409423172474,
          "train_metric": 0.0015400080010294914,
          "validation_metric": 0.0012548661325126886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5459733640000195
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.005843791546794819,
          "train_loss": 0.007712712977081537,
          "validation_loss": 0.005100273992866278,
          "train_metric": 0.0007767961942590773,
          "validation_metric": 0.0005136809777468443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.561225782999827
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.00594037242262857,
          "train_loss": 0.0035832738503813744,
          "validation_loss": 0.0032170782797038555,
          "train_metric": 0.0003608942497521639,
          "validation_metric": 0.0003240123623982072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5733672970000043
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.00484516433061799,
          "train_loss": 0.004274833481758833,
          "validation_loss": 0.0028759643901139498,
          "train_metric": 0.0004305455950088799,
          "validation_metric": 0.0002896566002164036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.58586111600016
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.005310069282131735,
          "train_loss": 0.005691259168088436,
          "validation_loss": 0.0057789417915046215,
          "train_metric": 0.000573202851228416,
          "validation_metric": 0.0005820338265039027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5991089939998346
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004563290200167103,
          "train_loss": 0.003815194359049201,
          "validation_loss": 0.0030684981029480696,
          "train_metric": 0.00038425237289629877,
          "validation_metric": 0.00030904790037311614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6109305980007775
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004012297711597057,
          "train_loss": 0.003504382912069559,
          "validation_loss": 0.00320256850682199,
          "train_metric": 0.00035294858389534056,
          "validation_metric": 0.00032255094265565276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6231692390010721
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0032777958404039964,
          "train_loss": 0.004561129026114941,
          "validation_loss": 0.003356299130246043,
          "train_metric": 0.0004593802150338888,
          "validation_metric": 0.0003380341804586351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6353310430004058
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004629755392670631,
          "train_loss": 0.005960050970315933,
          "validation_loss": 0.0048429411835968494,
          "train_metric": 0.0006002745358273387,
          "validation_metric": 0.0004877632309217006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6475224980003986
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0035409520496614277,
          "train_loss": 0.003726726397871971,
          "validation_loss": 0.00293556647375226,
          "train_metric": 0.0003753423225134611,
          "validation_metric": 0.0002956594980787486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.659865436000473
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0036656349839176983,
          "train_loss": 0.0038230500649660826,
          "validation_loss": 0.0024837576784193516,
          "train_metric": 0.000385043618734926,
          "validation_metric": 0.00025015498977154493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6728056330011896
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003467368755082134,
          "train_loss": 0.002849121345207095,
          "validation_loss": 0.0024419031105935574,
          "train_metric": 0.0002869530289899558,
          "validation_metric": 0.0002459395327605307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6856077880001976
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0032352217749576084,
          "train_loss": 0.0027196642477065325,
          "validation_loss": 0.002125456929206848,
          "train_metric": 0.00027391465846449137,
          "validation_metric": 0.00021406824816949666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6986629940001876
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.003116513697023038,
          "train_loss": 0.002760651521384716,
          "validation_loss": 0.0021605254150927067,
          "train_metric": 0.0002780427457764745,
          "validation_metric": 0.0002176002017222345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7103678450002917
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.005705245959688909,
          "train_loss": 0.004132214933633804,
          "validation_loss": 0.0030958782881498337,
          "train_metric": 0.0004161815741099417,
          "validation_metric": 0.00031180554651655257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7223655270008749
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.004206506768241525,
          "train_loss": 0.0035363517235964537,
          "validation_loss": 0.0032301112078130245,
          "train_metric": 0.00035616836976259947,
          "validation_metric": 0.00032532497425563633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7345906290011044
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.003316440124763176,
          "train_loss": 0.004232580773532391,
          "validation_loss": 0.003365066135302186,
          "train_metric": 0.0004262900329194963,
          "validation_metric": 0.00033891716157086194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7465695820010296
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0029487705978681333,
          "train_loss": 0.003671113168820739,
          "validation_loss": 0.00261243199929595,
          "train_metric": 0.00036974114482291043,
          "validation_metric": 0.00026311457622796297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.758919224001147
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.003354367312567774,
          "train_loss": 0.003242547158151865,
          "validation_loss": 0.002836269559338689,
          "train_metric": 0.00032657745759934187,
          "validation_metric": 0.0002856587234418839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.771212944000581
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.003354460666741943,
          "train_loss": 0.0037531645502895117,
          "validation_loss": 0.0027949183713644743,
          "train_metric": 0.00037800506106577814,
          "validation_metric": 0.0002814939944073558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7864306720002787
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0030763311733608134,
          "train_loss": 0.0025999019853770733,
          "validation_loss": 0.0020833457820117474,
          "train_metric": 0.00026185260503552854,
          "validation_metric": 0.00020982696150895208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7987050330011698
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.004421924091730034,
          "train_loss": 0.002980317221954465,
          "validation_loss": 0.0023505401331931353,
          "train_metric": 0.00030016666278243065,
          "validation_metric": 0.00023673780378885567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8105669640008273
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.008317957348481286,
          "train_loss": 0.0046837227419018745,
          "validation_loss": 0.003456052392721176,
          "train_metric": 0.0004717273695860058,
          "validation_metric": 0.0003480809391476214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8239784039997176
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.007357078400673345,
          "train_loss": 0.0038785915821790695,
          "validation_loss": 0.002715475857257843,
          "train_metric": 0.00039063754957169294,
          "validation_metric": 0.00027349276933819056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8356192970004486
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.004551877253106795,
          "train_loss": 0.0053229378536343575,
          "validation_loss": 0.004790982697159052,
          "train_metric": 0.0005361068178899586,
          "validation_metric": 0.0004825302166864276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8477903490002063
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.005986768133880105,
          "train_loss": 0.014548912644386292,
          "validation_loss": 0.0105509702116251,
          "train_metric": 0.0014653131365776062,
          "validation_metric": 0.001062655122950673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8595846230000461
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.007334941801673267,
          "train_loss": 0.004672658629715443,
          "validation_loss": 0.004329536110162735,
          "train_metric": 0.0004706130421254784,
          "validation_metric": 0.0004360550665296614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8716806570009794
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.00324465858284384,
          "train_loss": 0.002510635182261467,
          "validation_loss": 0.0019880342297255993,
          "train_metric": 0.00025286199524998665,
          "validation_metric": 0.00020022752869408578,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8843426010007533
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.003071971204917645,
          "train_loss": 0.0058914064429700375,
          "validation_loss": 0.0052557531744241714,
          "train_metric": 0.0005933609209023416,
          "validation_metric": 0.000529340235516429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8967164839996258
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0035451394396659452,
          "train_loss": 0.0030678356997668743,
          "validation_loss": 0.0021457774564623833,
          "train_metric": 0.0003089811943937093,
          "validation_metric": 0.00021611485863104463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.909006069001407
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0032065153227449628,
          "train_loss": 0.0029508129227906466,
          "validation_loss": 0.002153229434043169,
          "train_metric": 0.0002971951034851372,
          "validation_metric": 0.00021686538821086287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9209313749997818
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.00402248500904534,
          "train_loss": 0.0027539979200810194,
          "validation_loss": 0.0024086283519864082,
          "train_metric": 0.0002773726300802082,
          "validation_metric": 0.00024258822668343782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9330018250002468
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.003144582577078836,
          "train_loss": 0.005476237740367651,
          "validation_loss": 0.004519245121628046,
          "train_metric": 0.0005515466327778995,
          "validation_metric": 0.00045516181853599846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9449984330003645
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.005625335092190653,
          "train_loss": 0.00974277313798666,
          "validation_loss": 0.00818098708987236,
          "train_metric": 0.0009812564821913838,
          "validation_metric": 0.000823959126137197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9569893030002277
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.009860139994998462,
          "train_loss": 0.008043156936764717,
          "validation_loss": 0.005140132736414671,
          "train_metric": 0.0008100773557089269,
          "validation_metric": 0.0005176953272894025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.969103649000317
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.011956247151829302,
          "train_loss": 0.007333043497055769,
          "validation_loss": 0.006551253143697977,
          "train_metric": 0.0007385573117062449,
          "validation_metric": 0.0006598182953894138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.984327927000777
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.004961812453984749,
          "train_loss": 0.003456316888332367,
          "validation_loss": 0.002732792403548956,
          "train_metric": 0.0003481075691524893,
          "validation_metric": 0.00027523687458597124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9963703590001387
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0035460891231196,
          "train_loss": 0.0040171341970562935,
          "validation_loss": 0.003364820033311844,
          "train_metric": 0.0004045910609420389,
          "validation_metric": 0.0003388923068996519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.009791395999855
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.005057909067545552,
          "train_loss": 0.00878403801470995,
          "validation_loss": 0.006471472326666117,
          "train_metric": 0.0008846960845403373,
          "validation_metric": 0.0006517829606309533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0216875019996223
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.003875917360346648,
          "train_loss": 0.0023570379707962275,
          "validation_loss": 0.001854380825534463,
          "train_metric": 0.00023739223252050579,
          "validation_metric": 0.00018676645413506776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0343389500012563
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0033368885378877167,
          "train_loss": 0.002295364160090685,
          "validation_loss": 0.0017325261142104864,
          "train_metric": 0.00023118067474570125,
          "validation_metric": 0.00017449371807742864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.047544826000376
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0029448385757859796,
          "train_loss": 0.0029684961773455143,
          "validation_loss": 0.002668909030035138,
          "train_metric": 0.0002989760832861066,
          "validation_metric": 0.00026880274526774883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.059937097999864
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.002738524752203375,
          "train_loss": 0.0023319413885474205,
          "validation_loss": 0.001844608224928379,
          "train_metric": 0.00023486460850108415,
          "validation_metric": 0.0001857821480371058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.071443894001277
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0025089344853768125,
          "train_loss": 0.002800591057166457,
          "validation_loss": 0.001950335456058383,
          "train_metric": 0.000282065273495391,
          "validation_metric": 0.00019643065752461553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.08374171000105
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0026724098170234356,
          "train_loss": 0.0027136437129229307,
          "validation_loss": 0.0020270904060453176,
          "train_metric": 0.0002733082219492644,
          "validation_metric": 0.00020416115876287222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.095823994000966
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.002482489269823418,
          "train_loss": 0.002450625877827406,
          "validation_loss": 0.0016597742214798927,
          "train_metric": 0.0002468180609866977,
          "validation_metric": 0.00016716639220248908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1083193690010376
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.004571318393573165,
          "train_loss": 0.0025558758061379194,
          "validation_loss": 0.0018675525207072496,
          "train_metric": 0.0002574184618424624,
          "validation_metric": 0.00018809305038303137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.119989737000651
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0026852921837416943,
          "train_loss": 0.0027902633883059025,
          "validation_loss": 0.0020299365278333426,
          "train_metric": 0.00028102516080252826,
          "validation_metric": 0.0002044477587332949,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1319161760002316
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0024759148072917014,
          "train_loss": 0.002745586447417736,
          "validation_loss": 0.0023442141246050596,
          "train_metric": 0.00027652541757561266,
          "validation_metric": 0.00023610067728441209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.143345730000874
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0033342227070534136,
          "train_loss": 0.0021538955625146627,
          "validation_loss": 0.0016573999309912324,
          "train_metric": 0.0002169324434362352,
          "validation_metric": 0.00016692727513145655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1556550820005214
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0031309123114624526,
          "train_loss": 0.002333182143047452,
          "validation_loss": 0.0018330601742491126,
          "train_metric": 0.00023498958034906536,
          "validation_metric": 0.0001846191007643938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.16747820800083
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.003075896733207628,
          "train_loss": 0.0022919191978871822,
          "validation_loss": 0.0018373990897089243,
          "train_metric": 0.00023083369887899607,
          "validation_metric": 0.0001850561093306169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.182202669000617
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0026255574412061833,
          "train_loss": 0.0023148709442466497,
          "validation_loss": 0.0018701546359807253,
          "train_metric": 0.00023314535792451352,
          "validation_metric": 0.00018835511582437903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1955996100004995
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.002453541790600866,
          "train_loss": 0.0030195657163858414,
          "validation_loss": 0.0027345698326826096,
          "train_metric": 0.0003041195741388947,
          "validation_metric": 0.0002754158340394497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2076694290008163
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0027930135984206572,
          "train_loss": 0.0030039579141885042,
          "validation_loss": 0.0017668992513790727,
          "train_metric": 0.0003025476180482656,
          "validation_metric": 0.0001779556187102571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.219492299000194
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.003609309824241791,
          "train_loss": 0.002380850026383996,
          "validation_loss": 0.0020816437900066376,
          "train_metric": 0.00023979049001354724,
          "validation_metric": 0.0002096555253956467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.231348806000824
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.003166824739309959,
          "train_loss": 0.0022575075272470713,
          "validation_loss": 0.001637637265957892,
          "train_metric": 0.0002273678983328864,
          "validation_metric": 0.00016493683506269008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2441946920007467
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0026641240474418737,
          "train_loss": 0.002463191282004118,
          "validation_loss": 0.0018227638211101294,
          "train_metric": 0.0002480836119502783,
          "validation_metric": 0.00018358211673330516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.256067198000892
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0025182617973769084,
          "train_loss": 0.006786719895899296,
          "validation_loss": 0.004171076230704784,
          "train_metric": 0.0006835336680524051,
          "validation_metric": 0.00042009554454125464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2680343049996736
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0036069715206394903,
          "train_loss": 0.0045612347312271595,
          "validation_loss": 0.0035666627809405327,
          "train_metric": 0.00045939089613966644,
          "validation_metric": 0.000359221245162189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2801428830007353
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0038533457700395957,
          "train_loss": 0.00272851693443954,
          "validation_loss": 0.0023455594200640917,
          "train_metric": 0.00027480622520670295,
          "validation_metric": 0.00023623615561518818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.291974964000474
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0036555626029439736,
          "train_loss": 0.005481651518493891,
          "validation_loss": 0.004034225828945637,
          "train_metric": 0.0005520919221453369,
          "validation_metric": 0.0004063124652020633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.303527493000729
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.00420777962426655,
          "train_loss": 0.00536805996671319,
          "validation_loss": 0.005281521938741207,
          "train_metric": 0.0005406513228081167,
          "validation_metric": 0.0005319355987012386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.315601577000052
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.003925576325855218,
          "train_loss": 0.0022218034137040377,
          "validation_loss": 0.001767649780958891,
          "train_metric": 0.0002237718872493133,
          "validation_metric": 0.00017803121590986848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.327744950000124
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.002755987297859974,
          "train_loss": 0.002186728175729513,
          "validation_loss": 0.001509235822595656,
          "train_metric": 0.0002202392352046445,
          "validation_metric": 0.00015200470807030797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.340721456999745
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0030173054183251224,
          "train_loss": 0.004009597469121218,
          "validation_loss": 0.003095917170867324,
          "train_metric": 0.00040383197483606637,
          "validation_metric": 0.00031180944642983377,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3525605329996324
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0024480073916492984,
          "train_loss": 0.0021202382631599903,
          "validation_loss": 0.0014480436220765114,
          "train_metric": 0.0002135427057510242,
          "validation_metric": 0.00014584165182895958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.364702164000846
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0023600348285981454,
          "train_loss": 0.003755304729565978,
          "validation_loss": 0.00252273166552186,
          "train_metric": 0.00037822051672264934,
          "validation_metric": 0.0002540802815929055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3769762390002143
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0031146819455898367,
          "train_loss": 0.0019089601701125503,
          "validation_loss": 0.0016199933597818017,
          "train_metric": 0.00019226349832024425,
          "validation_metric": 0.0001631597988307476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.388837913000316
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.001952450144017348,
          "train_loss": 0.0023568507749587297,
          "validation_loss": 0.001724149682559073,
          "train_metric": 0.00023737340234220028,
          "validation_metric": 0.00017365004168823361,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4042801880004845
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.003045435772946803,
          "train_loss": 0.002871612785384059,
          "validation_loss": 0.002158284652978182,
          "train_metric": 0.0002892183547373861,
          "validation_metric": 0.00021737453062087297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.417815974000405
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0033883517025969923,
          "train_loss": 0.003249666653573513,
          "validation_loss": 0.002186691155657172,
          "train_metric": 0.0003272945177741349,
          "validation_metric": 0.00022023550991434604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.429709226000341
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.004607505201420281,
          "train_loss": 0.0026782762724906206,
          "validation_loss": 0.0024905595928430557,
          "train_metric": 0.00026974614593200386,
          "validation_metric": 0.00025084003573283553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4415002140012803
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.002356762342969887,
          "train_loss": 0.0021130628883838654,
          "validation_loss": 0.0016810566885396838,
          "train_metric": 0.00021281997032929212,
          "validation_metric": 0.0001693098747637123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4534227599997394
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0020863676982116885,
          "train_loss": 0.0028579707723110914,
          "validation_loss": 0.0017721292097121477,
          "train_metric": 0.00028784433379769325,
          "validation_metric": 0.0001784823543857783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4654046160012513
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.003414187773159938,
          "train_loss": 0.0031427238136529922,
          "validation_loss": 0.00238797371275723,
          "train_metric": 0.0003165236848872155,
          "validation_metric": 0.00024050797219388187,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4773971490012627
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0031459692545467988,
          "train_loss": 0.0029351734556257725,
          "validation_loss": 0.0026866858825087547,
          "train_metric": 0.0002956199168693274,
          "validation_metric": 0.00027059318381361663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.489028419000533
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.003104230265307706,
          "train_loss": 0.0019396462012082338,
          "validation_loss": 0.0014020093949511647,
          "train_metric": 0.00019535406318027526,
          "validation_metric": 0.00014120523701421916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5012808480005333
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0029773654587188503,
          "train_loss": 0.002212008461356163,
          "validation_loss": 0.0014210757799446583,
          "train_metric": 0.00022278539836406708,
          "validation_metric": 0.00014312555140350014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5135454880000907
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.00253698150117998,
          "train_loss": 0.010097015649080276,
          "validation_loss": 0.007581114768981934,
          "train_metric": 0.0010169345187023282,
          "validation_metric": 0.000763542135246098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.525459069000135
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0034752343490254134,
          "train_loss": 0.005185493268072605,
          "validation_loss": 0.003139883978292346,
          "train_metric": 0.0005222639301791787,
          "validation_metric": 0.0003162375942338258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5373976770006266
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0031758504919707775,
          "train_loss": 0.002167424187064171,
          "validation_loss": 0.0019152493914589286,
          "train_metric": 0.00021829502657055855,
          "validation_metric": 0.0001928969140863046,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.549673053001243
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0023626493857591413,
          "train_loss": 0.005400130525231361,
          "validation_loss": 0.003406630363315344,
          "train_metric": 0.0005438813823275268,
          "validation_metric": 0.0003431033401284367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.561921615000756
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0068640263361885445,
          "train_loss": 0.011208554729819298,
          "validation_loss": 0.01059107668697834,
          "train_metric": 0.001128884614445269,
          "validation_metric": 0.001066694618202746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5737544410003466
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.009373934284667484,
          "train_loss": 0.008373755030333996,
          "validation_loss": 0.005756470840424299,
          "train_metric": 0.0008433739421889186,
          "validation_metric": 0.0005797705962322652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.585845079000137
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.009511430122074671,
          "train_loss": 0.007316497154533863,
          "validation_loss": 0.005829921457916498,
          "train_metric": 0.0007368907681666315,
          "validation_metric": 0.0005871683824807405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6007533240008343
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.004448422558198217,
          "train_loss": 0.0028821639716625214,
          "validation_loss": 0.0027408364694565535,
          "train_metric": 0.00029028099379502237,
          "validation_metric": 0.0002760470088105649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.614271534001091
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.003829468183539575,
          "train_loss": 0.0024534312542527914,
          "validation_loss": 0.0015636099269613624,
          "train_metric": 0.00024710060097277164,
          "validation_metric": 0.00015748107398394495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.626229425000929
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0035713333236344624,
          "train_loss": 0.00317514524795115,
          "validation_loss": 0.0021918495185673237,
          "train_metric": 0.0003197889309376478,
          "validation_metric": 0.00022075507149565965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6384014960003697
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.0031809776310183224,
          "train_loss": 0.0020567988976836205,
          "validation_loss": 0.0018109752563759685,
          "train_metric": 0.00020715325081255287,
          "validation_metric": 0.00018239478231407702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.650673715001176
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0031279829927370884,
          "train_loss": 0.002347909612581134,
          "validation_loss": 0.0021234452724456787,
          "train_metric": 0.00023647288617212325,
          "validation_metric": 0.00021386565640568733,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.662707216000854
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.004342647447629133,
          "train_loss": 0.002785523422062397,
          "validation_loss": 0.002424859441816807,
          "train_metric": 0.000280547741567716,
          "validation_metric": 0.00024422293063253164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.674743245001082
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0054159601422725245,
          "train_loss": 0.005487749353051186,
          "validation_loss": 0.003518091980367899,
          "train_metric": 0.000552706012967974,
          "validation_metric": 0.0003543293278198689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.686505499001214
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.002893480792408809,
          "train_loss": 0.0019354225369170308,
          "validation_loss": 0.0017573285149410367,
          "train_metric": 0.00019492865249048918,
          "validation_metric": 0.00017699171439744532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6984536979998666
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0022311666725727264,
          "train_loss": 0.00242044311016798,
          "validation_loss": 0.002210731850937009,
          "train_metric": 0.00024377816589549184,
          "validation_metric": 0.00022265681764110923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.710310919001131
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.002259138029330643,
          "train_loss": 0.003122135763987899,
          "validation_loss": 0.002119766315445304,
          "train_metric": 0.00031445009517483413,
          "validation_metric": 0.00021349509188439697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7221989200006647
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.002871214055630844,
          "train_loss": 0.0017705890350043774,
          "validation_loss": 0.0014744269428774714,
          "train_metric": 0.00017832726007327437,
          "validation_metric": 0.00014849888975732028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.734209988000657
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.002729321662627626,
          "train_loss": 0.0016614559572190046,
          "validation_loss": 0.0013091258006170392,
          "train_metric": 0.0001673357473919168,
          "validation_metric": 0.0001318503636866808,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7489935780013184
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.002084121973894071,
          "train_loss": 0.0017543258145451546,
          "validation_loss": 0.0012665134854614735,
          "train_metric": 0.00017668928194325417,
          "validation_metric": 0.00012755859643220901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.761998568001218
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.002011180495173903,
          "train_loss": 0.002578475745394826,
          "validation_loss": 0.0017211195081472397,
          "train_metric": 0.0002596946433186531,
          "validation_metric": 0.00017334484437014908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7747594450011093
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0026425494615978096,
          "train_loss": 0.0026833037845790386,
          "validation_loss": 0.002517719753086567,
          "train_metric": 0.00027025252347812057,
          "validation_metric": 0.00025357550475746393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.786987069999668
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0020294832920626504,
          "train_loss": 0.001590525615029037,
          "validation_loss": 0.0011259678285568953,
          "train_metric": 0.00016019190661609173,
          "validation_metric": 0.00011340333730913699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7995523080007843
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.001684982204096741,
          "train_loss": 0.0016525359824299812,
          "validation_loss": 0.0013264664448797703,
          "train_metric": 0.00016643738490529358,
          "validation_metric": 0.00013359685544855893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8210414870009117
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0019567185372579843,
          "train_loss": 0.0017174723325297236,
          "validation_loss": 0.0011486882576718926,
          "train_metric": 0.0001729775103740394,
          "validation_metric": 0.00011569166963454336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.835826728000029
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0023870476652518846,
          "train_loss": 0.0029884218238294125,
          "validation_loss": 0.0018859049305319786,
          "train_metric": 0.0003009829088114202,
          "validation_metric": 0.00018994143465533853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8685603020003327
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0020386705073178746,
          "train_loss": 0.002154183341190219,
          "validation_loss": 0.0018627791432663798,
          "train_metric": 0.00021696145995520055,
          "validation_metric": 0.00018761228420771658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.891672137000569
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.002037708294665208,
          "train_loss": 0.0014645286137238145,
          "validation_loss": 0.0011382753727957606,
          "train_metric": 0.00014750193804502487,
          "validation_metric": 0.00011464292037999257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9038278220014035
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0018139660060114693,
          "train_loss": 0.00162126321811229,
          "validation_loss": 0.0013307060580700636,
          "train_metric": 0.00016328769561368972,
          "validation_metric": 0.00013402382319327444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9196566179998626
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.002057664027233841,
          "train_loss": 0.0014535424998030066,
          "validation_loss": 0.0011239721206948161,
          "train_metric": 0.000146395483170636,
          "validation_metric": 0.0001132023535319604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9365362570006255
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.002267247728013899,
          "train_loss": 0.0040548574179410934,
          "validation_loss": 0.0037157898768782616,
          "train_metric": 0.00040839039138518274,
          "validation_metric": 0.00037424074253067374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.954324059999635
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0027996203971270006,
          "train_loss": 0.0020392388105392456,
          "validation_loss": 0.0014943601563572884,
          "train_metric": 0.00020538465469144285,
          "validation_metric": 0.000150506486534141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9737043130007805
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.003379970634341589,
          "train_loss": 0.0019691395573318005,
          "validation_loss": 0.001409543096087873,
          "train_metric": 0.00019832451653201133,
          "validation_metric": 0.00014196400297805667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9924445369997557
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0028826999114244245,
          "train_loss": 0.0023024408146739006,
          "validation_loss": 0.0014231392415240407,
          "train_metric": 0.0002318933984497562,
          "validation_metric": 0.0001433333964087069,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0141134269997565
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.005877944509848021,
          "train_loss": 0.003120494307950139,
          "validation_loss": 0.002485416829586029,
          "train_metric": 0.00031428472721017897,
          "validation_metric": 0.00025032207486219704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0330582540009345
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0037769229311379604,
          "train_loss": 0.0030922023579478264,
          "validation_loss": 0.002755495021119714,
          "train_metric": 0.000311435287585482,
          "validation_metric": 0.00027752338792197406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.048180461000811
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0024311173110618256,
          "train_loss": 0.0014997466932982206,
          "validation_loss": 0.0010408074595034122,
          "train_metric": 0.00015104901103768498,
          "validation_metric": 0.00010482633661013097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0637511590011854
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.001683561466052197,
          "train_loss": 0.003287322586402297,
          "validation_loss": 0.0026625636965036392,
          "train_metric": 0.0003310871252324432,
          "validation_metric": 0.00026816362515091896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0775452660000155
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.003940716775105102,
          "train_loss": 0.00550519023090601,
          "validation_loss": 0.003377291141077876,
          "train_metric": 0.0005544626037590206,
          "validation_metric": 0.0003401484282221645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.096121339000092
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0023957669545779936,
          "train_loss": 0.0018594050779938698,
          "validation_loss": 0.0014552470529451966,
          "train_metric": 0.00018727245333138853,
          "validation_metric": 0.00014656713756266981,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.10971860599966
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0027011621423298493,
          "train_loss": 0.003796285018324852,
          "validation_loss": 0.003278596792370081,
          "train_metric": 0.0003823479055427015,
          "validation_metric": 0.00033020824776031077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.125802364000265
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.004363019528682344,
          "train_loss": 0.004163819365203381,
          "validation_loss": 0.0032841493375599384,
          "train_metric": 0.0004193646600469947,
          "validation_metric": 0.00033076750696636736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.139708385000631
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.004638688153136172,
          "train_loss": 0.0026194469537585974,
          "validation_loss": 0.001589111634530127,
          "train_metric": 0.00026382110081613064,
          "validation_metric": 0.00016004954522941262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.154435790000207
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.002272043309858418,
          "train_loss": 0.0017885668203234673,
          "validation_loss": 0.001483784057199955,
          "train_metric": 0.00018013788212556392,
          "validation_metric": 0.00014944130089133978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1688458820008236
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0020316103436925914,
          "train_loss": 0.004961203783750534,
          "validation_loss": 0.0038918808568269014,
          "train_metric": 0.0004996742936782539,
          "validation_metric": 0.0003919759765267372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.18284340900027
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.0036232361308066174,
          "train_loss": 0.0020653612446039915,
          "validation_loss": 0.0015738274669274688,
          "train_metric": 0.00020801558275707066,
          "validation_metric": 0.0001585101563250646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1966902149997622
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.00266950151490164,
          "train_loss": 0.0017800843343138695,
          "validation_loss": 0.0012196549214422703,
          "train_metric": 0.00017928356828633696,
          "validation_metric": 0.000122839177493006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2105308500013052
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.0038317037106025964,
          "train_loss": 0.002770568011328578,
          "validation_loss": 0.0028661799151450396,
          "train_metric": 0.00027904147282242775,
          "validation_metric": 0.00028867117362096906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.223654143999738
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.001918889829539694,
          "train_loss": 0.0018469180213287473,
          "validation_loss": 0.0013900757767260075,
          "train_metric": 0.00018601481860969216,
          "validation_metric": 0.00014000333612784743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.239525812001375
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.001604038958248566,
          "train_loss": 0.0014191364170983434,
          "validation_loss": 0.001121375011280179,
          "train_metric": 0.00014293025014922023,
          "validation_metric": 0.00011294077557977289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2527783890000137
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.005004211423511151,
          "train_loss": 0.010321778245270252,
          "validation_loss": 0.006524376571178436,
          "train_metric": 0.0010395717108622193,
          "validation_metric": 0.0006571112317033112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.270658297000409
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.012533734276075847,
          "train_loss": 0.00258064572699368,
          "validation_loss": 0.0021549744997173548,
          "train_metric": 0.0002599131839815527,
          "validation_metric": 0.0002170411462429911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.288754291999794
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.0032322926999768242,
          "train_loss": 0.0024953342508524656,
          "validation_loss": 0.0017288628732785583,
          "train_metric": 0.00025132091832347214,
          "validation_metric": 0.00017412472516298294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3136001900002157
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.002008601564739365,
          "train_loss": 0.0015948291402310133,
          "validation_loss": 0.0011312742717564106,
          "train_metric": 0.00016062533541116863,
          "validation_metric": 0.00011393780005164444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3314976439996826
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.003208886128049926,
          "train_loss": 0.004896655213087797,
          "validation_loss": 0.0037379751447588205,
          "train_metric": 0.0004931731382384896,
          "validation_metric": 0.00037647518911398947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3515498510005273
        }
      ],
      "best": {
        "epoch": 222,
        "updates": 3552
      },
      "best_validation_metric": 0.00010482633661013097,
      "config": {
        "label": "Residual",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": true,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010238473353490303
    },
    "a7bb18d2c173_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7814145870506763,
          "train_loss": 0.3954164981842041,
          "validation_loss": 0.30804216861724854,
          "train_metric": 0.03982490301132202,
          "validation_metric": 0.031024876981973648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.032834257999638794
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.22260148031637073,
          "train_loss": 0.12675222754478455,
          "validation_loss": 0.10941827297210693,
          "train_metric": 0.012766020372509956,
          "validation_metric": 0.011020207777619362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05189606299973093
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13687859661877155,
          "train_loss": 0.0998186320066452,
          "validation_loss": 0.08325222134590149,
          "train_metric": 0.010053366422653198,
          "validation_metric": 0.008384859189391136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09023794399945473
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10449196049012244,
          "train_loss": 0.08566039800643921,
          "validation_loss": 0.06854356080293655,
          "train_metric": 0.008627401664853096,
          "validation_metric": 0.006903456524014473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10868286699951568
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08928567566908896,
          "train_loss": 0.06895682960748672,
          "validation_loss": 0.05694973096251488,
          "train_metric": 0.006945078261196613,
          "validation_metric": 0.0057357680052518845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12738890999935393
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.07694822817575186,
          "train_loss": 0.06550305336713791,
          "validation_loss": 0.05250849947333336,
          "train_metric": 0.006597227416932583,
          "validation_metric": 0.005288464017212391,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14632918399911432
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.07500088540837169,
          "train_loss": 0.06622104346752167,
          "validation_loss": 0.0527079738676548,
          "train_metric": 0.006669540423899889,
          "validation_metric": 0.005308554507791996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16402488099993207
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.07500583142973483,
          "train_loss": 0.0584404319524765,
          "validation_loss": 0.04720449447631836,
          "train_metric": 0.0058859060518443584,
          "validation_metric": 0.004754263907670975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18215814299946942
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.06561703816987574,
          "train_loss": 0.05409875139594078,
          "validation_loss": 0.04502923786640167,
          "train_metric": 0.005448628216981888,
          "validation_metric": 0.0045351795852184296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20101351599987538
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.062038135481998324,
          "train_loss": 0.053244877606630325,
          "validation_loss": 0.04421956464648247,
          "train_metric": 0.0053626298904418945,
          "validation_metric": 0.004453632980585098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21971928600032697
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.06293298676609993,
          "train_loss": 0.05307386815547943,
          "validation_loss": 0.041624516248703,
          "train_metric": 0.005345406010746956,
          "validation_metric": 0.004192268941551447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2389495249990432
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.05762085015885532,
          "train_loss": 0.0455588698387146,
          "validation_loss": 0.03780829533934593,
          "train_metric": 0.004588522482663393,
          "validation_metric": 0.0038079130463302135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25745900199945027
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.05690541840158403,
          "train_loss": 0.042855244129896164,
          "validation_loss": 0.035773131996393204,
          "train_metric": 0.004316223785281181,
          "validation_metric": 0.003602938959375024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2763016680000874
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.05131055030506104,
          "train_loss": 0.042439259588718414,
          "validation_loss": 0.03465304151177406,
          "train_metric": 0.004274326842278242,
          "validation_metric": 0.0034901273902505636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29486346100020455
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.054770662216469646,
          "train_loss": 0.041851624846458435,
          "validation_loss": 0.03368465229868889,
          "train_metric": 0.00421514268964529,
          "validation_metric": 0.0033925948664546013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3145346069995867
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.051612536306492984,
          "train_loss": 0.03780415281653404,
          "validation_loss": 0.03091178461909294,
          "train_metric": 0.003807495580986142,
          "validation_metric": 0.003113321727141738,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33339051099937933
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.04799352714326233,
          "train_loss": 0.03501547873020172,
          "validation_loss": 0.028802592307329178,
          "train_metric": 0.0035266310442239046,
          "validation_metric": 0.0029008914716541767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3517056069995306
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.05227307602763176,
          "train_loss": 0.03318672627210617,
          "validation_loss": 0.025992732495069504,
          "train_metric": 0.003342445706948638,
          "validation_metric": 0.0026178930420428514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.370558730999619
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.04297623340971768,
          "train_loss": 0.029593322426080704,
          "validation_loss": 0.023705298081040382,
          "train_metric": 0.002980530960485339,
          "validation_metric": 0.0023875113110989332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38951528899997356
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.04155672492925078,
          "train_loss": 0.028391817584633827,
          "validation_loss": 0.022333141416311264,
          "train_metric": 0.0028595193289220333,
          "validation_metric": 0.0022493123542517424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4082111569987319
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.03856165299657732,
          "train_loss": 0.025596683844923973,
          "validation_loss": 0.02065998502075672,
          "train_metric": 0.002578004263341427,
          "validation_metric": 0.0020807981491088867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4272947829995246
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.038278600201010704,
          "train_loss": 0.02302641049027443,
          "validation_loss": 0.019143711775541306,
          "train_metric": 0.0023191357031464577,
          "validation_metric": 0.0019280845299363136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4461683890003769
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.03669173165690154,
          "train_loss": 0.0214802548289299,
          "validation_loss": 0.01708994247019291,
          "train_metric": 0.0021634125150740147,
          "validation_metric": 0.0017212366219609976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4653606369993213
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.03640718385577202,
          "train_loss": 0.02107406221330166,
          "validation_loss": 0.017444832250475883,
          "train_metric": 0.0021225023083388805,
          "validation_metric": 0.0017569795018061996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4843241449998459
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.03218612493947148,
          "train_loss": 0.021091826260089874,
          "validation_loss": 0.01672000251710415,
          "train_metric": 0.002124291379004717,
          "validation_metric": 0.0016839775489643216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5029456969987223
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.03800134756602347,
          "train_loss": 0.01980690285563469,
          "validation_loss": 0.015556933358311653,
          "train_metric": 0.001994878752157092,
          "validation_metric": 0.0015668373089283705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5214693460002309
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.03123316739220172,
          "train_loss": 0.01942870207130909,
          "validation_loss": 0.015438446775078773,
          "train_metric": 0.001956787658855319,
          "validation_metric": 0.0015549039235338569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5402545819997613
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.032131158397533,
          "train_loss": 0.021418439224362373,
          "validation_loss": 0.018161386251449585,
          "train_metric": 0.00215718662366271,
          "validation_metric": 0.0018291482701897621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5583147819997976
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.03933149657677859,
          "train_loss": 0.018975161015987396,
          "validation_loss": 0.015143267810344696,
          "train_metric": 0.00191110884770751,
          "validation_metric": 0.001525174593552947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5770204370001011
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.03526399866677821,
          "train_loss": 0.019753102213144302,
          "validation_loss": 0.016103994101285934,
          "train_metric": 0.001989459851756692,
          "validation_metric": 0.0016219354001805186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.594974440000442
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.02951533324085176,
          "train_loss": 0.016126777976751328,
          "validation_loss": 0.012799660675227642,
          "train_metric": 0.0016242299461737275,
          "validation_metric": 0.0012891348451375961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6135396249992482
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.02734902547672391,
          "train_loss": 0.016341358423233032,
          "validation_loss": 0.013206817209720612,
          "train_metric": 0.0016458416357636452,
          "validation_metric": 0.0013301422586664557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6317507599997043
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.027824038290418684,
          "train_loss": 0.014144368469715118,
          "validation_loss": 0.01162651926279068,
          "train_metric": 0.0014245690545067191,
          "validation_metric": 0.0011709805112332106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6500631589988188
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.028974946937523782,
          "train_loss": 0.017411917448043823,
          "validation_loss": 0.013200105167925358,
          "train_metric": 0.0017536646919324994,
          "validation_metric": 0.0013294662348926067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6676034329993854
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.02724970376584679,
          "train_loss": 0.015514892525970936,
          "validation_loss": 0.011972897686064243,
          "train_metric": 0.001562603167258203,
          "validation_metric": 0.0012058663414791226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.685467344999779
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.028658742317929864,
          "train_loss": 0.019842924550175667,
          "validation_loss": 0.014598979614675045,
          "train_metric": 0.0019985067192465067,
          "validation_metric": 0.0014703557826578617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7033509319990117
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.029375709709711373,
          "train_loss": 0.013993279077112675,
          "validation_loss": 0.011487230658531189,
          "train_metric": 0.0014093518257141113,
          "validation_metric": 0.001156951766461134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.722512360998735
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.02810075442539528,
          "train_loss": 0.014933343976736069,
          "validation_loss": 0.012235065922141075,
          "train_metric": 0.0015040317084640265,
          "validation_metric": 0.0012322711991146207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7409163339998486
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.0234252234804444,
          "train_loss": 0.012983136810362339,
          "validation_loss": 0.010560228489339352,
          "train_metric": 0.0013076140312477946,
          "validation_metric": 0.0010635876096785069,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7601303510000434
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.02682081627426669,
          "train_loss": 0.021471649408340454,
          "validation_loss": 0.016231216490268707,
          "train_metric": 0.0021625456865876913,
          "validation_metric": 0.0016347490018233657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7779085019992635
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.027600820758379996,
          "train_loss": 0.015250857919454575,
          "validation_loss": 0.011041841469705105,
          "train_metric": 0.0015360104152932763,
          "validation_metric": 0.0011120937997475266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7961136140002054
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.029368348128627986,
          "train_loss": 0.015445204451680183,
          "validation_loss": 0.011064745485782623,
          "train_metric": 0.0015555842546746135,
          "validation_metric": 0.0011144005693495274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8136965539997618
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.028424076735973358,
          "train_loss": 0.01464200671762228,
          "validation_loss": 0.010070476680994034,
          "train_metric": 0.0014746893430128694,
          "validation_metric": 0.00101426150649786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8328763019999315
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.024283423786982894,
          "train_loss": 0.012977142818272114,
          "validation_loss": 0.009241647087037563,
          "train_metric": 0.0013070105342194438,
          "validation_metric": 0.0009307849104516208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8513818119990901
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.022799318481702358,
          "train_loss": 0.0124423298984766,
          "validation_loss": 0.009252254851162434,
          "train_metric": 0.0012531457468867302,
          "validation_metric": 0.0009318532538600266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8697773679996317
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.024682429968379438,
          "train_loss": 0.01464198250323534,
          "validation_loss": 0.01079506240785122,
          "train_metric": 0.001474686898291111,
          "validation_metric": 0.0010872392449527979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8876800899997761
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.022979407047387213,
          "train_loss": 0.010441064834594727,
          "validation_loss": 0.008079484105110168,
          "train_metric": 0.0010515856556594372,
          "validation_metric": 0.000813735940027982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9064557419987977
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.02270357945235446,
          "train_loss": 0.018030976876616478,
          "validation_loss": 0.017610006034374237,
          "train_metric": 0.00181601382791996,
          "validation_metric": 0.0017736153677105904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9243533019998722
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.027099511469714344,
          "train_loss": 0.012335589155554771,
          "validation_loss": 0.011126410216093063,
          "train_metric": 0.001242395373992622,
          "validation_metric": 0.0011206114431843162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0419974149990594
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.024867851519957185,
          "train_loss": 0.010457474738359451,
          "validation_loss": 0.00876133143901825,
          "train_metric": 0.001053238520398736,
          "validation_metric": 0.0008824092801660299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0606704999991052
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.021779756410978734,
          "train_loss": 0.010378357023000717,
          "validation_loss": 0.008733291178941727,
          "train_metric": 0.0010452700080350041,
          "validation_metric": 0.0008795849862508476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0790855399991415
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.020304681733250618,
          "train_loss": 0.010018500499427319,
          "validation_loss": 0.00760723976418376,
          "train_metric": 0.0010090266587212682,
          "validation_metric": 0.0007661732379347086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0980898429988883
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.025931533717084676,
          "train_loss": 0.01247833389788866,
          "validation_loss": 0.009837978519499302,
          "train_metric": 0.001256772200576961,
          "validation_metric": 0.000990845263004303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1157738080000854
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.025020118278916925,
          "train_loss": 0.010233359411358833,
          "validation_loss": 0.008693724870681763,
          "train_metric": 0.001030666520819068,
          "validation_metric": 0.0008756001479923725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1333294979995117
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.023948683869093657,
          "train_loss": 0.0123439384624362,
          "validation_loss": 0.009686515666544437,
          "train_metric": 0.0012432362418621778,
          "validation_metric": 0.0009755903738550842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1515408969989949
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.022297492600046098,
          "train_loss": 0.013308493420481682,
          "validation_loss": 0.009280153550207615,
          "train_metric": 0.001340382732450962,
          "validation_metric": 0.0009346629958599806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.169990100999712
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.028274539043195546,
          "train_loss": 0.011884147301316261,
          "validation_loss": 0.008829011581838131,
          "train_metric": 0.001196927740238607,
          "validation_metric": 0.0008892258629202843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1878686000000016
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.021310870884917676,
          "train_loss": 0.010663533583283424,
          "validation_loss": 0.00692822365090251,
          "train_metric": 0.0010739919962361455,
          "validation_metric": 0.0006977852899581194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2061344260000624
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.01985327957663685,
          "train_loss": 0.010526522994041443,
          "validation_loss": 0.00786462053656578,
          "train_metric": 0.0010601928224787116,
          "validation_metric": 0.0007920957868918777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2237985379997554
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.01974426582455635,
          "train_loss": 0.008947912603616714,
          "validation_loss": 0.006870916113257408,
          "train_metric": 0.0009012009250000119,
          "validation_metric": 0.0006920135929249227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2419890780001879
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.02425143786240369,
          "train_loss": 0.008935658261179924,
          "validation_loss": 0.007910225540399551,
          "train_metric": 0.0008999668643809855,
          "validation_metric": 0.0007966889534145594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2596076949994313
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.019647888373583555,
          "train_loss": 0.009295937605202198,
          "validation_loss": 0.007510178256779909,
          "train_metric": 0.000936252879910171,
          "validation_metric": 0.000756397726945579,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2773507649999374
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.02019442879827693,
          "train_loss": 0.008071908727288246,
          "validation_loss": 0.006768073886632919,
          "train_metric": 0.000812973128631711,
          "validation_metric": 0.0006816555978730321,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.296132105999277
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.02233931695809588,
          "train_loss": 0.009246367029845715,
          "validation_loss": 0.0072311596013605595,
          "train_metric": 0.0009312601760029793,
          "validation_metric": 0.0007282959413714707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3147874320002302
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.02016249345615506,
          "train_loss": 0.010913573205471039,
          "validation_loss": 0.009060950018465519,
          "train_metric": 0.0010991750750690699,
          "validation_metric": 0.0009125856449827552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.332338971000354
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.022800956037826836,
          "train_loss": 0.01144286710768938,
          "validation_loss": 0.009895021095871925,
          "train_metric": 0.0011524837464094162,
          "validation_metric": 0.0009965903591364622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3498707139988255
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.020995680009946227,
          "train_loss": 0.013012913055717945,
          "validation_loss": 0.008059314452111721,
          "train_metric": 0.0013106130063533783,
          "validation_metric": 0.0008117046672850847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3672271569994336
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.02162119557033293,
          "train_loss": 0.010237767361104488,
          "validation_loss": 0.007141855079680681,
          "train_metric": 0.0010311102960258722,
          "validation_metric": 0.000719301460776478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.385072871000375
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.019293166347779334,
          "train_loss": 0.008072350174188614,
          "validation_loss": 0.006442527286708355,
          "train_metric": 0.0008130174828693271,
          "validation_metric": 0.0006488676881417632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4042864049988566
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.02043244574451819,
          "train_loss": 0.00821908749639988,
          "validation_loss": 0.006741788703948259,
          "train_metric": 0.0008277964661829174,
          "validation_metric": 0.0006790082552470267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.422336498999357
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.01991628808900714,
          "train_loss": 0.008091294206678867,
          "validation_loss": 0.006208683829754591,
          "train_metric": 0.0008149255299940705,
          "validation_metric": 0.0006253158790059388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4409610959992278
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.02028303372208029,
          "train_loss": 0.007624619174748659,
          "validation_loss": 0.00621773349121213,
          "train_metric": 0.0007679237169213593,
          "validation_metric": 0.0006262273527681828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4592252359998383
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.021837981301359832,
          "train_loss": 0.015077633783221245,
          "validation_loss": 0.012063510715961456,
          "train_metric": 0.0015185640659183264,
          "validation_metric": 0.0012149926042184234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.477633499998774
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.022418678097892553,
          "train_loss": 0.017863601446151733,
          "validation_loss": 0.010505378246307373,
          "train_metric": 0.0017991566564887762,
          "validation_metric": 0.0010580632369965315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4959174759987945
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.02074051642557606,
          "train_loss": 0.008713796734809875,
          "validation_loss": 0.00679820217192173,
          "train_metric": 0.0008776217582635581,
          "validation_metric": 0.0006846900214441121,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.513961972999823
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.02091357414610684,
          "train_loss": 0.00695782620459795,
          "validation_loss": 0.005414278712123632,
          "train_metric": 0.0007007667445577681,
          "validation_metric": 0.0005453063640743494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5322915469987493
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.01872300123795867,
          "train_loss": 0.008782991208136082,
          "validation_loss": 0.006837807595729828,
          "train_metric": 0.000884590728674084,
          "validation_metric": 0.000688678992446512,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5498572500000591
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.01792324607959017,
          "train_loss": 0.010058198124170303,
          "validation_loss": 0.008246536366641521,
          "train_metric": 0.0010130249429494143,
          "validation_metric": 0.0008305609226226807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5676035509986832
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.02047830674564466,
          "train_loss": 0.006938540376722813,
          "validation_loss": 0.005666838958859444,
          "train_metric": 0.0006988244713284075,
          "validation_metric": 0.0005707432865165174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.585731916999066
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.017742946598445997,
          "train_loss": 0.007146947085857391,
          "validation_loss": 0.005159191787242889,
          "train_metric": 0.0007198142120614648,
          "validation_metric": 0.0005196148995310068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.604414180999811
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.015859986538998783,
          "train_loss": 0.006620819680392742,
          "validation_loss": 0.004844682291150093,
          "train_metric": 0.000666824693325907,
          "validation_metric": 0.0004879386688116938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6237366850000399
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.017675598501227796,
          "train_loss": 0.007042163982987404,
          "validation_loss": 0.006047866307199001,
          "train_metric": 0.0007092609303072095,
          "validation_metric": 0.0006091188988648355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6418652459997247
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.018704302900005132,
          "train_loss": 0.007439001929014921,
          "validation_loss": 0.004698177799582481,
          "train_metric": 0.0007492291042581201,
          "validation_metric": 0.00047318320139311254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6606193400002667
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.017801414127461612,
          "train_loss": 0.0077289920300245285,
          "validation_loss": 0.004696471616625786,
          "train_metric": 0.0007784357876516879,
          "validation_metric": 0.00047301140148192644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6796263569995062
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.017327988578472286,
          "train_loss": 0.007159565109759569,
          "validation_loss": 0.005311921238899231,
          "train_metric": 0.0007210852345451713,
          "validation_metric": 0.0005349971470423043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6978525220001757
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.019240056513808668,
          "train_loss": 0.007069647312164307,
          "validation_loss": 0.006071386858820915,
          "train_metric": 0.0007120289374142885,
          "validation_metric": 0.0006114877760410309,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7158600040002057
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.01702152210054919,
          "train_loss": 0.006339521612972021,
          "validation_loss": 0.004493992775678635,
          "train_metric": 0.000638493278529495,
          "validation_metric": 0.0004526184930000454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7346890400003758
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.01652356644626707,
          "train_loss": 0.0069059100933372974,
          "validation_loss": 0.004701982252299786,
          "train_metric": 0.0006955379503779113,
          "validation_metric": 0.0004735664988402277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7531454709987884
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.018434127094224095,
          "train_loss": 0.006065586116164923,
          "validation_loss": 0.004679351579397917,
          "train_metric": 0.0006109036039561033,
          "validation_metric": 0.00047128720325417817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7713711630003672
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.016224962600972503,
          "train_loss": 0.00593546824529767,
          "validation_loss": 0.004206928424537182,
          "train_metric": 0.0005977986729703844,
          "validation_metric": 0.0004237063985783607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7903642829987803
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.016688763222191483,
          "train_loss": 0.006623825058341026,
          "validation_loss": 0.004288805183023214,
          "train_metric": 0.000667127373162657,
          "validation_metric": 0.00043195279431529343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.808746364999024
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.01788523275172338,
          "train_loss": 0.00751730240881443,
          "validation_loss": 0.006066605914384127,
          "train_metric": 0.0007571151363663375,
          "validation_metric": 0.0006110062822699547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.826921609999772
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.016076770785730332,
          "train_loss": 0.007309159263968468,
          "validation_loss": 0.004302409943193197,
          "train_metric": 0.000736151821911335,
          "validation_metric": 0.0004333229153417051,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8452740819993778
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.016659508750308305,
          "train_loss": 0.007744107861071825,
          "validation_loss": 0.005458468105643988,
          "train_metric": 0.0007799582090228796,
          "validation_metric": 0.0005497568636201322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8634048459989572
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.019642953178845346,
          "train_loss": 0.008689500391483307,
          "validation_loss": 0.008334516547620296,
          "train_metric": 0.000875174708198756,
          "validation_metric": 0.0008394219912588596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8813689999988128
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.016104584094136953,
          "train_loss": 0.005973420105874538,
          "validation_loss": 0.003991016652435064,
          "train_metric": 0.0006016209372319281,
          "validation_metric": 0.0004019605403300375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8991935419999209
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.020247525884769857,
          "train_loss": 0.007227648980915546,
          "validation_loss": 0.0049487529322505,
          "train_metric": 0.0007279424462467432,
          "validation_metric": 0.0004984202678315341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.917739870999867
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.01642485655611381,
          "train_loss": 0.006453614681959152,
          "validation_loss": 0.004913081414997578,
          "train_metric": 0.0006499843439087272,
          "validation_metric": 0.0004948275745846331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9365617359999305
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.017593999160453677,
          "train_loss": 0.005405389238148928,
          "validation_loss": 0.004419546574354172,
          "train_metric": 0.0005444110138341784,
          "validation_metric": 0.00044512058957479894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9545250859991938
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.01870020740898326,
          "train_loss": 0.006970438174903393,
          "validation_loss": 0.005532586947083473,
          "train_metric": 0.0007020370103418827,
          "validation_metric": 0.0005572219379246235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9721469919986703
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.019734516507014632,
          "train_loss": 0.00834713876247406,
          "validation_loss": 0.006787051446735859,
          "train_metric": 0.0008406933047808707,
          "validation_metric": 0.0006835670210421085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9901134989995626
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.018270554661285132,
          "train_loss": 0.009488333947956562,
          "validation_loss": 0.008507412858307362,
          "train_metric": 0.0009556301520206034,
          "validation_metric": 0.0008568353950977325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0078014909995545
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.01848405721830204,
          "train_loss": 0.005419939756393433,
          "validation_loss": 0.003980170004069805,
          "train_metric": 0.0005458765081129968,
          "validation_metric": 0.0004008680989500135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0265986129998055
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.016767044668085873,
          "train_loss": 0.005923427641391754,
          "validation_loss": 0.004707443527877331,
          "train_metric": 0.0005965859163552523,
          "validation_metric": 0.0004741165030281991,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0446207639997738
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.019296595943160355,
          "train_loss": 0.010118525475263596,
          "validation_loss": 0.009607422165572643,
          "train_metric": 0.0010191007750108838,
          "validation_metric": 0.0009676243644207716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0623316149994935
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.018195914279203862,
          "train_loss": 0.007178700063377619,
          "validation_loss": 0.004543110262602568,
          "train_metric": 0.0007230123737826943,
          "validation_metric": 0.00045756547478958964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0801150429997506
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.016450313210953027,
          "train_loss": 0.0060655646957457066,
          "validation_loss": 0.004672654438763857,
          "train_metric": 0.0006109015084803104,
          "validation_metric": 0.00047061266377568245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0980377919986495
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.016832669032737613,
          "train_loss": 0.00535861449316144,
          "validation_loss": 0.003162505105137825,
          "train_metric": 0.000539699976798147,
          "validation_metric": 0.0003185159293934703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1168021069988754
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.016760822210926563,
          "train_loss": 0.005894267000257969,
          "validation_loss": 0.004963962361216545,
          "train_metric": 0.0005936490488238633,
          "validation_metric": 0.000499952060636133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1351831880001555
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.016752287279814482,
          "train_loss": 0.008461756631731987,
          "validation_loss": 0.005172857083380222,
          "train_metric": 0.0008522372227162123,
          "validation_metric": 0.0005209912196733057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.153362847999233
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.02162050415063277,
          "train_loss": 0.016243886202573776,
          "validation_loss": 0.013013671152293682,
          "train_metric": 0.001636024913750589,
          "validation_metric": 0.0013106891419738531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.171314531999087
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.021231312362942845,
          "train_loss": 0.007031260058283806,
          "validation_loss": 0.004676053300499916,
          "train_metric": 0.0007081628427840769,
          "validation_metric": 0.0004709549539256841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1894978339987574
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.016307440353557467,
          "train_loss": 0.010559296235442162,
          "validation_loss": 0.004843863192945719,
          "train_metric": 0.0010634935460984707,
          "validation_metric": 0.0004878561303485185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2074019179999596
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.015765173418913037,
          "train_loss": 0.008880356326699257,
          "validation_loss": 0.005378429777920246,
          "train_metric": 0.0008943970315158367,
          "validation_metric": 0.000541695742867887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.225514978999854
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.020129133481532335,
          "train_loss": 0.005907667335122824,
          "validation_loss": 0.004508297424763441,
          "train_metric": 0.0005949986516498029,
          "validation_metric": 0.0004540592199191451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2434752589997515
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.02019098005257547,
          "train_loss": 0.011833691969513893,
          "validation_loss": 0.0103979567065835,
          "train_metric": 0.0011918462114408612,
          "validation_metric": 0.0010472440626472235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2614651509993564
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.019031270872801542,
          "train_loss": 0.006276782602071762,
          "validation_loss": 0.004721841309219599,
          "train_metric": 0.0006321744876913726,
          "validation_metric": 0.00047556651406921446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2798257379999995
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.015632274036761373,
          "train_loss": 0.008499611169099808,
          "validation_loss": 0.0060354373417794704,
          "train_metric": 0.0008560497662983835,
          "validation_metric": 0.0006078670849092305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2973999779987935
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.018648781056981534,
          "train_loss": 0.008180755190551281,
          "validation_loss": 0.00552435452118516,
          "train_metric": 0.0008239357266575098,
          "validation_metric": 0.0005563927115872502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4196535859991855
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.019756145775318146,
          "train_loss": 0.006423559505492449,
          "validation_loss": 0.005913176573812962,
          "train_metric": 0.0006469573709182441,
          "validation_metric": 0.0005955534870736301,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4381239889989956
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.016560457821469754,
          "train_loss": 0.004551413003355265,
          "validation_loss": 0.0032286308705806732,
          "train_metric": 0.0004584016278386116,
          "validation_metric": 0.0003251758753322065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.456265037999401
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.013629371183924377,
          "train_loss": 0.008811712265014648,
          "validation_loss": 0.006569406483322382,
          "train_metric": 0.0008874834165908396,
          "validation_metric": 0.0006616465398110449,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4746703679993516
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.016748552268836647,
          "train_loss": 0.006338904146105051,
          "validation_loss": 0.0038602508138865232,
          "train_metric": 0.0006384311127476394,
          "validation_metric": 0.0003887903585564345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4927150160001474
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.016429827257525176,
          "train_loss": 0.005176292732357979,
          "validation_loss": 0.00341459340415895,
          "train_metric": 0.0005213372642174363,
          "validation_metric": 0.00034390532528050244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5110793849999027
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.015995135327102616,
          "train_loss": 0.005196546670049429,
          "validation_loss": 0.004011306446045637,
          "train_metric": 0.0005233771516941488,
          "validation_metric": 0.0004040040948893875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5286491560000286
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.01654911378864199,
          "train_loss": 0.006015643011778593,
          "validation_loss": 0.004039371386170387,
          "train_metric": 0.0006058734725229442,
          "validation_metric": 0.0004068307171110064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.546049773998675
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.014816525799687952,
          "train_loss": 0.004868783056735992,
          "validation_loss": 0.0034089363180100918,
          "train_metric": 0.0004903659573756158,
          "validation_metric": 0.0003433355886954814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.563696836999952
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.013942702324129641,
          "train_loss": 0.0053566331043839455,
          "validation_loss": 0.0038690674118697643,
          "train_metric": 0.0005395004409365356,
          "validation_metric": 0.00038967837463133037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.581261507999443
        }
      ],
      "best": {
        "epoch": 108,
        "updates": 1728
      },
      "best_validation_metric": 0.0003185159293934703,
      "config": {
        "label": "Dropout 0.1",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.1,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.017847014579292254
    },
    "d7d20cfbbc18_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.728500097990036,
          "train_loss": 0.32748350501060486,
          "validation_loss": 0.25605568289756775,
          "train_metric": 0.03298294171690941,
          "validation_metric": 0.025788988918066025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.025571853999281302
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.19025728153064847,
          "train_loss": 0.12283626198768616,
          "validation_loss": 0.10460524260997772,
          "train_metric": 0.012371617369353771,
          "validation_metric": 0.01053545717149973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.053104949998669326
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.12047328054904938,
          "train_loss": 0.09659405797719955,
          "validation_loss": 0.0805262103676796,
          "train_metric": 0.009728599339723587,
          "validation_metric": 0.008110305294394493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07546776599883742
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.08932839496992528,
          "train_loss": 0.0805170089006424,
          "validation_loss": 0.06445010006427765,
          "train_metric": 0.00810937862843275,
          "validation_metric": 0.006491177715361118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0878561350000382
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07180124777369201,
          "train_loss": 0.06417941302061081,
          "validation_loss": 0.052139848470687866,
          "train_metric": 0.006463915575295687,
          "validation_metric": 0.005251334514468908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09999494899966521
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06325704464688897,
          "train_loss": 0.0624699704349041,
          "validation_loss": 0.050321128219366074,
          "train_metric": 0.006291746627539396,
          "validation_metric": 0.0050681596621870995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11251469499984523
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06037090043537319,
          "train_loss": 0.06244666874408722,
          "validation_loss": 0.05068628862500191,
          "train_metric": 0.006289399694651365,
          "validation_metric": 0.005104937124997377,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12458792100005667
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.059591412311419845,
          "train_loss": 0.0551791712641716,
          "validation_loss": 0.0441298708319664,
          "train_metric": 0.005557443480938673,
          "validation_metric": 0.004444599151611328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13692752799943264
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05139579763635993,
          "train_loss": 0.04721299931406975,
          "validation_loss": 0.03837432712316513,
          "train_metric": 0.004755120724439621,
          "validation_metric": 0.0038649216294288635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1493093239987502
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.04574782948475331,
          "train_loss": 0.04561600834131241,
          "validation_loss": 0.03735892102122307,
          "train_metric": 0.004594277124851942,
          "validation_metric": 0.0037626540288329124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16186448999906133
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0437198638683185,
          "train_loss": 0.04204858839511871,
          "validation_loss": 0.033534687012434006,
          "train_metric": 0.0042349798604846,
          "validation_metric": 0.0033774906769394875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17454504999841447
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.03942157095298171,
          "train_loss": 0.03619734197854996,
          "validation_loss": 0.029675962403416634,
          "train_metric": 0.0036456636153161526,
          "validation_metric": 0.0029888541903346777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1869354589998693
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03510583657771349,
          "train_loss": 0.03373274207115173,
          "validation_loss": 0.027667436748743057,
          "train_metric": 0.003397438209503889,
          "validation_metric": 0.002786562778055668,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19965108699943812
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03235916834091768,
          "train_loss": 0.03133251145482063,
          "validation_loss": 0.02592913992702961,
          "train_metric": 0.0031556959729641676,
          "validation_metric": 0.002611488103866577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21246613899893418
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03129324276233092,
          "train_loss": 0.0326116718351841,
          "validation_loss": 0.02754851244390011,
          "train_metric": 0.003284528385847807,
          "validation_metric": 0.0027745855040848255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22452740399967297
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.028896714851725847,
          "train_loss": 0.030652284622192383,
          "validation_loss": 0.02598833292722702,
          "train_metric": 0.00308718578889966,
          "validation_metric": 0.002617449499666691,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23624466299952473
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.026621754543157294,
          "train_loss": 0.024554364383220673,
          "validation_loss": 0.02126566879451275,
          "train_metric": 0.0024730258155614138,
          "validation_metric": 0.002141800010576844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2484281289998762
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.026915886672213674,
          "train_loss": 0.029723035171628,
          "validation_loss": 0.023281192407011986,
          "train_metric": 0.002993595087900758,
          "validation_metric": 0.002344796434044838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26025082799969823
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.027702209452399984,
          "train_loss": 0.024008754640817642,
          "validation_loss": 0.017625000327825546,
          "train_metric": 0.0024180738255381584,
          "validation_metric": 0.0017751253908500075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2723504639998282
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.020750796975335106,
          "train_loss": 0.019556082785129547,
          "validation_loss": 0.016577966511249542,
          "train_metric": 0.0019696170929819345,
          "validation_metric": 0.0016696719685569406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28456637899944326
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.020576069422531873,
          "train_loss": 0.017873117700219154,
          "validation_loss": 0.01555734034627676,
          "train_metric": 0.0018001149874180555,
          "validation_metric": 0.0015668784035369754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29704544199921656
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.020247126056347042,
          "train_loss": 0.017944050952792168,
          "validation_loss": 0.014023429714143276,
          "train_metric": 0.0018072591628879309,
          "validation_metric": 0.0014123886357992887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.309486473999641
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.018817331874743104,
          "train_loss": 0.016332851722836494,
          "validation_loss": 0.014480787329375744,
          "train_metric": 0.001644984818994999,
          "validation_metric": 0.0014584519667550921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32140201999936835
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.015988841303624213,
          "train_loss": 0.01532297395169735,
          "validation_loss": 0.013524141162633896,
          "train_metric": 0.0015432738000527024,
          "validation_metric": 0.001362101873382926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33359952899991185
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.01738046415266581,
          "train_loss": 0.018397396430373192,
          "validation_loss": 0.013566923327744007,
          "train_metric": 0.0018529182998463511,
          "validation_metric": 0.0013664108701050282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34559877199899347
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.01646898075705394,
          "train_loss": 0.016528358682990074,
          "validation_loss": 0.015661781653761864,
          "train_metric": 0.001664675772190094,
          "validation_metric": 0.00157739722635597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3573511750000762
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.014316251908894628,
          "train_loss": 0.013044876977801323,
          "validation_loss": 0.011559470556676388,
          "train_metric": 0.0013138321228325367,
          "validation_metric": 0.0011642276076599956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36945534299957217
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.013307352201081812,
          "train_loss": 0.012899477034807205,
          "validation_loss": 0.011417489498853683,
          "train_metric": 0.001299188006669283,
          "validation_metric": 0.001149927731603384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3820756299992354
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.015201686095679179,
          "train_loss": 0.015896521508693695,
          "validation_loss": 0.012359514832496643,
          "train_metric": 0.001601039431989193,
          "validation_metric": 0.0012448051711544394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3940792989997135
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.014801724842982367,
          "train_loss": 0.01260439120233059,
          "validation_loss": 0.010798940435051918,
          "train_metric": 0.001269468106329441,
          "validation_metric": 0.0010876297019422054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4063752409983863
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.018144490546546876,
          "train_loss": 0.011910106055438519,
          "validation_loss": 0.009505937807261944,
          "train_metric": 0.001199542311951518,
          "validation_metric": 0.0009574032737873495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4188675159984996
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.015545382717391476,
          "train_loss": 0.012557906098663807,
          "validation_loss": 0.008931681513786316,
          "train_metric": 0.0012647862313315272,
          "validation_metric": 0.0008995661046355963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4314005829983216
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.013211223966209218,
          "train_loss": 0.010604162700474262,
          "validation_loss": 0.008407311514019966,
          "train_metric": 0.0010680124396458268,
          "validation_metric": 0.0008467536536045372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44426257599843666
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.011625336133874953,
          "train_loss": 0.013516976498067379,
          "validation_loss": 0.009021398611366749,
          "train_metric": 0.0013613804476335645,
          "validation_metric": 0.0009086023201234639,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4563923319983587
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.012681668944424018,
          "train_loss": 0.014462439343333244,
          "validation_loss": 0.01360971387475729,
          "train_metric": 0.0014566038735210896,
          "validation_metric": 0.0013707205653190613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46815794599933724
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.01557280879933387,
          "train_loss": 0.02201736718416214,
          "validation_loss": 0.015061598271131516,
          "train_metric": 0.0022175086196511984,
          "validation_metric": 0.0015169490361586213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4799506549989019
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.011511159260408022,
          "train_loss": 0.009732195176184177,
          "validation_loss": 0.008330872282385826,
          "train_metric": 0.000980190932750702,
          "validation_metric": 0.0008390549337491393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49201369599904865
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.01032086274062749,
          "train_loss": 0.00870879739522934,
          "validation_loss": 0.007191747892647982,
          "train_metric": 0.0008771181455813348,
          "validation_metric": 0.0007243264699354768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5046021469988773
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.009547622292302549,
          "train_loss": 0.008846729062497616,
          "validation_loss": 0.0067849052138626575,
          "train_metric": 0.0008910102187655866,
          "validation_metric": 0.0006833508377894759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5193487029991957
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.011415118613513187,
          "train_loss": 0.012171742506325245,
          "validation_loss": 0.008104430511593819,
          "train_metric": 0.0012258932692930102,
          "validation_metric": 0.0008162487065419555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5467938459987636
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.00963783822953701,
          "train_loss": 0.008970755152404308,
          "validation_loss": 0.006819006521254778,
          "train_metric": 0.0009035015245899558,
          "validation_metric": 0.0006867853808216751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5761824679993879
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.009801773325307295,
          "train_loss": 0.008915501646697521,
          "validation_loss": 0.006469665560871363,
          "train_metric": 0.0008979365229606628,
          "validation_metric": 0.0006516010034829378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6142224169998372
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.010288793899235316,
          "train_loss": 0.007891234010457993,
          "validation_loss": 0.006398339290171862,
          "train_metric": 0.0007947761914692819,
          "validation_metric": 0.0006444172468036413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6393481459999748
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.00868916958279442,
          "train_loss": 0.008011811412870884,
          "validation_loss": 0.006478559225797653,
          "train_metric": 0.0008069202885963023,
          "validation_metric": 0.0006524967611767352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.65573939999922
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.012220922129927203,
          "train_loss": 0.009155090898275375,
          "validation_loss": 0.007426150608807802,
          "train_metric": 0.0009220673236995935,
          "validation_metric": 0.0007479347405023873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6694905139993352
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.011436141168815084,
          "train_loss": 0.010045844130218029,
          "validation_loss": 0.007035197224467993,
          "train_metric": 0.001011780695989728,
          "validation_metric": 0.0007085593533702195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.68213963999915
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.010630432283505797,
          "train_loss": 0.011331791058182716,
          "validation_loss": 0.010522670112550259,
          "train_metric": 0.0011412964668124914,
          "validation_metric": 0.0010598046937957406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6943759609985136
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.009336302377050743,
          "train_loss": 0.014058622531592846,
          "validation_loss": 0.013844212517142296,
          "train_metric": 0.0014159330166876316,
          "validation_metric": 0.0013943383237347007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7061414129984769
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.011501986533403397,
          "train_loss": 0.009885628707706928,
          "validation_loss": 0.007653104607015848,
          "train_metric": 0.0009956443682312965,
          "validation_metric": 0.0007707925979048014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.717964816998574
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.010613130398269277,
          "train_loss": 0.007742475718259811,
          "validation_loss": 0.005785346496850252,
          "train_metric": 0.0007797937723807991,
          "validation_metric": 0.0005826789420098066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7304372679991502
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008738365635508671,
          "train_loss": 0.01197832077741623,
          "validation_loss": 0.011343628168106079,
          "train_metric": 0.0012064125621691346,
          "validation_metric": 0.0011424885597079992,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7421099819985102
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008023067421163432,
          "train_loss": 0.007562712766230106,
          "validation_loss": 0.005323607940226793,
          "train_metric": 0.0007616888033226132,
          "validation_metric": 0.0005361742805689573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7544015479998052
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.007868653468904085,
          "train_loss": 0.009297807700932026,
          "validation_loss": 0.006549007259309292,
          "train_metric": 0.0009364410652779043,
          "validation_metric": 0.0006595919840037823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7664889319985377
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.008324112699483521,
          "train_loss": 0.009426431730389595,
          "validation_loss": 0.009254010394215584,
          "train_metric": 0.0009493957622908056,
          "validation_metric": 0.0009320300887338817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7783386389983207
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.012368215539027005,
          "train_loss": 0.008601157926023006,
          "validation_loss": 0.0064111435785889626,
          "train_metric": 0.0008662770851515234,
          "validation_metric": 0.000645706953946501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7900879439985147
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.009687915415270254,
          "train_loss": 0.007840674370527267,
          "validation_loss": 0.005630690138787031,
          "train_metric": 0.0007896839524619281,
          "validation_metric": 0.0005671025137417018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.801700563999475
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008077269536443055,
          "train_loss": 0.006787789985537529,
          "validation_loss": 0.0054641603492200375,
          "train_metric": 0.000683641352225095,
          "validation_metric": 0.0005503302672877908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.813396802999705
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.006685821295832284,
          "train_loss": 0.006263366434723139,
          "validation_loss": 0.004662930499762297,
          "train_metric": 0.0006308233714662492,
          "validation_metric": 0.0004696332907769829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.825323686000047
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.007334431415074505,
          "train_loss": 0.006555630825459957,
          "validation_loss": 0.005838551092892885,
          "train_metric": 0.0006602591020055115,
          "validation_metric": 0.0005880374810658395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8370048289998522
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006905647707753815,
          "train_loss": 0.006005297414958477,
          "validation_loss": 0.004659541416913271,
          "train_metric": 0.0006048315553925931,
          "validation_metric": 0.0004692919319495559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.849229227998876
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.008627982999314554,
          "train_loss": 0.006474621593952179,
          "validation_loss": 0.005489842034876347,
          "train_metric": 0.0006521001923829317,
          "validation_metric": 0.0005529167247004807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8612551139995048
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.00690813015535241,
          "train_loss": 0.005752439610660076,
          "validation_loss": 0.004629974253475666,
          "train_metric": 0.0005793646560050547,
          "validation_metric": 0.0004663140862248838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8736670809994393
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006324854024569504,
          "train_loss": 0.0058545880019664764,
          "validation_loss": 0.0046927728690207005,
          "train_metric": 0.0005896526272408664,
          "validation_metric": 0.0004726388433482498,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8859745359986846
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0068351525405887514,
          "train_loss": 0.008917616680264473,
          "validation_loss": 0.008220354095101357,
          "train_metric": 0.0008981497376225889,
          "validation_metric": 0.0008279239409603179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8982630259997677
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.008000331319635734,
          "train_loss": 0.008382441475987434,
          "validation_loss": 0.00753337936475873,
          "train_metric": 0.000844248803332448,
          "validation_metric": 0.0007587344152852893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.910134632998961
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.006737448842613958,
          "train_loss": 0.006288199685513973,
          "validation_loss": 0.004485901445150375,
          "train_metric": 0.0006333244382403791,
          "validation_metric": 0.0004518035857472569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9230422639993776
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007921891658043023,
          "train_loss": 0.008712094277143478,
          "validation_loss": 0.005809281952679157,
          "train_metric": 0.0008774501620791852,
          "validation_metric": 0.000585089554078877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9354525599992485
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00778336486837361,
          "train_loss": 0.006150304339826107,
          "validation_loss": 0.004038909450173378,
          "train_metric": 0.0006194360903464258,
          "validation_metric": 0.00040678418008610606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9480631849983183
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.005680954760464374,
          "train_loss": 0.005995623301714659,
          "validation_loss": 0.005130578298121691,
          "train_metric": 0.0006038572755642235,
          "validation_metric": 0.000516733096446842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9603560319992539
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0060421959205996245,
          "train_loss": 0.005446858238428831,
          "validation_loss": 0.0039938450790941715,
          "train_metric": 0.0005485875881277025,
          "validation_metric": 0.00040224549593403935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9727538600000116
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006552721722982824,
          "train_loss": 0.005445307586342096,
          "validation_loss": 0.0039313375018537045,
          "train_metric": 0.0005484314169734716,
          "validation_metric": 0.0003959499590564519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9859602199994697
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006271286911214702,
          "train_loss": 0.0051545994356274605,
          "validation_loss": 0.004310379736125469,
          "train_metric": 0.0005191523814573884,
          "validation_metric": 0.0004341256571933627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0779430129987304
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.009956596666597761,
          "train_loss": 0.017113812267780304,
          "validation_loss": 0.013151607476174831,
          "train_metric": 0.001723640481941402,
          "validation_metric": 0.0013245816808193922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0908429369992518
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.009974227752536535,
          "train_loss": 0.006625129841268063,
          "validation_loss": 0.005291996989399195,
          "train_metric": 0.0006672588060609996,
          "validation_metric": 0.0005329904961399734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.106879842998751
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006168481479107868,
          "train_loss": 0.0054809353314340115,
          "validation_loss": 0.004028170369565487,
          "train_metric": 0.0005520197446458042,
          "validation_metric": 0.000405702565331012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1192835709989595
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.008603171445429325,
          "train_loss": 0.006894308608025312,
          "validation_loss": 0.006176698487251997,
          "train_metric": 0.0006943694897927344,
          "validation_metric": 0.0006220944924280047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1318771309997828
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007957234207424335,
          "train_loss": 0.008646661415696144,
          "validation_loss": 0.0065186526626348495,
          "train_metric": 0.000870860181748867,
          "validation_metric": 0.0006565348012372851,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1441537299997435
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.008243853430030867,
          "train_loss": 0.00981737207621336,
          "validation_loss": 0.010102519765496254,
          "train_metric": 0.0009887696942314506,
          "validation_metric": 0.0010174887720495462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1558185649992083
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006789916064008139,
          "train_loss": 0.005592592526227236,
          "validation_loss": 0.004652687348425388,
          "train_metric": 0.0005632654065266252,
          "validation_metric": 0.0004686016764026135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1676143639997463
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.00580319495202275,
          "train_loss": 0.005281102377921343,
          "validation_loss": 0.004064279142767191,
          "train_metric": 0.0005318932817317545,
          "validation_metric": 0.0004093392926733941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1797830469986366
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005682255345163867,
          "train_loss": 0.004670701455324888,
          "validation_loss": 0.0037764455191791058,
          "train_metric": 0.00047041592188179493,
          "validation_metric": 0.00038034978206269443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1922955179998098
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.005088636578875594,
          "train_loss": 0.005166700575500727,
          "validation_loss": 0.0035962266847491264,
          "train_metric": 0.0005203711334615946,
          "validation_metric": 0.0003621987416408956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.205220998999721
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006093093528761528,
          "train_loss": 0.005610364954918623,
          "validation_loss": 0.00537495082244277,
          "train_metric": 0.0005650554085150361,
          "validation_metric": 0.0005413453909568489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2178716879989224
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.007265872249263339,
          "train_loss": 0.007519181817770004,
          "validation_loss": 0.005037151742726564,
          "train_metric": 0.0007573043694719672,
          "validation_metric": 0.0005073235370218754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.229659463999269
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0061193279034341685,
          "train_loss": 0.004973108880221844,
          "validation_loss": 0.004296155646443367,
          "train_metric": 0.0005008733132854104,
          "validation_metric": 0.00043269302113913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2412870169991947
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.008844511321512982,
          "train_loss": 0.005378494970500469,
          "validation_loss": 0.004169309511780739,
          "train_metric": 0.0005417022621259093,
          "validation_metric": 0.0004199176328256726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.253447563998634
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005669804915669374,
          "train_loss": 0.0045824116095900536,
          "validation_loss": 0.003294224850833416,
          "train_metric": 0.00046152371214702725,
          "validation_metric": 0.0003317822702229023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2660152029984602
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.005271380869089626,
          "train_loss": 0.0048345536924898624,
          "validation_loss": 0.0033399772364646196,
          "train_metric": 0.00048691852134652436,
          "validation_metric": 0.00033639022149145603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2780232449986215
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005104001036670525,
          "train_loss": 0.004279932472854853,
          "validation_loss": 0.0037712878547608852,
          "train_metric": 0.0004310591612011194,
          "validation_metric": 0.00037983033689670265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2900782179985981
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004830286867218092,
          "train_loss": 0.004640891216695309,
          "validation_loss": 0.003402076195925474,
          "train_metric": 0.0004674135416280478,
          "validation_metric": 0.00034264466376043856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3018886559984821
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.00471690806443803,
          "train_loss": 0.004531655926257372,
          "validation_loss": 0.003670204896479845,
          "train_metric": 0.0004564117407426238,
          "validation_metric": 0.000369649613276124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3140055499989103
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004665426327846944,
          "train_loss": 0.004014521837234497,
          "validation_loss": 0.003242441453039646,
          "train_metric": 0.0004043279041070491,
          "validation_metric": 0.0003265668638050556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.326491856998473
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.005101119488244876,
          "train_loss": 0.004324267618358135,
          "validation_loss": 0.003883620258420706,
          "train_metric": 0.0004355243581812829,
          "validation_metric": 0.00039114392711780965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3388364109996473
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004682626677094959,
          "train_loss": 0.003964405041188002,
          "validation_loss": 0.0031244794372469187,
          "train_metric": 0.0003992803394794464,
          "validation_metric": 0.0003146860981360078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3519043829983275
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.00585375265654875,
          "train_loss": 0.005264627747237682,
          "validation_loss": 0.004929744638502598,
          "train_metric": 0.0005302341305650771,
          "validation_metric": 0.0004965058760717511,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.364608710999164
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.005366782104829326,
          "train_loss": 0.004235393833369017,
          "validation_loss": 0.0033937899861484766,
          "train_metric": 0.0004265733587089926,
          "validation_metric": 0.0003418100532144308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3766192339990084
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005604555051831994,
          "train_loss": 0.004185291938483715,
          "validation_loss": 0.003764444263651967,
          "train_metric": 0.0004215272783767432,
          "validation_metric": 0.0003791410999838263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.38932627199938
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.00465164179331623,
          "train_loss": 0.003927567973732948,
          "validation_loss": 0.0028497339226305485,
          "train_metric": 0.00039557021227665246,
          "validation_metric": 0.00028701472911052406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4016766379991168
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.004213943822833244,
          "train_loss": 0.003922795876860619,
          "validation_loss": 0.0029739185702055693,
          "train_metric": 0.00039508964982815087,
          "validation_metric": 0.0002995222166646272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4144635109987576
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0044291301019256935,
          "train_loss": 0.006234404630959034,
          "validation_loss": 0.00416173879057169,
          "train_metric": 0.0006279062945395708,
          "validation_metric": 0.0004191550542600453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4262567549994856
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005240917707851622,
          "train_loss": 0.00418194942176342,
          "validation_loss": 0.0036627103108912706,
          "train_metric": 0.00042119063436985016,
          "validation_metric": 0.0003688947472255677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4389139029990474
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.006090985014452599,
          "train_loss": 0.01083461195230484,
          "validation_loss": 0.010316397063434124,
          "train_metric": 0.0010912223951891065,
          "validation_metric": 0.0010390296811237931,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4512270809991605
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.005212657910306007,
          "train_loss": 0.004508474376052618,
          "validation_loss": 0.002871852368116379,
          "train_metric": 0.00045407703146338463,
          "validation_metric": 0.00028924239450134337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.46359234099873
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.004414431263285223,
          "train_loss": 0.003557204268872738,
          "validation_loss": 0.002832036232575774,
          "train_metric": 0.00035826858947984874,
          "validation_metric": 0.00028523229411803186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.475460571999065
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0039505833556177095,
          "train_loss": 0.006458071991801262,
          "validation_loss": 0.006466188933700323,
          "train_metric": 0.0006504334742203355,
          "validation_metric": 0.0006512508261948824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4874811099998624
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004890365264145657,
          "train_loss": 0.003806433640420437,
          "validation_loss": 0.0028659908566623926,
          "train_metric": 0.00038337006117217243,
          "validation_metric": 0.0002886521106120199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4995002409996232
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004308742587454617,
          "train_loss": 0.00396642554551363,
          "validation_loss": 0.003019891679286957,
          "train_metric": 0.0003994838916696608,
          "validation_metric": 0.00030415243236348033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5117708429988852
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0038823579234303907,
          "train_loss": 0.004179538693279028,
          "validation_loss": 0.00401037372648716,
          "train_metric": 0.0004209478502161801,
          "validation_metric": 0.00040391014772467315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6091292949986382
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.005163143418030813,
          "train_loss": 0.003510484006255865,
          "validation_loss": 0.0027991815004497766,
          "train_metric": 0.0003535630530677736,
          "validation_metric": 0.00028192330501042306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6220772699998633
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.006352085554681253,
          "train_loss": 0.005032664630562067,
          "validation_loss": 0.004722778219729662,
          "train_metric": 0.0005068714963272214,
          "validation_metric": 0.00047566095599904656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.63437327399879
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.011078066076152027,
          "train_loss": 0.013898147270083427,
          "validation_loss": 0.011521977372467518,
          "train_metric": 0.0013997703790664673,
          "validation_metric": 0.0011604512110352516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6465215230000467
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.009402522948221304,
          "train_loss": 0.00472923694178462,
          "validation_loss": 0.0032230508513748646,
          "train_metric": 0.00047631142660975456,
          "validation_metric": 0.00032461388036608696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6585711909992824
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.005292398578603752,
          "train_loss": 0.004137364216148853,
          "validation_loss": 0.0031439349986612797,
          "train_metric": 0.00041670011705718935,
          "validation_metric": 0.00031664560083299875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6713529799999378
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004527361114014639,
          "train_loss": 0.004239290952682495,
          "validation_loss": 0.0030514614190906286,
          "train_metric": 0.00042696588207036257,
          "validation_metric": 0.00030733199673704803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6833823079996364
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0045561344813904725,
          "train_loss": 0.004688340239226818,
          "validation_loss": 0.004473282489925623,
          "train_metric": 0.00047219247790053487,
          "validation_metric": 0.00045053267967887223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6955957729987858
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0062237804449978285,
          "train_loss": 0.007977918721735477,
          "validation_loss": 0.007022134028375149,
          "train_metric": 0.0008035067585296929,
          "validation_metric": 0.0007072435691952705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7075709369983088
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.00468945296597667,
          "train_loss": 0.004213213454931974,
          "validation_loss": 0.0027932808734476566,
          "train_metric": 0.0004243393777869642,
          "validation_metric": 0.00028132900479249656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.72024432499893
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.004842383888899349,
          "train_loss": 0.004966671112924814,
          "validation_loss": 0.0036464938893914223,
          "train_metric": 0.0005002248799428344,
          "validation_metric": 0.0003672615275718272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.732997090999561
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.006367673631757498,
          "train_loss": 0.004723533987998962,
          "validation_loss": 0.00405119126662612,
          "train_metric": 0.00047573703341186047,
          "validation_metric": 0.00040802120929583907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7453211960000772
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.004751844535348937,
          "train_loss": 0.0034999230410903692,
          "validation_loss": 0.0027982774190604687,
          "train_metric": 0.00035249945358373225,
          "validation_metric": 0.0002818322682287544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7568922109985579
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004092620045412332,
          "train_loss": 0.0035254410468041897,
          "validation_loss": 0.002990034641698003,
          "train_metric": 0.0003550694964360446,
          "validation_metric": 0.0003011453081853688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7695658399989043
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0046533222775906324,
          "train_loss": 0.0054655554704368114,
          "validation_loss": 0.003977558575570583,
          "train_metric": 0.000550470722373575,
          "validation_metric": 0.00040060514584183693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7817032650000328
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.005958599802397657,
          "train_loss": 0.0040320660918951035,
          "validation_loss": 0.003433543723076582,
          "train_metric": 0.0004060949431732297,
          "validation_metric": 0.00034581395448185503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7934756899994682
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.00544908291703905,
          "train_loss": 0.0035949128214269876,
          "validation_loss": 0.0025188811123371124,
          "train_metric": 0.0003620664356276393,
          "validation_metric": 0.0002536924439482391,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8059185669990256
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.004435299335455056,
          "train_loss": 0.004218417219817638,
          "validation_loss": 0.0030528202187269926,
          "train_metric": 0.00042486353777348995,
          "validation_metric": 0.00030746887205168605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8179886889993213
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.004748308077978436,
          "train_loss": 0.0034238756634294987,
          "validation_loss": 0.0027917467523366213,
          "train_metric": 0.00034484025673009455,
          "validation_metric": 0.0002811745216604322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.829717778999111
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0049132774729514495,
          "train_loss": 0.003089688718318939,
          "validation_loss": 0.002417254727333784,
          "train_metric": 0.0003111821715719998,
          "validation_metric": 0.000243457019678317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8419822209998529
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.003958311845053686,
          "train_loss": 0.002940010279417038,
          "validation_loss": 0.0023474423214793205,
          "train_metric": 0.00029610705678351223,
          "validation_metric": 0.0002364258107263595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8545765259987093
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0036050653143320233,
          "train_loss": 0.004166924394667149,
          "validation_loss": 0.002754914341494441,
          "train_metric": 0.00041967740980908275,
          "validation_metric": 0.00027746488922275603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8664900280000438
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.004890520103799645,
          "train_loss": 0.0033935927785933018,
          "validation_loss": 0.003176419995725155,
          "train_metric": 0.0003417902044020593,
          "validation_metric": 0.0003199174243491143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8783100599994214
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.006259809131734073,
          "train_loss": 0.004467905033379793,
          "validation_loss": 0.0031075661536306143,
          "train_metric": 0.0004499910573940724,
          "validation_metric": 0.00031298265093937516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.889975323998442
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.004180362833722029,
          "train_loss": 0.0033807591535151005,
          "validation_loss": 0.0031126164831221104,
          "train_metric": 0.00034049764508381486,
          "validation_metric": 0.0003134913567919284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.902124599999297
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.006184936122735962,
          "train_loss": 0.0037688957527279854,
          "validation_loss": 0.002745382022112608,
          "train_metric": 0.0003795894735958427,
          "validation_metric": 0.0002765048120636493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9141257369992672
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.003961908783821855,
          "train_loss": 0.0036944595631211996,
          "validation_loss": 0.0025934597942978144,
          "train_metric": 0.00037209244328550994,
          "validation_metric": 0.0002612037642393261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.925884292999399
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.004193726119410712,
          "train_loss": 0.0031704436987638474,
          "validation_loss": 0.00274392357096076,
          "train_metric": 0.0003193155280314386,
          "validation_metric": 0.00027635792503133416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.937798720999126
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0038887530972715467,
          "train_loss": 0.00336978817358613,
          "validation_loss": 0.0027734320610761642,
          "train_metric": 0.00033939271816052496,
          "validation_metric": 0.0002793299499899149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9496249929998157
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0034807822230504826,
          "train_loss": 0.00327123561874032,
          "validation_loss": 0.0027881201822310686,
          "train_metric": 0.0003294668858870864,
          "validation_metric": 0.00028080923948436975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9614709229990694
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0033738553465809673,
          "train_loss": 0.0034786055330187082,
          "validation_loss": 0.0032120542600750923,
          "train_metric": 0.00035035243490710855,
          "validation_metric": 0.0003235063049942255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9731648529996164
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.003201572166290134,
          "train_loss": 0.0038107712753117085,
          "validation_loss": 0.0032442156225442886,
          "train_metric": 0.000383806909667328,
          "validation_metric": 0.0003267454740125686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.98491472799833
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.00295423873103573,
          "train_loss": 0.0035151962656527758,
          "validation_loss": 0.002453622641041875,
          "train_metric": 0.000354037678334862,
          "validation_metric": 0.0002471199259161949,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9967244089984888
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.003758968072361313,
          "train_loss": 0.0027505101170390844,
          "validation_loss": 0.0020132006611675024,
          "train_metric": 0.00027702131774276495,
          "validation_metric": 0.00020276221039239317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0091422199984663
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0030502243316732347,
          "train_loss": 0.0037231820169836283,
          "validation_loss": 0.0033959904685616493,
          "train_metric": 0.00037498524761758745,
          "validation_metric": 0.0003420317079871893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0209822269989672
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.002943591527582612,
          "train_loss": 0.0033974749967455864,
          "validation_loss": 0.003091791644692421,
          "train_metric": 0.000342181243468076,
          "validation_metric": 0.0003113939019385725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0333228039999085
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.003140099513984751,
          "train_loss": 0.002829902805387974,
          "validation_loss": 0.0023846884723752737,
          "train_metric": 0.00028501744964160025,
          "validation_metric": 0.0002401771052973345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.123680622999018
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0026708814766607247,
          "train_loss": 0.0031014317646622658,
          "validation_loss": 0.003110106335952878,
          "train_metric": 0.0003123648348264396,
          "validation_metric": 0.0003132385027129203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1360485209988838
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.005460345055325888,
          "train_loss": 0.01333299744874239,
          "validation_loss": 0.01045945379883051,
          "train_metric": 0.0013428505044430494,
          "validation_metric": 0.0010534379398450255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1479805119997764
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.009052199573488906,
          "train_loss": 0.0063875457271933556,
          "validation_loss": 0.006198588293045759,
          "train_metric": 0.0006433302187360823,
          "validation_metric": 0.0006242991657927632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1598331319983117
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.004626363392162602,
          "train_loss": 0.0075950613245368,
          "validation_loss": 0.0046640420332551,
          "train_metric": 0.0007649468025192618,
          "validation_metric": 0.00046974525321274996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.171899326998755
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.003685606825456489,
          "train_loss": 0.0029441642109304667,
          "validation_loss": 0.0024685771204531193,
          "train_metric": 0.0002965253952424973,
          "validation_metric": 0.0002486260491423309,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.18376279199947
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.002916460383858066,
          "train_loss": 0.005076117813587189,
          "validation_loss": 0.0033803831320255995,
          "train_metric": 0.0005112479557283223,
          "validation_metric": 0.00034045983920805156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1958934599988424
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.005792865480543696,
          "train_loss": 0.0033669525291770697,
          "validation_loss": 0.003035246394574642,
          "train_metric": 0.0003391071513760835,
          "validation_metric": 0.00030569895170629025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2078573049984698
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0034446123318048194,
          "train_loss": 0.003191647818312049,
          "validation_loss": 0.002429027808830142,
          "train_metric": 0.00032145107979886234,
          "validation_metric": 0.0002446427824907005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2323251179986983
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0035875356261385605,
          "train_loss": 0.0026020770892500877,
          "validation_loss": 0.0020087999291718006,
          "train_metric": 0.0002620716695673764,
          "validation_metric": 0.0002023190027102828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.268019489998551
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0028663186349149328,
          "train_loss": 0.0025293934158980846,
          "validation_loss": 0.0019956915639340878,
          "train_metric": 0.00025475124130025506,
          "validation_metric": 0.00020099875109735876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2953770989988698
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0028502725435828324,
          "train_loss": 0.0027129235677421093,
          "validation_loss": 0.002031296957284212,
          "train_metric": 0.0002732357825152576,
          "validation_metric": 0.0002045847795670852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3234271689998423
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0042524987875367515,
          "train_loss": 0.003872842527925968,
          "validation_loss": 0.0037787158507853746,
          "train_metric": 0.00039005849976092577,
          "validation_metric": 0.00038057842175476253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.346989843999836
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.003962184899137355,
          "train_loss": 0.004055355675518513,
          "validation_loss": 0.0029193286318331957,
          "train_metric": 0.00040844056638889015,
          "validation_metric": 0.00029402406653389335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3708805779988325
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.003953211162297521,
          "train_loss": 0.0025397089775651693,
          "validation_loss": 0.0022614081390202045,
          "train_metric": 0.00025579021894373,
          "validation_metric": 0.0002277607418363914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.383233914999437
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.002719601947319461,
          "train_loss": 0.0026238360442221165,
          "validation_loss": 0.0019035885343328118,
          "train_metric": 0.0002642631297931075,
          "validation_metric": 0.00019172250176779926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3965265679998993
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0032076858951768372,
          "train_loss": 0.0027902680449187756,
          "validation_loss": 0.0019289092160761356,
          "train_metric": 0.00028102556825615466,
          "validation_metric": 0.00019427266670390964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.409029129999908
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0030867182431393303,
          "train_loss": 0.002391109010204673,
          "validation_loss": 0.002165204146876931,
          "train_metric": 0.0002408237342024222,
          "validation_metric": 0.0002180714363930747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.42131937099839
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0033158883488795254,
          "train_loss": 0.00237888447009027,
          "validation_loss": 0.0016955232713371515,
          "train_metric": 0.0002395925112068653,
          "validation_metric": 0.00017076691437978297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4341394809998746
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.002880521169572603,
          "train_loss": 0.003373422659933567,
          "validation_loss": 0.003332486841827631,
          "train_metric": 0.00033975878614000976,
          "validation_metric": 0.0003356358502060175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4500136299993756
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0027995321433991194,
          "train_loss": 0.0022942558862268925,
          "validation_loss": 0.001852122601121664,
          "train_metric": 0.00023106903245206922,
          "validation_metric": 0.00018653899314813316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4808435429986275
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0027970887131232303,
          "train_loss": 0.0025605058763176203,
          "validation_loss": 0.0017409902065992355,
          "train_metric": 0.0002578847634140402,
          "validation_metric": 0.00017534616927150637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5052091349989496
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.003096605360042304,
          "train_loss": 0.005839838180691004,
          "validation_loss": 0.005989121273159981,
          "train_metric": 0.0005881671095266938,
          "validation_metric": 0.0006032023229636252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.534713548999207
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.005634894463582896,
          "train_loss": 0.005951581988483667,
          "validation_loss": 0.0038655921816825867,
          "train_metric": 0.0005994215025566518,
          "validation_metric": 0.00038932825555093586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5553901029998087
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.003944108571886318,
          "train_loss": 0.002415234921500087,
          "validation_loss": 0.0016975969774648547,
          "train_metric": 0.00024325359845533967,
          "validation_metric": 0.00017097573436331004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5681424719987263
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0053483923693420365,
          "train_loss": 0.0035635398235172033,
          "validation_loss": 0.002356210956349969,
          "train_metric": 0.00035890669096261263,
          "validation_metric": 0.00023730892280582339,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5807074469994404
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.004016427541500889,
          "train_loss": 0.003468751674517989,
          "validation_loss": 0.003071770304813981,
          "train_metric": 0.000349359936080873,
          "validation_metric": 0.0003093774721492082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.592325550998794
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0028282462735660374,
          "train_loss": 0.0024225222878158092,
          "validation_loss": 0.0022868881933391094,
          "train_metric": 0.00024398756795562804,
          "validation_metric": 0.00023032700119074434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.604437815998608
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.002529031626181677,
          "train_loss": 0.002386258216574788,
          "validation_loss": 0.001739665400236845,
          "train_metric": 0.00024033518275246024,
          "validation_metric": 0.00017521271365694702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6169770099986636
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.00394903509732103,
          "train_loss": 0.002652904251590371,
          "validation_loss": 0.0024135904386639595,
          "train_metric": 0.00026719082961790264,
          "validation_metric": 0.0002430879685562104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.629645293998692
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0029760080215055496,
          "train_loss": 0.002289547584950924,
          "validation_loss": 0.0017027707071974874,
          "train_metric": 0.00023059482919052243,
          "validation_metric": 0.00017149682389572263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6420703969997703
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0033333667925035115,
          "train_loss": 0.002790773520246148,
          "validation_loss": 0.001941721304319799,
          "train_metric": 0.00028107649995945394,
          "validation_metric": 0.00019556304323486984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6542549509995297
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.003451092583418358,
          "train_loss": 0.0027972699608653784,
          "validation_loss": 0.002302397508174181,
          "train_metric": 0.00028173081227578223,
          "validation_metric": 0.00023188904742710292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6667022919991723
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.006145377796201501,
          "train_loss": 0.005973841995000839,
          "validation_loss": 0.004446700215339661,
          "train_metric": 0.0006016636034473777,
          "validation_metric": 0.0004478553310036659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6797664589994383
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0033833492270787247,
          "train_loss": 0.002412480302155018,
          "validation_loss": 0.0020494197960942984,
          "train_metric": 0.0002429761952953413,
          "validation_metric": 0.00020641006994992495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6922149099991657
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0032946736937446985,
          "train_loss": 0.0021842254791408777,
          "validation_loss": 0.0018576375441625714,
          "train_metric": 0.00021998718148097396,
          "validation_metric": 0.00018709445430431515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.704493270999592
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0022994441533228382,
          "train_loss": 0.0027969246730208397,
          "validation_loss": 0.002489302773028612,
          "train_metric": 0.000281696004094556,
          "validation_metric": 0.0002507134631741792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.812827801999447
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002320041909115389,
          "train_loss": 0.0024129333905875683,
          "validation_loss": 0.002209965605288744,
          "train_metric": 0.00024302180099766701,
          "validation_metric": 0.0002225796488346532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.827236458999323
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0023475970738218166,
          "train_loss": 0.002201341325417161,
          "validation_loss": 0.0015944390324875712,
          "train_metric": 0.0002217110595665872,
          "validation_metric": 0.0001605860743438825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8417631199990865
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0027625904913293198,
          "train_loss": 0.0026560709811747074,
          "validation_loss": 0.0026658019050955772,
          "train_metric": 0.00026750972028821707,
          "validation_metric": 0.0002684897917788476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8547586139993655
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0027907500334549695,
          "train_loss": 0.0019415380666032434,
          "validation_loss": 0.0015534512931481004,
          "train_metric": 0.0001955446059582755,
          "validation_metric": 0.00015645792882423848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.867755060999116
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0024239311169367284,
          "train_loss": 0.0024284161627292633,
          "validation_loss": 0.002476495923474431,
          "train_metric": 0.0002445811696816236,
          "validation_metric": 0.00024942358140833676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8810194249999768
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.002325865578313824,
          "train_loss": 0.001956499880179763,
          "validation_loss": 0.0015367130981758237,
          "train_metric": 0.0001970514713320881,
          "validation_metric": 0.00015477211854886264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8958361389995844
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0028790376818506047,
          "train_loss": 0.002869058633223176,
          "validation_loss": 0.0029741995967924595,
          "train_metric": 0.0002889610768761486,
          "validation_metric": 0.00029955050558783114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.908132801998363
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.00255901031778194,
          "train_loss": 0.0024907321203500032,
          "validation_loss": 0.0017521464033052325,
          "train_metric": 0.0002508574107196182,
          "validation_metric": 0.00017646976630203426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9209349069988093
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.00285765016451478,
          "train_loss": 0.0024944462347775698,
          "validation_loss": 0.0018187678651884198,
          "train_metric": 0.00025123151135630906,
          "validation_metric": 0.00018317965441383421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.934508908998396
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.003064761323912535,
          "train_loss": 0.005674334242939949,
          "validation_loss": 0.006037655286490917,
          "train_metric": 0.0005714981816709042,
          "validation_metric": 0.0006080905441194773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9483641859987983
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.003508273199258838,
          "train_loss": 0.004737573210150003,
          "validation_loss": 0.003270583925768733,
          "train_metric": 0.0004771510139107704,
          "validation_metric": 0.00032940125674940646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.962335329999405
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.003049729777558241,
          "train_loss": 0.002487736754119396,
          "validation_loss": 0.0022855119314044714,
          "train_metric": 0.00025055574951693416,
          "validation_metric": 0.00023018837964627892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.974483819998568
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0023796298082743306,
          "train_loss": 0.001998920226469636,
          "validation_loss": 0.0018319382797926664,
          "train_metric": 0.00020132391364313662,
          "validation_metric": 0.0001845060905907303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.986870310998711
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0024417492240900174,
          "train_loss": 0.005391971208155155,
          "validation_loss": 0.004610136616975069,
          "train_metric": 0.0005430596647784114,
          "validation_metric": 0.000464316108264029,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.998788702998354
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.005641180418024305,
          "train_loss": 0.003403585171326995,
          "validation_loss": 0.0029747069347649813,
          "train_metric": 0.0003427966439630836,
          "validation_metric": 0.0002996015828102827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.081566902999839
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.004231832575896988,
          "train_loss": 0.0035790707916021347,
          "validation_loss": 0.0024940939620137215,
          "train_metric": 0.00036047090543434024,
          "validation_metric": 0.0002511959755793214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.214778847999696
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.004203739263175521,
          "train_loss": 0.003965888172388077,
          "validation_loss": 0.0038141377735882998,
          "train_metric": 0.0003994297294411808,
          "validation_metric": 0.00038414596929214895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.309776060999866
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.002577638872026,
          "train_loss": 0.0029582171700894833,
          "validation_loss": 0.002494297455996275,
          "train_metric": 0.0002979407727252692,
          "validation_metric": 0.00025121652288362384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3298095259997353
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.002700876437302213,
          "train_loss": 0.0020670127123594284,
          "validation_loss": 0.001548741478472948,
          "train_metric": 0.00020818194025196135,
          "validation_metric": 0.0001559835800435394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.349501128999691
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.002641384104208555,
          "train_loss": 0.0022863061167299747,
          "validation_loss": 0.0016294368542730808,
          "train_metric": 0.00023026838607620448,
          "validation_metric": 0.00016411091201007366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3732272469987947
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0020552628811856266,
          "train_loss": 0.002190658589825034,
          "validation_loss": 0.0022713581565767527,
          "train_metric": 0.00022063512005843222,
          "validation_metric": 0.00022876290313433856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.402328394999131
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0017349017380183795,
          "train_loss": 0.002510425169020891,
          "validation_loss": 0.0016018972964957356,
          "train_metric": 0.0002528408367652446,
          "validation_metric": 0.00016133721510414034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.421163813998646
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0022354315406118985,
          "train_loss": 0.0019184598932042718,
          "validation_loss": 0.0017449004808440804,
          "train_metric": 0.00019322028674650937,
          "validation_metric": 0.00017573997320141643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4395006919985462
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0022512596478918567,
          "train_loss": 0.0017983666621148586,
          "validation_loss": 0.001620680559426546,
          "train_metric": 0.00018112490943167359,
          "validation_metric": 0.00016322900773957372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4632968039986736
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0021359733345889254,
          "train_loss": 0.002333016600459814,
          "validation_loss": 0.0016282597789540887,
          "train_metric": 0.00023497290385421365,
          "validation_metric": 0.00016399238666053861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4985096709988284
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.002858362793631386,
          "train_loss": 0.0025403928011655807,
          "validation_loss": 0.002290292177349329,
          "train_metric": 0.00025585907860659063,
          "validation_metric": 0.00023066984431352466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5257173769987276
        }
      ],
      "best": {
        "epoch": 186,
        "updates": 2976
      },
      "best_validation_metric": 0.00015477211854886264,
      "config": {
        "label": "Concatenated skip",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": true,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2243,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012440744292399174
    },
    "5bc849f25b25_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.41845904290676117,
          "train_loss": 0.21851415932178497,
          "validation_loss": 0.20473019778728485,
          "train_metric": 0.02200794778764248,
          "validation_metric": 0.020619675517082214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01297916899966367
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.1676428052596748,
          "train_loss": 0.1442738175392151,
          "validation_loss": 0.1083851158618927,
          "train_metric": 0.014530733227729797,
          "validation_metric": 0.010916152969002724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02550512199923105
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.1271183155477047,
          "train_loss": 0.11397996544837952,
          "validation_loss": 0.09012553840875626,
          "train_metric": 0.011479645036160946,
          "validation_metric": 0.00907711498439312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03795292999893718
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.09986551781184971,
          "train_loss": 0.08589127659797668,
          "validation_loss": 0.07211899757385254,
          "train_metric": 0.00865065585821867,
          "validation_metric": 0.007263561710715294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.050384521999149
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.07878399966284633,
          "train_loss": 0.07695572078227997,
          "validation_loss": 0.0637139156460762,
          "train_metric": 0.007750697899609804,
          "validation_metric": 0.006417031399905682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06249396299972432
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06999824510421604,
          "train_loss": 0.06526762992143631,
          "validation_loss": 0.06084335595369339,
          "train_metric": 0.006573516875505447,
          "validation_metric": 0.006127920001745224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07458778500040353
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06485200382303447,
          "train_loss": 0.05935068055987358,
          "validation_loss": 0.049333974719047546,
          "train_metric": 0.005977582652121782,
          "validation_metric": 0.004968736786395311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08691791199998988
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.05878081580158323,
          "train_loss": 0.0555548258125782,
          "validation_loss": 0.04710673913359642,
          "train_metric": 0.0055952779948711395,
          "validation_metric": 0.0047444188967347145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0994384909990913
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.05506116268225014,
          "train_loss": 0.0535321980714798,
          "validation_loss": 0.047991134226322174,
          "train_metric": 0.0053915660828351974,
          "validation_metric": 0.0048334915190935135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11156623199894966
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05404108914081007,
          "train_loss": 0.04821320250630379,
          "validation_loss": 0.04330015555024147,
          "train_metric": 0.004855857230722904,
          "validation_metric": 0.0043610334396362305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12410549299966078
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04961593635380268,
          "train_loss": 0.04657549038529396,
          "validation_loss": 0.04151332378387451,
          "train_metric": 0.004690913017839193,
          "validation_metric": 0.004181069787591696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13636816200050816
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04924166726414114,
          "train_loss": 0.04423417150974274,
          "validation_loss": 0.039078958332538605,
          "train_metric": 0.004455103538930416,
          "validation_metric": 0.003935889806598425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1487224079992302
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.04533920937683433,
          "train_loss": 0.04269801080226898,
          "validation_loss": 0.03462360426783562,
          "train_metric": 0.00430038757622242,
          "validation_metric": 0.003487162757664919,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1613318210002035
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.04167148354463279,
          "train_loss": 0.0352061502635479,
          "validation_loss": 0.029675012454390526,
          "train_metric": 0.0035458344500511885,
          "validation_metric": 0.0029887589626014233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1736963570001535
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03770395915489644,
          "train_loss": 0.03286425396800041,
          "validation_loss": 0.02688632532954216,
          "train_metric": 0.003309967229142785,
          "validation_metric": 0.002707892330363393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1860949490001076
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.032078508636914194,
          "train_loss": 0.03048131614923477,
          "validation_loss": 0.025485079735517502,
          "train_metric": 0.003069966332986951,
          "validation_metric": 0.0025667641311883926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19835356700059492
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03021253680344671,
          "train_loss": 0.028504734858870506,
          "validation_loss": 0.024020284414291382,
          "train_metric": 0.002870892407372594,
          "validation_metric": 0.0024192347191274166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2109869649993925
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.02774184598820284,
          "train_loss": 0.02671283856034279,
          "validation_loss": 0.022039813920855522,
          "train_metric": 0.0026904193218797445,
          "validation_metric": 0.002219769172370434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22345650299939734
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.027848627767525613,
          "train_loss": 0.025834523141384125,
          "validation_loss": 0.0215616412460804,
          "train_metric": 0.0026019588112831116,
          "validation_metric": 0.0021716090850532055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2358979589989758
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.0350184211274609,
          "train_loss": 0.03225558251142502,
          "validation_loss": 0.024449199438095093,
          "train_metric": 0.003248664317652583,
          "validation_metric": 0.002462433883920312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24802063400056795
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.03112356533529237,
          "train_loss": 0.022610845044255257,
          "validation_loss": 0.021048566326498985,
          "train_metric": 0.002277281368151307,
          "validation_metric": 0.0021199346520006657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2604972770004679
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.025856521155219525,
          "train_loss": 0.026957785710692406,
          "validation_loss": 0.022114522755146027,
          "train_metric": 0.0027150893583893776,
          "validation_metric": 0.0022272937931120396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27262256299945875
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.02578889933647588,
          "train_loss": 0.018482059240341187,
          "validation_loss": 0.015332325361669064,
          "train_metric": 0.001861445140093565,
          "validation_metric": 0.0015442156000062823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2852406210004119
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.020806329644983634,
          "train_loss": 0.01968742161989212,
          "validation_loss": 0.014386219903826714,
          "train_metric": 0.0019828451331704855,
          "validation_metric": 0.0014489273307844996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2977585789994919
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.021072010247735307,
          "train_loss": 0.0196637324988842,
          "validation_loss": 0.01980803720653057,
          "train_metric": 0.0019804590847343206,
          "validation_metric": 0.0019949928391724825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30981821900059003
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.017936955904588103,
          "train_loss": 0.017420340329408646,
          "validation_loss": 0.01280991081148386,
          "train_metric": 0.0017545126611366868,
          "validation_metric": 0.0012901673326268792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3226371479995578
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015697771799750626,
          "train_loss": 0.01741812564432621,
          "validation_loss": 0.016201676800847054,
          "train_metric": 0.0017542896093800664,
          "validation_metric": 0.0016317734261974692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33456504000059795
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.017510911042336375,
          "train_loss": 0.014142591506242752,
          "validation_loss": 0.011252922005951405,
          "train_metric": 0.0014243900077417493,
          "validation_metric": 0.0011333531001582742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3472985200005496
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.01734317565569654,
          "train_loss": 0.020783117040991783,
          "validation_loss": 0.017709095031023026,
          "train_metric": 0.0020931994076818228,
          "validation_metric": 0.0017835950711742043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3595932059997722
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.017934594419784844,
          "train_loss": 0.020063497126102448,
          "validation_loss": 0.015239039435982704,
          "train_metric": 0.002020722022280097,
          "validation_metric": 0.0015348203014582396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3718684329996904
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.013932761736214161,
          "train_loss": 0.01618463732302189,
          "validation_loss": 0.014340869151055813,
          "train_metric": 0.001630057580769062,
          "validation_metric": 0.0014443598920479417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3837836509992485
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.018347828387049958,
          "train_loss": 0.015032259747385979,
          "validation_loss": 0.011295442469418049,
          "train_metric": 0.0015139938332140446,
          "validation_metric": 0.0011376356706023216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39578176099894335
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.016564108576858416,
          "train_loss": 0.013337207026779652,
          "validation_loss": 0.011447574011981487,
          "train_metric": 0.0013432747218757868,
          "validation_metric": 0.0011529577895998955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.407528605999687
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01198665820993483,
          "train_loss": 0.014576696790754795,
          "validation_loss": 0.012972936034202576,
          "train_metric": 0.0014681116444990039,
          "validation_metric": 0.001306586666032672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4195940730005532
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.013125977464369498,
          "train_loss": 0.010193406604230404,
          "validation_loss": 0.008512240834534168,
          "train_metric": 0.0010266425088047981,
          "validation_metric": 0.0008573217201046646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4320338659999834
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.011527594688232057,
          "train_loss": 0.009771285578608513,
          "validation_loss": 0.00789465568959713,
          "train_metric": 0.000984127982519567,
          "validation_metric": 0.0007951207808218896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4450057609992655
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.011893673072336242,
          "train_loss": 0.012662642635405064,
          "validation_loss": 0.011704102158546448,
          "train_metric": 0.001275335089303553,
          "validation_metric": 0.0011787941912189126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45692011199935223
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.01044908793119248,
          "train_loss": 0.009098139591515064,
          "validation_loss": 0.007869894616305828,
          "train_metric": 0.0009163313079625368,
          "validation_metric": 0.0007926268735900521,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4693962509991252
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.009184351773001254,
          "train_loss": 0.009985133074223995,
          "validation_loss": 0.007681173738092184,
          "train_metric": 0.0010056659812107682,
          "validation_metric": 0.0007736196857877076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48199377699893375
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.010291059108567424,
          "train_loss": 0.008394632488489151,
          "validation_loss": 0.00731263030320406,
          "train_metric": 0.0008454765775240958,
          "validation_metric": 0.0007365013589151204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49480671799938136
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010177273696172051,
          "train_loss": 0.010530607774853706,
          "validation_loss": 0.00973808579146862,
          "train_metric": 0.0010606043506413698,
          "validation_metric": 0.0009807843016460538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5068599329988501
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.009132841019891202,
          "train_loss": 0.007123306859284639,
          "validation_loss": 0.005853293463587761,
          "train_metric": 0.0007174334023147821,
          "validation_metric": 0.0005895223002880812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5193425050001679
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.010474994371179491,
          "train_loss": 0.01721721701323986,
          "validation_loss": 0.01308069285005331,
          "train_metric": 0.0017340549966320395,
          "validation_metric": 0.001317439367994666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.530921555000532
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.009325740160420537,
          "train_loss": 0.0074717802926898,
          "validation_loss": 0.006883339025080204,
          "train_metric": 0.0007525302935391665,
          "validation_metric": 0.0006932646501809359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5425207230000524
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.0067742622923105955,
          "train_loss": 0.006942267995327711,
          "validation_loss": 0.0051760622300207615,
          "train_metric": 0.0006991998525336385,
          "validation_metric": 0.0005213140393607318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5558080959999643
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.006978391291340813,
          "train_loss": 0.0067337495274841785,
          "validation_loss": 0.0052259876392781734,
          "train_metric": 0.0006781985866837204,
          "validation_metric": 0.0005263423081487417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5684415319992695
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.007889736829383764,
          "train_loss": 0.011199242435395718,
          "validation_loss": 0.008247263729572296,
          "train_metric": 0.0011279468890279531,
          "validation_metric": 0.0008306340896524489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5805758469996363
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.008029950855416246,
          "train_loss": 0.00894999410957098,
          "validation_loss": 0.0066956630907952785,
          "train_metric": 0.0009014107054099441,
          "validation_metric": 0.0006743627018295228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5924440910002886
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00819560862146318,
          "train_loss": 0.007321988232433796,
          "validation_loss": 0.0061864545568823814,
          "train_metric": 0.0007374438573606312,
          "validation_metric": 0.0006230769795365632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6040887069993914
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.008165046412614174,
          "train_loss": 0.005631379783153534,
          "validation_loss": 0.0037961399648338556,
          "train_metric": 0.0005671719554811716,
          "validation_metric": 0.00038233332452364266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.616167709000365
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007459649466909468,
          "train_loss": 0.009484965354204178,
          "validation_loss": 0.009746874682605267,
          "train_metric": 0.0009552909759804606,
          "validation_metric": 0.0009816695237532258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6281065549992491
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.012182177626527846,
          "train_loss": 0.008039084263145924,
          "validation_loss": 0.006014736834913492,
          "train_metric": 0.0008096671663224697,
          "validation_metric": 0.0006057822029106319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6401041459994303
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.007669617742067203,
          "train_loss": 0.007949156686663628,
          "validation_loss": 0.005213141907006502,
          "train_metric": 0.0008006099378690124,
          "validation_metric": 0.0005250485264696181,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.652172485999472
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.007302980244276114,
          "train_loss": 0.006993729621171951,
          "validation_loss": 0.005370179191231728,
          "train_metric": 0.0007043828372843564,
          "validation_metric": 0.0005408647120930254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6641040889990109
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.008774790214374661,
          "train_loss": 0.008321143686771393,
          "validation_loss": 0.00849241018295288,
          "train_metric": 0.0008380751241929829,
          "validation_metric": 0.0008553244406357408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6759353979996376
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0091457652306417,
          "train_loss": 0.010506180115044117,
          "validation_loss": 0.00792794581502676,
          "train_metric": 0.0010581441456452012,
          "validation_metric": 0.0007984737167134881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6880674669992004
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.007904297963250428,
          "train_loss": 0.004490549210458994,
          "validation_loss": 0.0035864971578121185,
          "train_metric": 0.00045227163354866207,
          "validation_metric": 0.00036121884477324784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7005665039996529
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.01053971435612766,
          "train_loss": 0.016787733882665634,
          "validation_loss": 0.016029685735702515,
          "train_metric": 0.0016907991375774145,
          "validation_metric": 0.0016144512919709086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7127360270005738
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.011966160644078627,
          "train_loss": 0.008228184655308723,
          "validation_loss": 0.0050873723812401295,
          "train_metric": 0.0008287126547656953,
          "validation_metric": 0.0005123814917169511,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7245914199993422
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006808063481003046,
          "train_loss": 0.004356035962700844,
          "validation_loss": 0.0030484579037874937,
          "train_metric": 0.00043872397509403527,
          "validation_metric": 0.00030702954973094165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7372933520000515
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004806755023309961,
          "train_loss": 0.004795592278242111,
          "validation_loss": 0.004299277905374765,
          "train_metric": 0.0004829944809898734,
          "validation_metric": 0.00043300751713104546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7503058139991481
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.004799488182470668,
          "train_loss": 0.005996539257466793,
          "validation_loss": 0.006070504896342754,
          "train_metric": 0.0006039494764991105,
          "validation_metric": 0.0006113990093581378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.845473649998894
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.007196154052508064,
          "train_loss": 0.01009567640721798,
          "validation_loss": 0.0078029632568359375,
          "train_metric": 0.0010167995933443308,
          "validation_metric": 0.0007858857861720026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.876367365999613
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007902053854195401,
          "train_loss": 0.0038587688468396664,
          "validation_loss": 0.002851600991562009,
          "train_metric": 0.00038864105590619147,
          "validation_metric": 0.00028720279806293547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9474516119989858
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.005327188533556182,
          "train_loss": 0.008495751768350601,
          "validation_loss": 0.009140483103692532,
          "train_metric": 0.0008556609391234815,
          "validation_metric": 0.0009205960086546838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9738390980000986
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.00560941151343286,
          "train_loss": 0.003661559894680977,
          "validation_loss": 0.002629627473652363,
          "train_metric": 0.0003687789139803499,
          "validation_metric": 0.0002648464578669518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9862350949988468
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.004790859886270482,
          "train_loss": 0.005243054125458002,
          "validation_loss": 0.00328547740355134,
          "train_metric": 0.0005280611803755164,
          "validation_metric": 0.0003309012681711465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9980017580001004
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0044937270795344375,
          "train_loss": 0.003521266859024763,
          "validation_loss": 0.0027356192003935575,
          "train_metric": 0.00035464909160509706,
          "validation_metric": 0.00027552159735932946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0095706339998287
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.006274908999330364,
          "train_loss": 0.005787645932286978,
          "validation_loss": 0.004313301295042038,
          "train_metric": 0.0005829104338772595,
          "validation_metric": 0.0004344198969192803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0211009499998909
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.006288113749178592,
          "train_loss": 0.0059364186599850655,
          "validation_loss": 0.005382617935538292,
          "train_metric": 0.0005978943663649261,
          "validation_metric": 0.0005421175737865269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0329746219995286
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.004155754912062548,
          "train_loss": 0.003410942852497101,
          "validation_loss": 0.002084902720525861,
          "train_metric": 0.000343537685694173,
          "validation_metric": 0.00020998375839553773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.045215857000585
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.004390210571727948,
          "train_loss": 0.004576707258820534,
          "validation_loss": 0.004220203496515751,
          "train_metric": 0.00046094920253381133,
          "validation_metric": 0.0004250434576533735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0568694969988428
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.004692218702984974,
          "train_loss": 0.0057886624708771706,
          "validation_loss": 0.0059942216612398624,
          "train_metric": 0.0005830129375681281,
          "validation_metric": 0.0006037160055711865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0720473519995721
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005631028252537362,
          "train_loss": 0.00707528879866004,
          "validation_loss": 0.004802821669727564,
          "train_metric": 0.0007125972188077867,
          "validation_metric": 0.00048372262972407043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0841016710000986
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00724928651470691,
          "train_loss": 0.0036445034202188253,
          "validation_loss": 0.0019371696980670094,
          "train_metric": 0.00036706108949147165,
          "validation_metric": 0.00019510462880134583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0964924149993749
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.00585300926832133,
          "train_loss": 0.0034800362773239613,
          "validation_loss": 0.0022447523660957813,
          "train_metric": 0.00035049652797169983,
          "validation_metric": 0.00022608324070461094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1090325000004668
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.005182861357752699,
          "train_loss": 0.004276298452168703,
          "validation_loss": 0.0038790320977568626,
          "train_metric": 0.00043069315142929554,
          "validation_metric": 0.00039068193291313946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1209440160000668
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.0037918172602076083,
          "train_loss": 0.0035881339572370052,
          "validation_loss": 0.002499943831935525,
          "train_metric": 0.0003613836597651243,
          "validation_metric": 0.0002517852117307484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1328543899999204
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.004138135293032974,
          "train_loss": 0.0033953338861465454,
          "validation_loss": 0.0025824028998613358,
          "train_metric": 0.0003419655840843916,
          "validation_metric": 0.00026009013527072966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1448338419995707
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004216473294945899,
          "train_loss": 0.002593801822513342,
          "validation_loss": 0.001671645324677229,
          "train_metric": 0.0002612382231745869,
          "validation_metric": 0.0001683619775576517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1572935089989187
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0035842264187522233,
          "train_loss": 0.005226960871368647,
          "validation_loss": 0.0041853152215480804,
          "train_metric": 0.00052644032984972,
          "validation_metric": 0.0004215295775793493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1693188350000128
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0061604706206708215,
          "train_loss": 0.006207527592778206,
          "validation_loss": 0.006199733354151249,
          "train_metric": 0.0006251994636841118,
          "validation_metric": 0.0006244144169613719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1811806860005163
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.007461132539901882,
          "train_loss": 0.0033099078573286533,
          "validation_loss": 0.0023432455491274595,
          "train_metric": 0.00033336179330945015,
          "validation_metric": 0.00023600310669280589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1929649529993185
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.00531732623494463,
          "train_loss": 0.0037077402230352163,
          "validation_loss": 0.003007907420396805,
          "train_metric": 0.00037343002622947097,
          "validation_metric": 0.0003029454092029482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2046960709994892
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.004911517891741823,
          "train_loss": 0.004097212105989456,
          "validation_loss": 0.0024468335323035717,
          "train_metric": 0.00041265616891905665,
          "validation_metric": 0.00024643613141961396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2165523969997594
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.003902873875631485,
          "train_loss": 0.0037494467105716467,
          "validation_loss": 0.0022370389197021723,
          "train_metric": 0.0003776305529754609,
          "validation_metric": 0.0002253063430543989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2282957329989586
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0037722915076301433,
          "train_loss": 0.0032451865263283253,
          "validation_loss": 0.0032056181225925684,
          "train_metric": 0.0003268432919867337,
          "validation_metric": 0.0003228581335861236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.239852927999891
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0039532116570626386,
          "train_loss": 0.00263984058983624,
          "validation_loss": 0.0014149402268230915,
          "train_metric": 0.00026587507454678416,
          "validation_metric": 0.0001425075897714123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.251942880999195
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.002851794648449868,
          "train_loss": 0.002350043971091509,
          "validation_loss": 0.001725166104733944,
          "train_metric": 0.000236687803408131,
          "validation_metric": 0.00017375241441186517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2637770849996741
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0027123029431095347,
          "train_loss": 0.0026279918383806944,
          "validation_loss": 0.0024731082376092672,
          "train_metric": 0.00026468170108273625,
          "validation_metric": 0.00024908239720389247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2755290570003126
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0028794847457902506,
          "train_loss": 0.0026803314685821533,
          "validation_loss": 0.0017537848325446248,
          "train_metric": 0.00026995319058187306,
          "validation_metric": 0.00017663479957263917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2870843099990452
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0022594907713937573,
          "train_loss": 0.003024031175300479,
          "validation_loss": 0.0021693366579711437,
          "train_metric": 0.0003045693156309426,
          "validation_metric": 0.0002184876357205212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2987213639989932
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0028446778815123253,
          "train_loss": 0.002624227898195386,
          "validation_loss": 0.0025788191705942154,
          "train_metric": 0.00026430265279486775,
          "validation_metric": 0.00025972924777306616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3104351389993099
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0034149009370594285,
          "train_loss": 0.004467193502932787,
          "validation_loss": 0.003138265572488308,
          "train_metric": 0.00044991940376348794,
          "validation_metric": 0.00031607464188709855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.322306910999032
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0029386202368186787,
          "train_loss": 0.0021464249584823847,
          "validation_loss": 0.0012431982904672623,
          "train_metric": 0.00021618003665935248,
          "validation_metric": 0.0001252103829756379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.334899540999686
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.003615044173784554,
          "train_loss": 0.005185762420296669,
          "validation_loss": 0.005212593823671341,
          "train_metric": 0.0005222909967415035,
          "validation_metric": 0.000524993403814733,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3469903169989266
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005184923589695245,
          "train_loss": 0.0032751087564975023,
          "validation_loss": 0.001826393068768084,
          "train_metric": 0.0003298569063190371,
          "validation_metric": 0.00018394761718809605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3587301489988022
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004303774199797772,
          "train_loss": 0.002455071546137333,
          "validation_loss": 0.0014496884541586041,
          "train_metric": 0.0002472658234182745,
          "validation_metric": 0.0001460073108319193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3706104579996463
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0032009280621423386,
          "train_loss": 0.0030889976769685745,
          "validation_loss": 0.002774082589894533,
          "train_metric": 0.0003111125552095473,
          "validation_metric": 0.00027939543360844254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3825543359998846
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0033595101958781015,
          "train_loss": 0.0029812166467309,
          "validation_loss": 0.0023152881767600775,
          "train_metric": 0.000300257233902812,
          "validation_metric": 0.00023318734019994736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3945829419990332
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.002320140596566489,
          "train_loss": 0.0018481028964743018,
          "validation_loss": 0.0010820503812283278,
          "train_metric": 0.00018613414431456476,
          "validation_metric": 0.00010898015898419544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4068386279996048
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.002527380216633901,
          "train_loss": 0.004209944512695074,
          "validation_loss": 0.004110608249902725,
          "train_metric": 0.0004240102134644985,
          "validation_metric": 0.00041400542249903083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4185884399994393
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0030730681683053263,
          "train_loss": 0.002078052842989564,
          "validation_loss": 0.0018968003569170833,
          "train_metric": 0.00020929389575030655,
          "validation_metric": 0.00019103878003079444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4303871379997872
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.002187505531765055,
          "train_loss": 0.0025526743847876787,
          "validation_loss": 0.0013351710513234138,
          "train_metric": 0.0002570959914010018,
          "validation_metric": 0.00013447355013340712,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.442134487999283
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.004240257207129616,
          "train_loss": 0.005715868901461363,
          "validation_loss": 0.005624698009341955,
          "train_metric": 0.0005756813916377723,
          "validation_metric": 0.000566499016713351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4536420330005058
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.005576602110522799,
          "train_loss": 0.008512349799275398,
          "validation_loss": 0.009294843301177025,
          "train_metric": 0.0008573327213525772,
          "validation_metric": 0.0009361426928080618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4655868299996655
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004364053522294853,
          "train_loss": 0.0022739493288099766,
          "validation_loss": 0.0014216684503480792,
          "train_metric": 0.00022902384807821363,
          "validation_metric": 0.00014318525791168213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4771890890006034
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.002021707878157031,
          "train_loss": 0.0020154849626123905,
          "validation_loss": 0.0010403741616755724,
          "train_metric": 0.00020299226162023842,
          "validation_metric": 0.00010478268086444587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.489176953000424
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.001993116547964746,
          "train_loss": 0.001786703011021018,
          "validation_loss": 0.0011510945623740554,
          "train_metric": 0.00017995020607486367,
          "validation_metric": 0.00011593403178267181,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5007586449992232
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.002797696964989882,
          "train_loss": 0.0029525053687393665,
          "validation_loss": 0.001914783613756299,
          "train_metric": 0.0002973655064124614,
          "validation_metric": 0.00019284998415969312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.512208391999593
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.002241526079160394,
          "train_loss": 0.001690941397100687,
          "validation_loss": 0.0011993085499852896,
          "train_metric": 0.00017030541494023055,
          "validation_metric": 0.00012078997679054737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5236999690005177
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.002246523341455031,
          "train_loss": 0.0028348537161946297,
          "validation_loss": 0.0020443385001271963,
          "train_metric": 0.000285516056464985,
          "validation_metric": 0.0002058982936432585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5352707950005424
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.002562055062298896,
          "train_loss": 0.0021932381205260754,
          "validation_loss": 0.0013424346689134836,
          "train_metric": 0.00022089490084908903,
          "validation_metric": 0.00013520510401576757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5467997020004987
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0017353311341139488,
          "train_loss": 0.0028307221364229918,
          "validation_loss": 0.0016564414836466312,
          "train_metric": 0.00028509992989711463,
          "validation_metric": 0.0001668307522777468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5580734579998534
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.003275410272181034,
          "train_loss": 0.0039473893120884895,
          "validation_loss": 0.0029942139517515898,
          "train_metric": 0.00039756661863066256,
          "validation_metric": 0.0003015663241967559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.570366281999668
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.004250979865901172,
          "train_loss": 0.0034918836317956448,
          "validation_loss": 0.002271275967359543,
          "train_metric": 0.00035168975591659546,
          "validation_metric": 0.00022875459399074316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.582274578000579
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003004139096447034,
          "train_loss": 0.002297036349773407,
          "validation_loss": 0.002474363660439849,
          "train_metric": 0.00023134908406063914,
          "validation_metric": 0.000249208853347227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5943258890001744
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0021435508388094604,
          "train_loss": 0.0017148456536233425,
          "validation_loss": 0.0009443670278415084,
          "train_metric": 0.00017271297110710293,
          "validation_metric": 9.511318785371259e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6069062289989233
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0016267797564069042,
          "train_loss": 0.0025180340744554996,
          "validation_loss": 0.0014884897973388433,
          "train_metric": 0.0002536071988288313,
          "validation_metric": 0.00014991524221841246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6190429910002422
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.001810647710954072,
          "train_loss": 0.002736328635364771,
          "validation_loss": 0.0017946050502359867,
          "train_metric": 0.0002755930181592703,
          "validation_metric": 0.00018074603576678783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6310955369990552
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0021341051360650454,
          "train_loss": 0.0014738220488652587,
          "validation_loss": 0.0010818217415362597,
          "train_metric": 0.00014843793178442866,
          "validation_metric": 0.00010895712330238894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6429525730000023
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0028309821464063134,
          "train_loss": 0.0036718433257192373,
          "validation_loss": 0.0035368145909160376,
          "train_metric": 0.00036981457378715277,
          "validation_metric": 0.0003562150232028216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6546224399990024
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0026914552472590003,
          "train_loss": 0.0026836558245122433,
          "validation_loss": 0.0016956690233200788,
          "train_metric": 0.00027028797194361687,
          "validation_metric": 0.0001707815536065027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6664716969989968
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.00266076807747595,
          "train_loss": 0.0021687692496925592,
          "validation_loss": 0.001156527898274362,
          "train_metric": 0.00021843051945324987,
          "validation_metric": 0.00011648125655483454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6781802539990167
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0020087515040358994,
          "train_loss": 0.002168552251532674,
          "validation_loss": 0.002001863205805421,
          "train_metric": 0.0002184086333727464,
          "validation_metric": 0.00020162033615633845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6902982329993392
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.004019592903205194,
          "train_loss": 0.004665459506213665,
          "validation_loss": 0.004182003904134035,
          "train_metric": 0.0004698879492934793,
          "validation_metric": 0.00042119616409763694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.701689591000104
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.002788319376122672,
          "train_loss": 0.004539731424301863,
          "validation_loss": 0.0035669885110110044,
          "train_metric": 0.0004572250763885677,
          "validation_metric": 0.00035925398697145283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7134180089997244
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.011576306183997076,
          "train_loss": 0.00882845837622881,
          "validation_loss": 0.008443949744105339,
          "train_metric": 0.0008891699253581464,
          "validation_metric": 0.0008504436118528247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.725170771998819
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.004691163878305815,
          "train_loss": 0.0019013509154319763,
          "validation_loss": 0.0011481735855340958,
          "train_metric": 0.00019149709260091186,
          "validation_metric": 0.00011563982116058469,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7370040400001017
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0028960897761862725,
          "train_loss": 0.0018767036963254213,
          "validation_loss": 0.001341586234048009,
          "train_metric": 0.00018901472503785044,
          "validation_metric": 0.0001351196551695466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7487259649988118
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.001469866907427786,
          "train_loss": 0.001704173511825502,
          "validation_loss": 0.0012667630799114704,
          "train_metric": 0.00017163810844067484,
          "validation_metric": 0.0001275837275898084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7605807250001817
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0018339827674935805,
          "train_loss": 0.0034625360276550055,
          "validation_loss": 0.0027892235666513443,
          "train_metric": 0.00034873391268774867,
          "validation_metric": 0.000280920387012884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7724584440002218
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.003211956143786665,
          "train_loss": 0.0026401642244309187,
          "validation_loss": 0.0018986702198162675,
          "train_metric": 0.00026590769994072616,
          "validation_metric": 0.00019122709636576474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7853359609998734
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0023504793098254595,
          "train_loss": 0.0015385554870590568,
          "validation_loss": 0.0010448372922837734,
          "train_metric": 0.00015495768457185477,
          "validation_metric": 0.0001052321822498925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7971430239995243
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0018201615002908511,
          "train_loss": 0.001775587908923626,
          "validation_loss": 0.001289681764319539,
          "train_metric": 0.00017883071268443018,
          "validation_metric": 0.00012989201059099287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8088158360005764
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.002526402229705127,
          "train_loss": 0.0029636439867317677,
          "validation_loss": 0.0030041022691875696,
          "train_metric": 0.0002984873717650771,
          "validation_metric": 0.00030256216996349394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8205274590000045
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.004234274892951362,
          "train_loss": 0.0055715544149279594,
          "validation_loss": 0.00441470742225647,
          "train_metric": 0.0005611465312540531,
          "validation_metric": 0.00044463324593380094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8319323510004324
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.004399651936182636,
          "train_loss": 0.001979161286726594,
          "validation_loss": 0.0014911037869751453,
          "train_metric": 0.00019933388102799654,
          "validation_metric": 0.00015017850091680884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.843842661999588
        }
      ],
      "best": {
        "epoch": 118,
        "updates": 1888
      },
      "best_validation_metric": 9.511318785371259e-05,
      "config": {
        "label": "Tanh",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009752599030705229
    },
    "d7d20cfbbc18_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.6851238450035453,
          "train_loss": 0.24134469032287598,
          "validation_loss": 0.1561749279499054,
          "train_metric": 0.024307353422045708,
          "validation_metric": 0.01572936773300171,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.013500346000000718
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.2109692096710205,
          "train_loss": 0.15131829679012299,
          "validation_loss": 0.12313321232795715,
          "train_metric": 0.015240224078297615,
          "validation_metric": 0.012401525862514973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02605107199997292
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13556771632283926,
          "train_loss": 0.11771136522293091,
          "validation_loss": 0.10192956030368805,
          "train_metric": 0.011855458840727806,
          "validation_metric": 0.010265971533954144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03883771000073466
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.11162484949454665,
          "train_loss": 0.10019880533218384,
          "validation_loss": 0.08469738066196442,
          "train_metric": 0.010091655887663364,
          "validation_metric": 0.008530410937964916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.052005195000674576
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.09353748336434364,
          "train_loss": 0.08390111476182938,
          "validation_loss": 0.06853823363780975,
          "train_metric": 0.008450212888419628,
          "validation_metric": 0.006902919616550207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06531413800075825
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.07856978673953563,
          "train_loss": 0.07078257948160172,
          "validation_loss": 0.05842211842536926,
          "train_metric": 0.007128962315618992,
          "validation_metric": 0.0058840615674853325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07813752400033991
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06878085504285991,
          "train_loss": 0.06412050127983093,
          "validation_loss": 0.052704308182001114,
          "train_metric": 0.0064579821191728115,
          "validation_metric": 0.005308184772729874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09121770900128467
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06267210748046637,
          "train_loss": 0.060872018337249756,
          "validation_loss": 0.049474086612463,
          "train_metric": 0.006130806170403957,
          "validation_metric": 0.004982849583029747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10426252000070235
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.06081061786971986,
          "train_loss": 0.05766541138291359,
          "validation_loss": 0.04891744256019592,
          "train_metric": 0.005807848647236824,
          "validation_metric": 0.004926786292344332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11732906100041873
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.05692271422594786,
          "train_loss": 0.05300955846905708,
          "validation_loss": 0.04441585764288902,
          "train_metric": 0.005338928662240505,
          "validation_metric": 0.004473403096199036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13056760700055747
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.05192438559606671,
          "train_loss": 0.04864034801721573,
          "validation_loss": 0.041195981204509735,
          "train_metric": 0.004898877814412117,
          "validation_metric": 0.004149108659476042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14357016000030853
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04821514885406941,
          "train_loss": 0.04538911208510399,
          "validation_loss": 0.0378991961479187,
          "train_metric": 0.004571424797177315,
          "validation_metric": 0.003817068412899971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15646096599994053
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.043889874359592795,
          "train_loss": 0.0417974516749382,
          "validation_loss": 0.03606892377138138,
          "train_metric": 0.004209686536341906,
          "validation_metric": 0.0036327301058918238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16933462000088184
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03815814096014947,
          "train_loss": 0.03506443277001381,
          "validation_loss": 0.03014182671904564,
          "train_metric": 0.003531561465933919,
          "validation_metric": 0.003035774454474449,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18218474900095316
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.036669535911642015,
          "train_loss": 0.03197959437966347,
          "validation_loss": 0.025836357846856117,
          "train_metric": 0.003220867831259966,
          "validation_metric": 0.0026021432131528854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19500888800030225
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.030701296171173453,
          "train_loss": 0.02973571978509426,
          "validation_loss": 0.026119384914636612,
          "train_metric": 0.002994873095303774,
          "validation_metric": 0.002630649134516716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2074864650003292
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.029187815554905683,
          "train_loss": 0.0282464399933815,
          "validation_loss": 0.0235819760710001,
          "train_metric": 0.0028448777738958597,
          "validation_metric": 0.002375090029090643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22047492900128418
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.03107411286327988,
          "train_loss": 0.0312570221722126,
          "validation_loss": 0.0255034938454628,
          "train_metric": 0.003148092655465007,
          "validation_metric": 0.0025686186272650957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23279084500063618
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.026603606034768745,
          "train_loss": 0.022891534492373466,
          "validation_loss": 0.019505923613905907,
          "train_metric": 0.0023055514320731163,
          "validation_metric": 0.001964565133675933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2455844910000451
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.021424310165457428,
          "train_loss": 0.020309915766119957,
          "validation_loss": 0.01723344251513481,
          "train_metric": 0.0020455403719097376,
          "validation_metric": 0.001735689234919846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2580280260008294
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.022649855935014784,
          "train_loss": 0.021446969360113144,
          "validation_loss": 0.018379081040620804,
          "train_metric": 0.0021600599866360426,
          "validation_metric": 0.0018510741647332907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27062791299977107
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.0217111871461384,
          "train_loss": 0.020019279792904854,
          "validation_loss": 0.01657511666417122,
          "train_metric": 0.0020162686705589294,
          "validation_metric": 0.0016693848883733153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2834101810003631
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.019672927679494023,
          "train_loss": 0.019388247281312943,
          "validation_loss": 0.01678743213415146,
          "train_metric": 0.0019527132390066981,
          "validation_metric": 0.0016907688695937395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29595034700105316
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.02046148377121426,
          "train_loss": 0.01898162253201008,
          "validation_loss": 0.015953736379742622,
          "train_metric": 0.0019117594929412007,
          "validation_metric": 0.0016068019904196262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.308202481999615
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.026537406258285046,
          "train_loss": 0.023082714527845383,
          "validation_loss": 0.017485501244664192,
          "train_metric": 0.002324806060642004,
          "validation_metric": 0.0017610754584893584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3202575129998877
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.020713517151307315,
          "train_loss": 0.019347740337252617,
          "validation_loss": 0.015051767230033875,
          "train_metric": 0.0019486336968839169,
          "validation_metric": 0.001515959040261805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3328915609999967
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01880419405642897,
          "train_loss": 0.016354875639081,
          "validation_loss": 0.013926394283771515,
          "train_metric": 0.0016472033457830548,
          "validation_metric": 0.001402615336701274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42992150700047205
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.016534791298909113,
          "train_loss": 0.015039922669529915,
          "validation_loss": 0.012758416123688221,
          "train_metric": 0.001514765783213079,
          "validation_metric": 0.0012849807972088456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4431478589995095
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.018539246986620128,
          "train_loss": 0.015150820836424828,
          "validation_loss": 0.012505925260484219,
          "train_metric": 0.0015259350184351206,
          "validation_metric": 0.0012595509178936481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4559508980000828
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01845882658381015,
          "train_loss": 0.0263831727206707,
          "validation_loss": 0.021692663431167603,
          "train_metric": 0.002657216275110841,
          "validation_metric": 0.0021848054602742195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46825766999972984
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.01887913089012727,
          "train_loss": 0.01708083227276802,
          "validation_loss": 0.014424415305256844,
          "train_metric": 0.0017203189199790359,
          "validation_metric": 0.0014527742750942707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4804072189999715
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.015706078353105113,
          "train_loss": 0.013623007573187351,
          "validation_loss": 0.01145058125257492,
          "train_metric": 0.001372059341520071,
          "validation_metric": 0.001153260818682611,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49259455100036575
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.018663879949599504,
          "train_loss": 0.01805170252919197,
          "validation_loss": 0.014781827107071877,
          "train_metric": 0.0018181015038862824,
          "validation_metric": 0.0014887715224176645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5051089270000375
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01752010887139477,
          "train_loss": 0.01663573645055294,
          "validation_loss": 0.014080896973609924,
          "train_metric": 0.0016754904063418508,
          "validation_metric": 0.0014181763399392366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5175977840008272
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.015418506023706868,
          "train_loss": 0.01448780857026577,
          "validation_loss": 0.011648541316390038,
          "train_metric": 0.0014591589570045471,
          "validation_metric": 0.0011731984559446573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5299449490012194
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.013958298251964152,
          "train_loss": 0.012688550166785717,
          "validation_loss": 0.010741646401584148,
          "train_metric": 0.001277944422326982,
          "validation_metric": 0.0010818593436852098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5425839930012444
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.014501343452138826,
          "train_loss": 0.015077909454703331,
          "validation_loss": 0.012522627599537373,
          "train_metric": 0.0015185917727649212,
          "validation_metric": 0.0012612331192940474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5554806020008982
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.015160513867158443,
          "train_loss": 0.015085410326719284,
          "validation_loss": 0.01296358834952116,
          "train_metric": 0.0015193474246188998,
          "validation_metric": 0.0013056453317403793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.567716563000431
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.0168740670196712,
          "train_loss": 0.012578879483044147,
          "validation_loss": 0.010490906424820423,
          "train_metric": 0.0012668987037613988,
          "validation_metric": 0.00105660583358258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5812301170008141
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012558586371596903,
          "train_loss": 0.011350860819220543,
          "validation_loss": 0.009651830419898033,
          "train_metric": 0.0011432170867919922,
          "validation_metric": 0.0009720969246700406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5938595549996535
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.012008170073386282,
          "train_loss": 0.01158946193754673,
          "validation_loss": 0.009304620325565338,
          "train_metric": 0.001167248236015439,
          "validation_metric": 0.000937127391807735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6067224819998955
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.01317415907396935,
          "train_loss": 0.011593138799071312,
          "validation_loss": 0.00916856899857521,
          "train_metric": 0.0011676185531541705,
          "validation_metric": 0.0009234249009750783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6201531080005225
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.012022785740555264,
          "train_loss": 0.017954226583242416,
          "validation_loss": 0.015289584174752235,
          "train_metric": 0.0018082840833812952,
          "validation_metric": 0.0015399110270664096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6326333480010362
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014909820165485144,
          "train_loss": 0.01431149709969759,
          "validation_loss": 0.012048734351992607,
          "train_metric": 0.0014414016623049974,
          "validation_metric": 0.0012135044671595097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6447515239997301
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013073766545858234,
          "train_loss": 0.010623696260154247,
          "validation_loss": 0.008926188573241234,
          "train_metric": 0.0010699797421693802,
          "validation_metric": 0.0008990128408186138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6578694410000026
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010758196411188692,
          "train_loss": 0.009987705387175083,
          "validation_loss": 0.008647993206977844,
          "train_metric": 0.0010059250053018332,
          "validation_metric": 0.0008709942921996117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6716319629995269
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.01085945064551197,
          "train_loss": 0.010169338434934616,
          "validation_loss": 0.008769248612225056,
          "train_metric": 0.0010242185089737177,
          "validation_metric": 0.0008832066087052226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.684572942000159
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.01109956661821343,
          "train_loss": 0.01147399004548788,
          "validation_loss": 0.00943567045032978,
          "train_metric": 0.0011556182289496064,
          "validation_metric": 0.0009503261535428464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6968502150011773
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.012931099772686139,
          "train_loss": 0.00975019484758377,
          "validation_loss": 0.007961606606841087,
          "train_metric": 0.0009820038685575128,
          "validation_metric": 0.0008018639055080712,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7093442090008466
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009785887697944418,
          "train_loss": 0.009279869496822357,
          "validation_loss": 0.007988525554537773,
          "train_metric": 0.000934634474106133,
          "validation_metric": 0.0008045749855227768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7214403420002782
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.009299000463215634,
          "train_loss": 0.010184766724705696,
          "validation_loss": 0.009129703044891357,
          "train_metric": 0.0010257724206894636,
          "validation_metric": 0.0009195100865326822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7337504440001794
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.010168677967158146,
          "train_loss": 0.00883601326495409,
          "validation_loss": 0.0075323511846363544,
          "train_metric": 0.0008899309323169291,
          "validation_metric": 0.0007586308638565242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7466740090003441
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.010722120001446456,
          "train_loss": 0.010039329528808594,
          "validation_loss": 0.008807784877717495,
          "train_metric": 0.0010111245792359114,
          "validation_metric": 0.0008870878373272717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7589067950011668
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.010452699178131297,
          "train_loss": 0.00902954675257206,
          "validation_loss": 0.007872484624385834,
          "train_metric": 0.000909422873519361,
          "validation_metric": 0.0007928877603262663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7710841160005657
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.009185833274386823,
          "train_loss": 0.009410109370946884,
          "validation_loss": 0.007505800575017929,
          "train_metric": 0.0009477517451159656,
          "validation_metric": 0.0007559568039141595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7840819129996817
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.008959966580732726,
          "train_loss": 0.008251712657511234,
          "validation_loss": 0.006896935403347015,
          "train_metric": 0.0008310822304338217,
          "validation_metric": 0.0006946339854039252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7967985779996525
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008822539093671367,
          "train_loss": 0.00829176977276802,
          "validation_loss": 0.006800055969506502,
          "train_metric": 0.0008351167198270559,
          "validation_metric": 0.0006848766934126616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8096155870007351
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.008354039629921317,
          "train_loss": 0.007905767299234867,
          "validation_loss": 0.006627353839576244,
          "train_metric": 0.0007962399977259338,
          "validation_metric": 0.0006674827891401947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8228402879994974
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.009528101363684982,
          "train_loss": 0.008216619491577148,
          "validation_loss": 0.007351606618613005,
          "train_metric": 0.0008275477448478341,
          "validation_metric": 0.0007404269417747855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8350792950004688
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.00792392689618282,
          "train_loss": 0.009343978017568588,
          "validation_loss": 0.00732021126896143,
          "train_metric": 0.0009410912753082812,
          "validation_metric": 0.0007372648105956614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8478841110008943
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.00862078758655116,
          "train_loss": 0.007861784659326077,
          "validation_loss": 0.006398169323801994,
          "train_metric": 0.0007918101036921144,
          "validation_metric": 0.0006444001919589937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8610213390002173
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.008397341953241266,
          "train_loss": 0.00785257201641798,
          "validation_loss": 0.006636479403823614,
          "train_metric": 0.0007908822735771537,
          "validation_metric": 0.0006684020045213401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8737848439996014
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.00792282038310077,
          "train_loss": 0.007761125452816486,
          "validation_loss": 0.006691877264529467,
          "train_metric": 0.0007816721918061376,
          "validation_metric": 0.0006739814998582006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8865085140005249
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.008166019440977834,
          "train_loss": 0.007140409667044878,
          "validation_loss": 0.006099289748817682,
          "train_metric": 0.0007191559998318553,
          "validation_metric": 0.000614298100117594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9001998840012675
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.009080607065698132,
          "train_loss": 0.008042611181735992,
          "validation_loss": 0.007286658510565758,
          "train_metric": 0.0008100224076770246,
          "validation_metric": 0.0007338855066336691,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9128535590007232
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.00811591005185619,
          "train_loss": 0.007554612122476101,
          "validation_loss": 0.006213771179318428,
          "train_metric": 0.0007608729065395892,
          "validation_metric": 0.00062582828104496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9249818330008566
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008824263059068471,
          "train_loss": 0.008121580816805363,
          "validation_loss": 0.006312828045338392,
          "train_metric": 0.0008179758442565799,
          "validation_metric": 0.0006358049577102065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9377588540010038
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.00832608449854888,
          "train_loss": 0.010289219208061695,
          "validation_loss": 0.008493580855429173,
          "train_metric": 0.0010362924076616764,
          "validation_metric": 0.0008554423111490905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9498950079996575
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.00877501253853552,
          "train_loss": 0.006870351266115904,
          "validation_loss": 0.006006564479321241,
          "train_metric": 0.0006919566658325493,
          "validation_metric": 0.0006049592047929764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9632804800003214
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.007299505450646393,
          "train_loss": 0.006946505978703499,
          "validation_loss": 0.005785537883639336,
          "train_metric": 0.0006996265728957951,
          "validation_metric": 0.0005826981505379081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9764455860004091
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.007085959150572307,
          "train_loss": 0.007358310278505087,
          "validation_loss": 0.005880382843315601,
          "train_metric": 0.0007411020924337208,
          "validation_metric": 0.0005922506679780781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9890196650012513
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.007447212090482935,
          "train_loss": 0.006344006396830082,
          "validation_loss": 0.0052243866957724094,
          "train_metric": 0.0006389450281858444,
          "validation_metric": 0.0005261811311356723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0017529400010972
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006892835786857177,
          "train_loss": 0.007097315043210983,
          "validation_loss": 0.005499411839991808,
          "train_metric": 0.000714815512765199,
          "validation_metric": 0.0005538805271498859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0142808279997553
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.007321704317291733,
          "train_loss": 0.008506550453603268,
          "validation_loss": 0.007234817836433649,
          "train_metric": 0.0008567486074753106,
          "validation_metric": 0.0007286643958650529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0262484160011809
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.008910495293093845,
          "train_loss": 0.006475388538092375,
          "validation_loss": 0.0055541484616696835,
          "train_metric": 0.0006521774921566248,
          "validation_metric": 0.0005593935493379831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0389458959998592
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.009247284091543406,
          "train_loss": 0.006639064289629459,
          "validation_loss": 0.005873969290405512,
          "train_metric": 0.0006686621927656233,
          "validation_metric": 0.0005916046793572605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0515486069998587
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.007987596996827051,
          "train_loss": 0.006319202948361635,
          "validation_loss": 0.0054232282564044,
          "train_metric": 0.0006364469300024211,
          "validation_metric": 0.0005462076514959335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0643070460009767
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006773542772862129,
          "train_loss": 0.006594016216695309,
          "validation_loss": 0.005362479016184807,
          "train_metric": 0.0006641251966357231,
          "validation_metric": 0.0005400892696343362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0766004440010875
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.007274927746038884,
          "train_loss": 0.009023490361869335,
          "validation_loss": 0.0061689093708992004,
          "train_metric": 0.0009088129736483097,
          "validation_metric": 0.0006213099113665521,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0889875320008287
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.007789886178215966,
          "train_loss": 0.005683409050107002,
          "validation_loss": 0.004555889405310154,
          "train_metric": 0.0005724121583625674,
          "validation_metric": 0.0004588525043800473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1023181340005976
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.006407412583939731,
          "train_loss": 0.005831774324178696,
          "validation_loss": 0.004895292688161135,
          "train_metric": 0.0005873549380339682,
          "validation_metric": 0.0004930360009893775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1147897520004335
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.008250024897279218,
          "train_loss": 0.007393504958599806,
          "validation_loss": 0.005451825447380543,
          "train_metric": 0.0007446467061527073,
          "validation_metric": 0.0005490879411809146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.126831449000747
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.007269675232237205,
          "train_loss": 0.005730052478611469,
          "validation_loss": 0.0045626298524439335,
          "train_metric": 0.0005771099240519106,
          "validation_metric": 0.00045953135122545063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1393459999999322
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.008450913534034044,
          "train_loss": 0.0053389607928693295,
          "validation_loss": 0.004476289264857769,
          "train_metric": 0.0005377205670811236,
          "validation_metric": 0.00045083544682711363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.152009667001039
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.006526127574034035,
          "train_loss": 0.005047039594501257,
          "validation_loss": 0.004337061196565628,
          "train_metric": 0.0005083194118924439,
          "validation_metric": 0.0004368129011709243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1648803309999494
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.00627539461129345,
          "train_loss": 0.005932082422077656,
          "validation_loss": 0.004661512095481157,
          "train_metric": 0.0005974575760774314,
          "validation_metric": 0.0004694904782809317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1771755610006949
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.00632219002727652,
          "train_loss": 0.006402188912034035,
          "validation_loss": 0.005544474348425865,
          "train_metric": 0.0006448050262406468,
          "validation_metric": 0.0005584191530942917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.189506769000218
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.005320466771081556,
          "train_loss": 0.004904362838715315,
          "validation_loss": 0.004219850059598684,
          "train_metric": 0.0004939495120197535,
          "validation_metric": 0.000425007805461064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2024047869999777
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.006083436084736604,
          "train_loss": 0.007824964821338654,
          "validation_loss": 0.0056642391718924046,
          "train_metric": 0.0007881018682383001,
          "validation_metric": 0.0005704814102500677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2154905759998655
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.00654576595115941,
          "train_loss": 0.004803966730833054,
          "validation_loss": 0.0038239862769842148,
          "train_metric": 0.00048383793910034,
          "validation_metric": 0.00038513782783411443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2288612709999143
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.005305877326463815,
          "train_loss": 0.004515648819506168,
          "validation_loss": 0.003774329088628292,
          "train_metric": 0.00045479965046979487,
          "validation_metric": 0.00038013659650459886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.241479980000804
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.005167598792468198,
          "train_loss": 0.0044187563471496105,
          "validation_loss": 0.0036996789276599884,
          "train_metric": 0.0004450409614946693,
          "validation_metric": 0.000372618087567389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2543453460002638
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0054132111254148185,
          "train_loss": 0.004610396921634674,
          "validation_loss": 0.003630898892879486,
          "train_metric": 0.0004643423017114401,
          "validation_metric": 0.00036569085204973817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2675057489996107
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004798470545210876,
          "train_loss": 0.0043950313702225685,
          "validation_loss": 0.003675587009638548,
          "train_metric": 0.00044265147880651057,
          "validation_metric": 0.0003701916430145502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2800814769998397
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004619571947841905,
          "train_loss": 0.005293216556310654,
          "validation_loss": 0.0039157564751803875,
          "train_metric": 0.0005331133725121617,
          "validation_metric": 0.0003943807096220553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2921536960002413
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0054721724591217935,
          "train_loss": 0.004695390351116657,
          "validation_loss": 0.004030286334455013,
          "train_metric": 0.00047290243674069643,
          "validation_metric": 0.0004059156635776162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3038506989996677
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.004466690545086749,
          "train_loss": 0.004083320032805204,
          "validation_loss": 0.0034895611461251974,
          "train_metric": 0.0004112569731660187,
          "validation_metric": 0.00035145581932738423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4002360180002142
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.005115434461913537,
          "train_loss": 0.007258309982717037,
          "validation_loss": 0.005510220304131508,
          "train_metric": 0.0007310304208658636,
          "validation_metric": 0.0005549692432396114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.414250752000953
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.009293196591897868,
          "train_loss": 0.005568325519561768,
          "validation_loss": 0.004445022903382778,
          "train_metric": 0.0005608212668448687,
          "validation_metric": 0.00044768641237169504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4275145290012006
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.005500021390616894,
          "train_loss": 0.008761043660342693,
          "validation_loss": 0.006037356331944466,
          "train_metric": 0.0008823800599202514,
          "validation_metric": 0.0006080605089664459,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4399565660005464
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0070423808647319674,
          "train_loss": 0.007777173537760973,
          "validation_loss": 0.007064220495522022,
          "train_metric": 0.0007832883857190609,
          "validation_metric": 0.000711482425685972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4533506970001326
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.005583135949564166,
          "train_loss": 0.004451039247214794,
          "validation_loss": 0.0038050871808081865,
          "train_metric": 0.00044829241232946515,
          "validation_metric": 0.00038323443732224405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4662624780012266
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.005064301345555577,
          "train_loss": 0.00429256446659565,
          "validation_loss": 0.0037014917470514774,
          "train_metric": 0.00043233134783804417,
          "validation_metric": 0.00037280068499967456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.479266975000428
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0042272973332728725,
          "train_loss": 0.003975697793066502,
          "validation_loss": 0.0034713027998805046,
          "train_metric": 0.000400417746277526,
          "validation_metric": 0.00034961692290380597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.492796205000559
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0043412219019955955,
          "train_loss": 0.003991357982158661,
          "validation_loss": 0.0030240491032600403,
          "train_metric": 0.0004019949701614678,
          "validation_metric": 0.0003045711782760918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5057952330007538
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004158979114436079,
          "train_loss": 0.0047474028542637825,
          "validation_loss": 0.0036211239639669657,
          "train_metric": 0.0004781410680152476,
          "validation_metric": 0.0003647063276730478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5188700159997097
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004463581753952894,
          "train_loss": 0.0042828344739973545,
          "validation_loss": 0.0032419017516076565,
          "train_metric": 0.0004313514509703964,
          "validation_metric": 0.0003265124396421015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5316377520011883
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.006689408182865009,
          "train_loss": 0.004371163435280323,
          "validation_loss": 0.003327373182401061,
          "train_metric": 0.00044024758972227573,
          "validation_metric": 0.0003351208579260856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5440056120005465
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.00480348538258113,
          "train_loss": 0.004481618292629719,
          "validation_loss": 0.003416909836232662,
          "train_metric": 0.00045137220877222717,
          "validation_metric": 0.00034413867979310453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.556377261000307
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.00430601309199119,
          "train_loss": 0.006890431046485901,
          "validation_loss": 0.004803858231753111,
          "train_metric": 0.0006939790328033268,
          "validation_metric": 0.00048382702516391873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.569465955000851
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.005239754376816563,
          "train_loss": 0.0041156671941280365,
          "validation_loss": 0.0032125820871442556,
          "train_metric": 0.00041451488505117595,
          "validation_metric": 0.0003235595067963004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.583424221000314
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.0047459820198128,
          "train_loss": 0.004670476540923119,
          "validation_loss": 0.00380503386259079,
          "train_metric": 0.0004703932791016996,
          "validation_metric": 0.00038322905311360955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5962476360000437
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004061326224473305,
          "train_loss": 0.0057152034714818,
          "validation_loss": 0.0050643268041312695,
          "train_metric": 0.0005756142782047391,
          "validation_metric": 0.0005100604612380266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.608753534999778
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0058341369513073005,
          "train_loss": 0.0037456953432410955,
          "validation_loss": 0.003334058914333582,
          "train_metric": 0.0003772526979446411,
          "validation_metric": 0.00033579423325136304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6211447910009156
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0038496894849231467,
          "train_loss": 0.0035776777658611536,
          "validation_loss": 0.0027867243625223637,
          "train_metric": 0.0003603305958677083,
          "validation_metric": 0.0002806686970870942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6351986800000304
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.003668425437354017,
          "train_loss": 0.003872461384162307,
          "validation_loss": 0.0035358238965272903,
          "train_metric": 0.00039002011180855334,
          "validation_metric": 0.000356115197064355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6481262689994765
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003431274635659065,
          "train_loss": 0.0034102704375982285,
          "validation_loss": 0.0025599426589906216,
          "train_metric": 0.0003434699319768697,
          "validation_metric": 0.00025782804004848003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6607152830001723
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003475960016658064,
          "train_loss": 0.0032833025325089693,
          "validation_loss": 0.0028894594870507717,
          "train_metric": 0.00033068223274312913,
          "validation_metric": 0.00029101583641022444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6731302339994727
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0040749234467512,
          "train_loss": 0.00403990363702178,
          "validation_loss": 0.002771301893517375,
          "train_metric": 0.00040688420995138586,
          "validation_metric": 0.0002791153674479574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6860636900000827
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.00426553058787249,
          "train_loss": 0.0032655070535838604,
          "validation_loss": 0.002753407694399357,
          "train_metric": 0.000328889989759773,
          "validation_metric": 0.00027731311274692416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.699096460999499
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.003762999862374272,
          "train_loss": 0.0033455542288720608,
          "validation_loss": 0.0027879609260708094,
          "train_metric": 0.0003369519836269319,
          "validation_metric": 0.00028079317416995764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.712023019001208
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.003976885753218085,
          "train_loss": 0.003281041979789734,
          "validation_loss": 0.0023889169096946716,
          "train_metric": 0.0003304545534774661,
          "validation_metric": 0.00024060296709649265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7250876190009876
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.004135214221605565,
          "train_loss": 0.0038991966284811497,
          "validation_loss": 0.0034921634942293167,
          "train_metric": 0.00039271279820241034,
          "validation_metric": 0.0003517179284244776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7382720300010988
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.003929747297661379,
          "train_loss": 0.0032486054114997387,
          "validation_loss": 0.0025915673468261957,
          "train_metric": 0.00032718764850869775,
          "validation_metric": 0.0002610131341498345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7512218629999552
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.004098122510185931,
          "train_loss": 0.0033484052401036024,
          "validation_loss": 0.0029053143225610256,
          "train_metric": 0.00033723912201821804,
          "validation_metric": 0.00029261261806823313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.764563699000064
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.004465967649593949,
          "train_loss": 0.0035924166440963745,
          "validation_loss": 0.0029879582580178976,
          "train_metric": 0.00036181503674015403,
          "validation_metric": 0.00030093619716353714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7770761330011737
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0038345800203387626,
          "train_loss": 0.00404722522944212,
          "validation_loss": 0.0035219790879637003,
          "train_metric": 0.00040762173011898994,
          "validation_metric": 0.00035472080344334245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7904670080006326
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.003698163513035979,
          "train_loss": 0.004568993113934994,
          "validation_loss": 0.0032586907036602497,
          "train_metric": 0.0004601722757797688,
          "validation_metric": 0.0003282034012954682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8034156869998696
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0031250577740138397,
          "train_loss": 0.0039499192498624325,
          "validation_loss": 0.003451134078204632,
          "train_metric": 0.00039782136445865035,
          "validation_metric": 0.00034758556284941733,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.816795662000004
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.005473515935591422,
          "train_loss": 0.004694379400461912,
          "validation_loss": 0.003842829493805766,
          "train_metric": 0.0004728006897494197,
          "validation_metric": 0.0003870357177220285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8298716870012868
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.004709988679678645,
          "train_loss": 0.0033802310936152935,
          "validation_loss": 0.003248101333156228,
          "train_metric": 0.00034044450148940086,
          "validation_metric": 0.0003271368332207203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8422697690002678
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.003481393046968151,
          "train_loss": 0.0027800349052995443,
          "validation_loss": 0.002337930491194129,
          "train_metric": 0.0002799949434120208,
          "validation_metric": 0.00023546782904304564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8553909029997158
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0029804148180119228,
          "train_loss": 0.003126707160845399,
          "validation_loss": 0.002280800137668848,
          "train_metric": 0.00031491051777265966,
          "validation_metric": 0.0002297138998983428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8688725669999258
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.003752286807866767,
          "train_loss": 0.007945841178297997,
          "validation_loss": 0.006590164266526699,
          "train_metric": 0.0008002759423106909,
          "validation_metric": 0.0006637371843680739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8822140350002883
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.00460196852509398,
          "train_loss": 0.005487068556249142,
          "validation_loss": 0.004278750624507666,
          "train_metric": 0.0005526375607587397,
          "validation_metric": 0.0004309401265345514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.895738868999615
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.006206100035342388,
          "train_loss": 0.007751016411930323,
          "validation_loss": 0.005675944034010172,
          "train_metric": 0.0007806538487784564,
          "validation_metric": 0.0005716602900065482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9085880580005323
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.006595768471015617,
          "train_loss": 0.004143876954913139,
          "validation_loss": 0.0035119052045047283,
          "train_metric": 0.00041735608829185367,
          "validation_metric": 0.0003537062439136207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9212244720001763
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0040652345269336365,
          "train_loss": 0.003048161743208766,
          "validation_loss": 0.00244358298368752,
          "train_metric": 0.0003069997183047235,
          "validation_metric": 0.00024610874243080616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9348487019997265
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.003774748998694122,
          "train_loss": 0.003310101106762886,
          "validation_loss": 0.0026716412976384163,
          "train_metric": 0.0003333812637720257,
          "validation_metric": 0.00026907792198471725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9482709000003524
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0031564516430080403,
          "train_loss": 0.0035146118607372046,
          "validation_loss": 0.0025201498065143824,
          "train_metric": 0.0003539788885973394,
          "validation_metric": 0.0002538202388677746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9609230540008866
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0033680221022223122,
          "train_loss": 0.002976024989038706,
          "validation_loss": 0.0021755786146968603,
          "train_metric": 0.0002997343544848263,
          "validation_metric": 0.0002191162930103019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9738653690001229
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.003679667101096129,
          "train_loss": 0.00657980190590024,
          "validation_loss": 0.005838960409164429,
          "train_metric": 0.0006626935210078955,
          "validation_metric": 0.0005880787502974272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9871179350011516
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0051701319753192365,
          "train_loss": 0.00899307057261467,
          "validation_loss": 0.006033634766936302,
          "train_metric": 0.0009057490387931466,
          "validation_metric": 0.0006076857098378241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9997663240010297
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.004010327807918657,
          "train_loss": 0.003366546705365181,
          "validation_loss": 0.0028376562986522913,
          "train_metric": 0.00033906628959812224,
          "validation_metric": 0.0002857983345165849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0131734750011674
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.003569387008610647,
          "train_loss": 0.005581708624958992,
          "validation_loss": 0.004538032691925764,
          "train_metric": 0.0005621691816486418,
          "validation_metric": 0.000457054004073143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.026505095000175
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0037843942336621694,
          "train_loss": 0.003170205047354102,
          "validation_loss": 0.0023849220015108585,
          "train_metric": 0.0003192914300598204,
          "validation_metric": 0.0002402006066404283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0392977290011913
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.003354438355017919,
          "train_loss": 0.00459077674895525,
          "validation_loss": 0.004043848719447851,
          "train_metric": 0.00046236629714258015,
          "validation_metric": 0.00040728162275627255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0523503229997004
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.004156243034231011,
          "train_loss": 0.0030869843903928995,
          "validation_loss": 0.0027414814103394747,
          "train_metric": 0.0003109097888227552,
          "validation_metric": 0.00027611193945631385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0656620259997
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.004081008482899051,
          "train_loss": 0.003294785274192691,
          "validation_loss": 0.002753706183284521,
          "train_metric": 0.00033183873165398836,
          "validation_metric": 0.0002773432061076164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.078465487000358
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0032478058710694313,
          "train_loss": 0.00279452046379447,
          "validation_loss": 0.0022255636285990477,
          "train_metric": 0.0002814538311213255,
          "validation_metric": 0.000224150629946962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.090735792000487
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0028702022100333124,
          "train_loss": 0.00276703923009336,
          "validation_loss": 0.0025073972065001726,
          "train_metric": 0.0002786860568448901,
          "validation_metric": 0.0002525358577258885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1038746550002543
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0027522632117324974,
          "train_loss": 0.0025232008192688227,
          "validation_loss": 0.0019318928243592381,
          "train_metric": 0.00025412754621356726,
          "validation_metric": 0.00019457317830529064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.117580304000512
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0025850631136563607,
          "train_loss": 0.0028609763830900192,
          "validation_loss": 0.002445514313876629,
          "train_metric": 0.0002881470718421042,
          "validation_metric": 0.0002463032433297485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1309651900010067
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.002829283388564363,
          "train_loss": 0.0025855842977762222,
          "validation_loss": 0.002067727269604802,
          "train_metric": 0.00026041065575554967,
          "validation_metric": 0.00020825389947276562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1439253500011546
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0031839741277508438,
          "train_loss": 0.0023646021727472544,
          "validation_loss": 0.0018151021795347333,
          "train_metric": 0.00023815403983462602,
          "validation_metric": 0.0001828104432206601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1573890290001145
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0027490845568536315,
          "train_loss": 0.002399064600467682,
          "validation_loss": 0.0018712787423282862,
          "train_metric": 0.00024162499175872654,
          "validation_metric": 0.0001884683733806014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1715032200008864
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0031400529442180414,
          "train_loss": 0.0028460060711950064,
          "validation_loss": 0.0025089415721595287,
          "train_metric": 0.0002866392896976322,
          "validation_metric": 0.00025269141769967973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1854900020007335
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.002789598307572305,
          "train_loss": 0.0028137885965406895,
          "validation_loss": 0.002434448804706335,
          "train_metric": 0.00028339450364001095,
          "validation_metric": 0.0002451887703500688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.198912185000154
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.003809572765021585,
          "train_loss": 0.0048950668424367905,
          "validation_loss": 0.003349293489009142,
          "train_metric": 0.0004930131835862994,
          "validation_metric": 0.00033732858719304204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2131564680003066
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.004055819161294494,
          "train_loss": 0.011800894513726234,
          "validation_loss": 0.010151294991374016,
          "train_metric": 0.0011885428102687001,
          "validation_metric": 0.001022401382215321,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.228045887999542
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.004853298640227877,
          "train_loss": 0.002235913649201393,
          "validation_loss": 0.0017751220148056746,
          "train_metric": 0.00022519305639434606,
          "validation_metric": 0.00017878379730973393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2422649529999035
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0026660376934160013,
          "train_loss": 0.0021206627134233713,
          "validation_loss": 0.0017292676493525505,
          "train_metric": 0.000213585386518389,
          "validation_metric": 0.0001741655432851985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.257033195999611
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.003485754870780511,
          "train_loss": 0.00348101113922894,
          "validation_loss": 0.003160313703119755,
          "train_metric": 0.0003505946951918304,
          "validation_metric": 0.0003182952059432864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2709578260000853
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.003244832194468472,
          "train_loss": 0.0024774083867669106,
          "validation_loss": 0.0020656567066907883,
          "train_metric": 0.0002495154913049191,
          "validation_metric": 0.0002080453559756279,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2849941849999595
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0030475716848741286,
          "train_loss": 0.005973652470856905,
          "validation_loss": 0.005114113446325064,
          "train_metric": 0.0006016443949192762,
          "validation_metric": 0.0005150748183950782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.299126914000226
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.004727905514300801,
          "train_loss": 0.003269131761044264,
          "validation_loss": 0.0025070724077522755,
          "train_metric": 0.00032925495179370046,
          "validation_metric": 0.0002525031450204551,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.312279829000545
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0027398399397497997,
          "train_loss": 0.002641512081027031,
          "validation_loss": 0.001916980487294495,
          "train_metric": 0.00026604338199831545,
          "validation_metric": 0.00019307126058265567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3256163719997858
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0029780917684547603,
          "train_loss": 0.00417494960129261,
          "validation_loss": 0.0035948825534433126,
          "train_metric": 0.0004204857104923576,
          "validation_metric": 0.0003620634088292718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3392187999997986
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0032674245521775447,
          "train_loss": 0.0021802710834890604,
          "validation_loss": 0.0017106401501223445,
          "train_metric": 0.00021958892466500401,
          "validation_metric": 0.00017228940851055086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3531495980005275
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.002684503182535991,
          "train_loss": 0.0035957361105829477,
          "validation_loss": 0.0025044127833098173,
          "train_metric": 0.0003621494397521019,
          "validation_metric": 0.00025223527336493134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.366791316000672
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.003255766854636022,
          "train_loss": 0.003337820991873741,
          "validation_loss": 0.0030949178617447615,
          "train_metric": 0.00033617307781241834,
          "validation_metric": 0.0003117088053841144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3807557720010664
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0037180712351982947,
          "train_loss": 0.002617832040414214,
          "validation_loss": 0.0022864998318254948,
          "train_metric": 0.0002636584686115384,
          "validation_metric": 0.00023028790019452572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.39395763200082
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.00374288151942892,
          "train_loss": 0.00292043830268085,
          "validation_loss": 0.0025179588701575994,
          "train_metric": 0.0002941358252428472,
          "validation_metric": 0.0002535996027290821,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.407065348999822
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.005458810861455277,
          "train_loss": 0.0034703221172094345,
          "validation_loss": 0.0031392856035381556,
          "train_metric": 0.0003495181445032358,
          "validation_metric": 0.00031617734930478036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.420798057999491
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0030551268355338834,
          "train_loss": 0.0029003082308918238,
          "validation_loss": 0.0018579765455797315,
          "train_metric": 0.0002921083942055702,
          "validation_metric": 0.00018712860764935613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.442467373999534
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.003702129108205554,
          "train_loss": 0.004418022930622101,
          "validation_loss": 0.003489421447739005,
          "train_metric": 0.0004449671250768006,
          "validation_metric": 0.0003514417912811041,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.455656015999921
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0029356873164942954,
          "train_loss": 0.0033378482330590487,
          "validation_loss": 0.0029822743963450193,
          "train_metric": 0.00033617584267631173,
          "validation_metric": 0.00030036375392228365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4687213600009272
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.004215932549413992,
          "train_loss": 0.004695879761129618,
          "validation_loss": 0.0033489190973341465,
          "train_metric": 0.00047295185504481196,
          "validation_metric": 0.00033729083952493966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4817023830000835
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.002804675095831044,
          "train_loss": 0.002440971089527011,
          "validation_loss": 0.0018740561790764332,
          "train_metric": 0.00024584567290730774,
          "validation_metric": 0.00018874806119129062,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4946692800003802
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.002649458205269184,
          "train_loss": 0.003677861765027046,
          "validation_loss": 0.003414456034079194,
          "train_metric": 0.00037042071926407516,
          "validation_metric": 0.00034389147185720503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.50708602099985
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.003220545178919565,
          "train_loss": 0.002599406288936734,
          "validation_loss": 0.0021601030603051186,
          "train_metric": 0.00026180269196629524,
          "validation_metric": 0.0002175576810259372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5202126929998485
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.003420785469643306,
          "train_loss": 0.0020872263703495264,
          "validation_loss": 0.0016189772868528962,
          "train_metric": 0.00021021779684815556,
          "validation_metric": 0.00016305746976286173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.534112855000785
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0020096653679502197,
          "train_loss": 0.003296771552413702,
          "validation_loss": 0.002508441684767604,
          "train_metric": 0.00033203873317688704,
          "validation_metric": 0.0002526410680729896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5486754850007856
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.004166206796071492,
          "train_loss": 0.0058622052893042564,
          "validation_loss": 0.004472094122320414,
          "train_metric": 0.0005904198042117059,
          "validation_metric": 0.00045041291741654277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5621508229996834
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0053228140277497005,
          "train_loss": 0.006543605588376522,
          "validation_loss": 0.005634008441120386,
          "train_metric": 0.000659047975204885,
          "validation_metric": 0.0005674366257153451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5758504630011885
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.006719931468978757,
          "train_loss": 0.0028222803957760334,
          "validation_loss": 0.002049076370894909,
          "train_metric": 0.00028424971969798207,
          "validation_metric": 0.00020637545094359666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.589191301000028
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.00525011863646796,
          "train_loss": 0.0032557626254856586,
          "validation_loss": 0.0021653936710208654,
          "train_metric": 0.00032790846307761967,
          "validation_metric": 0.00021809054305776954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6023331270007475
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.004999043594580144,
          "train_loss": 0.002472524531185627,
          "validation_loss": 0.0015654508024454117,
          "train_metric": 0.00024902360746636987,
          "validation_metric": 0.0001576664944877848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.616847769000742
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.003833059781754855,
          "train_loss": 0.003911611158400774,
          "validation_loss": 0.00292683532461524,
          "train_metric": 0.0003939630987588316,
          "validation_metric": 0.0002947801258414984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6305343300009554
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0031884511481621303,
          "train_loss": 0.0018098277505487204,
          "validation_loss": 0.0013686121674254537,
          "train_metric": 0.0001822791964514181,
          "validation_metric": 0.00013784160546492785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6445033220006735
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0023114253854146227,
          "train_loss": 0.002630658447742462,
          "validation_loss": 0.0017429918516427279,
          "train_metric": 0.00026495035854168236,
          "validation_metric": 0.0001755477424012497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6582370409996656
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002199592161559849,
          "train_loss": 0.0024730751756578684,
          "validation_loss": 0.0020491760224103928,
          "train_metric": 0.0002490790793672204,
          "validation_metric": 0.00020638546266127378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6741713379997236
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0021314984442142304,
          "train_loss": 0.0019061783095821738,
          "validation_loss": 0.0016618564259260893,
          "train_metric": 0.00019198330119252205,
          "validation_metric": 0.00016737609985284507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6916825240004982
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0024492800403095316,
          "train_loss": 0.0017344809602946043,
          "validation_loss": 0.0013823441695421934,
          "train_metric": 0.00017469056183472276,
          "validation_metric": 0.00013922463404014707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.712072325999543
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.00215211639260815,
          "train_loss": 0.0016594822518527508,
          "validation_loss": 0.001273983740247786,
          "train_metric": 0.00016713698278181255,
          "validation_metric": 0.00012831095955334604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.731717071001185
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.0033403686647943687,
          "train_loss": 0.002002072287723422,
          "validation_loss": 0.0015701944939792156,
          "train_metric": 0.00020164137822575867,
          "validation_metric": 0.0001581442338647321,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.750863481000124
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0029010401412961073,
          "train_loss": 0.0027965770568698645,
          "validation_loss": 0.001773236901499331,
          "train_metric": 0.00028166096308268607,
          "validation_metric": 0.0001785939239198342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7699299859996245
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0020268031494197203,
          "train_loss": 0.0018602947238832712,
          "validation_loss": 0.001512595103122294,
          "train_metric": 0.00018736207857728004,
          "validation_metric": 0.0001523430400993675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.790239179001219
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.002067210058157798,
          "train_loss": 0.0019682401325553656,
          "validation_loss": 0.0015329151647165418,
          "train_metric": 0.00019823393085971475,
          "validation_metric": 0.00015438960690516979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8113299460001144
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.002127755098626949,
          "train_loss": 0.001761319930665195,
          "validation_loss": 0.0014397413469851017,
          "train_metric": 0.00017739369650371373,
          "validation_metric": 0.00014500545512419194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8294484500002
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0022900044823472854,
          "train_loss": 0.0019422061741352081,
          "validation_loss": 0.0014567248290404677,
          "train_metric": 0.00019561187946237624,
          "validation_metric": 0.00014671598910354078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9572088889999577
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.002909713821281912,
          "train_loss": 0.005193569231778383,
          "validation_loss": 0.003823384642601013,
          "train_metric": 0.0005230772658251226,
          "validation_metric": 0.0003850772918667644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9747357640007976
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0042544346215436235,
          "train_loss": 0.0032404998783022165,
          "validation_loss": 0.0024744258262217045,
          "train_metric": 0.00032637128606438637,
          "validation_metric": 0.0002492151106707752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9920245249995787
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0031298808607971296,
          "train_loss": 0.001954565057530999,
          "validation_loss": 0.0014890625607222319,
          "train_metric": 0.0001968566357390955,
          "validation_metric": 0.00014997291145846248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0095798050006124
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0032545053254580125,
          "train_loss": 0.00252527859993279,
          "validation_loss": 0.0021444070152938366,
          "train_metric": 0.0002543367736507207,
          "validation_metric": 0.00021597681916318834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0269425200003752
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0026315269169572275,
          "train_loss": 0.0017835613107308745,
          "validation_loss": 0.0013034214498475194,
          "train_metric": 0.0001796337601263076,
          "validation_metric": 0.00013127582496963441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0447157100006734
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.002809586214425508,
          "train_loss": 0.0016668526222929358,
          "validation_loss": 0.0013281241990625858,
          "train_metric": 0.000167879305081442,
          "validation_metric": 0.00013376379502005875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.062275333000798
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.002136265622539213,
          "train_loss": 0.002837948966771364,
          "validation_loss": 0.0023136939853429794,
          "train_metric": 0.000285827845800668,
          "validation_metric": 0.00023302678891923279,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0799405810012104
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0019474297550914343,
          "train_loss": 0.0019107358530163765,
          "validation_loss": 0.0016152081079781055,
          "train_metric": 0.00019244232680648565,
          "validation_metric": 0.0001626778393983841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.097863461000088
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0022119953027868178,
          "train_loss": 0.0021867691539227962,
          "validation_loss": 0.0018229717388749123,
          "train_metric": 0.00022024339705239981,
          "validation_metric": 0.00018360304238740355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1152005960011593
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0033263006880588364,
          "train_loss": 0.002485361648723483,
          "validation_loss": 0.00171163328923285,
          "train_metric": 0.00025031648692674935,
          "validation_metric": 0.00017238942382391542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1326781650004705
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0034498853856348433,
          "train_loss": 0.001908981241285801,
          "validation_loss": 0.0014071702025830746,
          "train_metric": 0.00019226560834795237,
          "validation_metric": 0.00014172504597809166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1499921920003544
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0020380226960696746,
          "train_loss": 0.0015811113407835364,
          "validation_loss": 0.0011952950153499842,
          "train_metric": 0.0001592437329236418,
          "validation_metric": 0.00012038574641337618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1682431370009
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0022952716681174934,
          "train_loss": 0.003408050164580345,
          "validation_loss": 0.0022343373857438564,
          "train_metric": 0.00034324638545513153,
          "validation_metric": 0.00022503428044728935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.185933398999623
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0038854503436596133,
          "train_loss": 0.0022654407657682896,
          "validation_loss": 0.001783787622116506,
          "train_metric": 0.0002281668857904151,
          "validation_metric": 0.00017965656297747046,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.203172061999794
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.003199253193088225,
          "train_loss": 0.0022154084872454405,
          "validation_loss": 0.0018022906733676791,
          "train_metric": 0.00022312781948130578,
          "validation_metric": 0.0001815201248973608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2201593080008024
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0019947820983361453,
          "train_loss": 0.0016443177592009306,
          "validation_loss": 0.0013611753238365054,
          "train_metric": 0.0001656096865190193,
          "validation_metric": 0.00013709258928429335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.237890813999911
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0017390602733939886,
          "train_loss": 0.0015646195970475674,
          "validation_loss": 0.00128281326033175,
          "train_metric": 0.00015758276276756078,
          "validation_metric": 0.00012920027074869722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2563299060002464
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.001830663579312386,
          "train_loss": 0.0021336607169359922,
          "validation_loss": 0.0015130754327401519,
          "train_metric": 0.00021489450591616333,
          "validation_metric": 0.00015239142521750182,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.274028224999711
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0019948674298575497,
          "train_loss": 0.002134566893801093,
          "validation_loss": 0.001876533729955554,
          "train_metric": 0.00021498576097656041,
          "validation_metric": 0.00018899759743362665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2918302710004355
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0027030309138353914,
          "train_loss": 0.005582877434790134,
          "validation_loss": 0.0037740753032267094,
          "train_metric": 0.0005622869357466698,
          "validation_metric": 0.0003801110142376274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.309315107000657
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.003215984492271673,
          "train_loss": 0.0031875513959676027,
          "validation_loss": 0.0028571104630827904,
          "train_metric": 0.0003210384747944772,
          "validation_metric": 0.000287757720798254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.326927613999942
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.002301576689205831,
          "train_loss": 0.002005688613280654,
          "validation_loss": 0.0015455991961061954,
          "train_metric": 0.00020200562721583992,
          "validation_metric": 0.00015566709043923765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.344413552000333
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0018091240654030116,
          "train_loss": 0.0014341870555654168,
          "validation_loss": 0.0011986740864813328,
          "train_metric": 0.0001444460649508983,
          "validation_metric": 0.000120726072054822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3615006520012685
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.00216110868495889,
          "train_loss": 0.001770049799233675,
          "validation_loss": 0.0012087645009160042,
          "train_metric": 0.00017827290866989642,
          "validation_metric": 0.00012174232688266784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3785033359999943
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.003269924935011659,
          "train_loss": 0.010835031047463417,
          "validation_loss": 0.0083487993106246,
          "train_metric": 0.0010912646539509296,
          "validation_metric": 0.0008408605353906751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3952819670012104
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0063757618991076015,
          "train_loss": 0.006677553057670593,
          "validation_loss": 0.0051550548523664474,
          "train_metric": 0.0006725386483594775,
          "validation_metric": 0.0005191981908865273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4123756749995664
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.0037992439574736636,
          "train_loss": 0.005839404184371233,
          "validation_loss": 0.004704599268734455,
          "train_metric": 0.0005881233373656869,
          "validation_metric": 0.00047383003402501345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.42966314899968
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.004972756600182038,
          "train_loss": 0.0029713630210608244,
          "validation_loss": 0.0019071276765316725,
          "train_metric": 0.0002992648514918983,
          "validation_metric": 0.00019207890727557242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4468528510005854
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0022777219201088883,
          "train_loss": 0.002162759890779853,
          "validation_loss": 0.001664292300119996,
          "train_metric": 0.00021782527619507164,
          "validation_metric": 0.0001676214305916801,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4644046440007514
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.002664235320480657,
          "train_loss": 0.0019585283007472754,
          "validation_loss": 0.001605831552296877,
          "train_metric": 0.00019725578022189438,
          "validation_metric": 0.000161733478307724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.481944442000895
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0014363633126777131,
          "train_loss": 0.0020091321785002947,
          "validation_loss": 0.0014007051941007376,
          "train_metric": 0.0002023524430114776,
          "validation_metric": 0.00014107390597928315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4994309510002495
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.002494968639439321,
          "train_loss": 0.001577925169840455,
          "validation_loss": 0.0013359715230762959,
          "train_metric": 0.00015892284864094108,
          "validation_metric": 0.00013455416774377227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5169326289997116
        }
      ],
      "best": {
        "epoch": 213,
        "updates": 3408
      },
      "best_validation_metric": 0.00012038574641337618,
      "config": {
        "label": "Concatenated skip",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": true,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2243,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010972043857612681
    },
    "c160e9485e77_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.4904455444775522,
          "train_loss": 0.13934959471225739,
          "validation_loss": 0.11717458069324493,
          "train_metric": 0.01403478067368269,
          "validation_metric": 0.011801394633948803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.019819928998913383
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.14093744941055775,
          "train_loss": 0.10362391173839569,
          "validation_loss": 0.08495587110519409,
          "train_metric": 0.010436620563268661,
          "validation_metric": 0.008556444197893143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03821827199863037
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.10152645176276565,
          "train_loss": 0.08222483098506927,
          "validation_loss": 0.06578671187162399,
          "train_metric": 0.008281383663415909,
          "validation_metric": 0.006625796668231487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05806985799972608
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.07833745097741485,
          "train_loss": 0.06921931356191635,
          "validation_loss": 0.0547175332903862,
          "train_metric": 0.00697151618078351,
          "validation_metric": 0.005510949995368719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07753075799882936
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.06660907529294491,
          "train_loss": 0.06183292716741562,
          "validation_loss": 0.04943029582500458,
          "train_metric": 0.006227585487067699,
          "validation_metric": 0.0049784379079937935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0957501829998364
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06221078592352569,
          "train_loss": 0.05955684930086136,
          "validation_loss": 0.04691355675458908,
          "train_metric": 0.005998347420245409,
          "validation_metric": 0.004724961705505848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11372048599878326
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05784499691799283,
          "train_loss": 0.05652565509080887,
          "validation_loss": 0.04484263062477112,
          "train_metric": 0.005693057086318731,
          "validation_metric": 0.004516384564340115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13172842699896137
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.05620387801900506,
          "train_loss": 0.04815591126680374,
          "validation_loss": 0.03781985118985176,
          "train_metric": 0.004850087221711874,
          "validation_metric": 0.0038090769667178392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1572287149992917
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04935957072302699,
          "train_loss": 0.045613013207912445,
          "validation_loss": 0.03637992963194847,
          "train_metric": 0.004593975841999054,
          "validation_metric": 0.0036640530452132225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17640882299929217
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.04284394963178784,
          "train_loss": 0.04061843082308769,
          "validation_loss": 0.03177601471543312,
          "train_metric": 0.004090939648449421,
          "validation_metric": 0.003200363600626588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.194032260998938
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04002338938880712,
          "train_loss": 0.03796352818608284,
          "validation_loss": 0.0293949656188488,
          "train_metric": 0.0038235478568822145,
          "validation_metric": 0.002960553392767906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21604170799946587
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.035883073578588665,
          "train_loss": 0.03283192589879036,
          "validation_loss": 0.025566520169377327,
          "train_metric": 0.0033067113254219294,
          "validation_metric": 0.002574966521933675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2347123189992999
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.031996352947317064,
          "train_loss": 0.030434872955083847,
          "validation_loss": 0.02382774092257023,
          "train_metric": 0.0030652887653559446,
          "validation_metric": 0.0023998429533094168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25637195099989185
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.02924633794464171,
          "train_loss": 0.0280045997351408,
          "validation_loss": 0.022035785019397736,
          "train_metric": 0.002820520894601941,
          "validation_metric": 0.0022193635813891888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2839006049998716
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.030054271162953228,
          "train_loss": 0.0365850105881691,
          "validation_loss": 0.0302193071693182,
          "train_metric": 0.003684708382934332,
          "validation_metric": 0.003043578239157796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3132111269987945
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.029217209026683122,
          "train_loss": 0.027735229581594467,
          "validation_loss": 0.022455986589193344,
          "train_metric": 0.002793390303850174,
          "validation_metric": 0.0022616845089942217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33067541499985964
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.025231220351997763,
          "train_loss": 0.023457027971744537,
          "validation_loss": 0.018747622147202492,
          "train_metric": 0.0023625059984624386,
          "validation_metric": 0.0018881921423599124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3499869369989028
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.025014831218868494,
          "train_loss": 0.030795931816101074,
          "validation_loss": 0.024499820545315742,
          "train_metric": 0.003101653652265668,
          "validation_metric": 0.0024675324093550444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3687792279997666
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.025055271456949413,
          "train_loss": 0.021437613293528557,
          "validation_loss": 0.016873743385076523,
          "train_metric": 0.002159117953851819,
          "validation_metric": 0.0016994616016745567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3874002599986852
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.021902769833104685,
          "train_loss": 0.019920842722058296,
          "validation_loss": 0.01573348604142666,
          "train_metric": 0.0020063542760908604,
          "validation_metric": 0.0015846191672608256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40770147599869233
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.020822126738494262,
          "train_loss": 0.01873175799846649,
          "validation_loss": 0.014964815229177475,
          "train_metric": 0.0018865938764065504,
          "validation_metric": 0.0015072014648467302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4276930559990433
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.021016723301727325,
          "train_loss": 0.01799447275698185,
          "validation_loss": 0.014211496338248253,
          "train_metric": 0.0018123373156413436,
          "validation_metric": 0.0014313298743218184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44675838799958
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.019918115460313857,
          "train_loss": 0.019539963454008102,
          "validation_loss": 0.01614227145910263,
          "train_metric": 0.0019679937977343798,
          "validation_metric": 0.0016257907263934612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4651273259987647
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.017020551662426442,
          "train_loss": 0.015961982309818268,
          "validation_loss": 0.012622656300663948,
          "train_metric": 0.0016076323809102178,
          "validation_metric": 0.0012713078176602721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48510548999911407
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.01761132344836369,
          "train_loss": 0.016268903389573097,
          "validation_loss": 0.012611706741154194,
          "train_metric": 0.0016385444905608892,
          "validation_metric": 0.0012702050153166056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5053440430001501
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.017567875154782087,
          "train_loss": 0.019173938781023026,
          "validation_loss": 0.01651631109416485,
          "train_metric": 0.0019311292562633753,
          "validation_metric": 0.0016634624917060137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5245800119992055
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01763389733969234,
          "train_loss": 0.01505590695887804,
          "validation_loss": 0.012550221756100655,
          "train_metric": 0.0015163758071139455,
          "validation_metric": 0.0012640124186873436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5443748099987715
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.016971300821751356,
          "train_loss": 0.015415839850902557,
          "validation_loss": 0.012932023964822292,
          "train_metric": 0.0015526267234236002,
          "validation_metric": 0.001302465912885964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5629153649988439
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016253514273557812,
          "train_loss": 0.020640984177589417,
          "validation_loss": 0.018503401428461075,
          "train_metric": 0.0020788847468793392,
          "validation_metric": 0.0018635950982570648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.582175785999425
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.016631856065941975,
          "train_loss": 0.01813647709786892,
          "validation_loss": 0.015576912090182304,
          "train_metric": 0.0018266397528350353,
          "validation_metric": 0.001568849547766149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6010744979994342
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.022961813519941643,
          "train_loss": 0.019094768911600113,
          "validation_loss": 0.01592588797211647,
          "train_metric": 0.0019231552723795176,
          "validation_metric": 0.00160399719607085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6197713900000963
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.020093899918720126,
          "train_loss": 0.013179982081055641,
          "validation_loss": 0.010243618860840797,
          "train_metric": 0.0013274394441395998,
          "validation_metric": 0.001031699706800282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6391327189994627
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01634510757867247,
          "train_loss": 0.011862652376294136,
          "validation_loss": 0.009460332803428173,
          "train_metric": 0.001194762997329235,
          "validation_metric": 0.0009528100490570068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6582605599996896
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013148405298124999,
          "train_loss": 0.014509058557450771,
          "validation_loss": 0.009694297797977924,
          "train_metric": 0.0014612991362810135,
          "validation_metric": 0.0009763741982169449,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6771056219986349
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.013423388299997896,
          "train_loss": 0.015370291657745838,
          "validation_loss": 0.01403712946921587,
          "train_metric": 0.0015480393776670098,
          "validation_metric": 0.0014137682737782598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6954153709993989
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.017114396585384384,
          "train_loss": 0.021662453189492226,
          "validation_loss": 0.01501722913235426,
          "train_metric": 0.0021817630622535944,
          "validation_metric": 0.0015124803176149726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7138013080002565
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.012511134380474687,
          "train_loss": 0.011520547792315483,
          "validation_loss": 0.010028948076069355,
          "train_metric": 0.0011603075545281172,
          "validation_metric": 0.00101007882039994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8337562779997825
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.0122513668902684,
          "train_loss": 0.010457387194037437,
          "validation_loss": 0.008981993421912193,
          "train_metric": 0.0010532296728342772,
          "validation_metric": 0.0009046336635947227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8543964059990685
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.010958608007058501,
          "train_loss": 0.010309784673154354,
          "validation_loss": 0.008481208235025406,
          "train_metric": 0.001038363785482943,
          "validation_metric": 0.000854196201544255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8745122549989901
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.01229957677423954,
          "train_loss": 0.011590210720896721,
          "validation_loss": 0.008584142662584782,
          "train_metric": 0.001167323556728661,
          "validation_metric": 0.00086456339340657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8948866139999154
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.01071275441790931,
          "train_loss": 0.009134098887443542,
          "validation_loss": 0.007157585117965937,
          "train_metric": 0.0009199529886245728,
          "validation_metric": 0.000720885640475899,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9141739549995691
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.010277624693117104,
          "train_loss": 0.009214832447469234,
          "validation_loss": 0.0072525241412222385,
          "train_metric": 0.0009280841914005578,
          "validation_metric": 0.0007304476457647979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9329751089990168
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.010706025146646425,
          "train_loss": 0.009006442502140999,
          "validation_loss": 0.006590125150978565,
          "train_metric": 0.0009070959640666842,
          "validation_metric": 0.0006637332844547927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9618213969988574
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01079377208952792,
          "train_loss": 0.010856346227228642,
          "validation_loss": 0.008697446435689926,
          "train_metric": 0.001093411585316062,
          "validation_metric": 0.0008759750053286552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9875263639987679
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.01509905670536682,
          "train_loss": 0.009674359112977982,
          "validation_loss": 0.0070261117070913315,
          "train_metric": 0.0009743660921230912,
          "validation_metric": 0.0007076442707329988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0120018549987435
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.014044214825844392,
          "train_loss": 0.009077363647520542,
          "validation_loss": 0.006410509347915649,
          "train_metric": 0.0009142388007603586,
          "validation_metric": 0.0006456431001424789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0429660930003593
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.011401034134905785,
          "train_loss": 0.013197113759815693,
          "validation_loss": 0.012916880659759045,
          "train_metric": 0.0013291650684550405,
          "validation_metric": 0.001300940872170031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0688894229988364
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.008963527165178675,
          "train_loss": 0.010732848197221756,
          "validation_loss": 0.010596615262329578,
          "train_metric": 0.0010809731902554631,
          "validation_metric": 0.0010672523640096188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0864697839988366
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.010197097464697435,
          "train_loss": 0.009429388679564,
          "validation_loss": 0.0076580834574997425,
          "train_metric": 0.0009496933780610561,
          "validation_metric": 0.0007712941151112318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.10398155799885
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009994707324949559,
          "train_loss": 0.0071529545821249485,
          "validation_loss": 0.00552271818742156,
          "train_metric": 0.0007204193971119821,
          "validation_metric": 0.0005562279256992042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1215975349987275
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008956704958109185,
          "train_loss": 0.011344822123646736,
          "validation_loss": 0.011179435066878796,
          "train_metric": 0.0011426089331507683,
          "validation_metric": 0.0011259517632424831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1389723920001416
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.00791679871326778,
          "train_loss": 0.008332045748829842,
          "validation_loss": 0.005514451302587986,
          "train_metric": 0.0008391730370931327,
          "validation_metric": 0.000555395323317498,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.15669376599908
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008290500540169887,
          "train_loss": 0.010323784314095974,
          "validation_loss": 0.007080139126628637,
          "train_metric": 0.001039773807860911,
          "validation_metric": 0.0007130856392905116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1736596319988166
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.008166912943124771,
          "train_loss": 0.01167335081845522,
          "validation_loss": 0.012087585404515266,
          "train_metric": 0.0011756970779970288,
          "validation_metric": 0.0012174174189567566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1904806249985995
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.01413898024475202,
          "train_loss": 0.012262470088899136,
          "validation_loss": 0.009281841106712818,
          "train_metric": 0.0012350310571491718,
          "validation_metric": 0.0009348329622298479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2076103979998152
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.012061389585142024,
          "train_loss": 0.006606051232665777,
          "validation_loss": 0.00462497491389513,
          "train_metric": 0.0006653373129665852,
          "validation_metric": 0.0004658105899579823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2251736620000884
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.007349479870754294,
          "train_loss": 0.006594234146177769,
          "validation_loss": 0.005790570750832558,
          "train_metric": 0.0006641470827162266,
          "validation_metric": 0.0005832050810568035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.242102086000159
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0065441620026831515,
          "train_loss": 0.005880288779735565,
          "validation_loss": 0.0048261163756251335,
          "train_metric": 0.0005922411801293492,
          "validation_metric": 0.00048606883501634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2589119059994118
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.0066380145799485035,
          "train_loss": 0.005898048169910908,
          "validation_loss": 0.005274666473269463,
          "train_metric": 0.0005940297851338983,
          "validation_metric": 0.0005312450230121613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2760250849987642
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.006321337568806484,
          "train_loss": 0.005584442522376776,
          "validation_loss": 0.00483600702136755,
          "train_metric": 0.0005624445620924234,
          "validation_metric": 0.00048706488450989127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.293398375999459
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.008528440012014471,
          "train_loss": 0.0057684145867824554,
          "validation_loss": 0.004623442888259888,
          "train_metric": 0.0005809736321680248,
          "validation_metric": 0.0004656562814489007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3122126460002619
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.00774960402486613,
          "train_loss": 0.00589545164257288,
          "validation_loss": 0.005343218334019184,
          "train_metric": 0.000593768258113414,
          "validation_metric": 0.0005381493829190731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3296348200001376
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006523394244140945,
          "train_loss": 0.005915304645895958,
          "validation_loss": 0.004787793383002281,
          "train_metric": 0.0005957678658887744,
          "validation_metric": 0.0004822090268135071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3464641479986312
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.006212917098309845,
          "train_loss": 0.00805151928216219,
          "validation_loss": 0.008130255155265331,
          "train_metric": 0.000810919445939362,
          "validation_metric": 0.0008188495994545519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3636856519988214
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0063179541466524825,
          "train_loss": 0.005207476671785116,
          "validation_loss": 0.004538180772215128,
          "train_metric": 0.0005244780331850052,
          "validation_metric": 0.0004570689343381673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3813691140003357
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005504364045918919,
          "train_loss": 0.006156617775559425,
          "validation_loss": 0.004406629595905542,
          "train_metric": 0.0006200720090419054,
          "validation_metric": 0.00044381959014572203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.398770202000378
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0058738759471452795,
          "train_loss": 0.00452030636370182,
          "validation_loss": 0.0035550387110561132,
          "train_metric": 0.0004552686295937747,
          "validation_metric": 0.000358050485374406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4164079309994122
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.006635569341597147,
          "train_loss": 0.006438794545829296,
          "validation_loss": 0.0039557237178087234,
          "train_metric": 0.0006484918994829059,
          "validation_metric": 0.0003984060022048652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.43386215099963
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.004754433815833181,
          "train_loss": 0.004848063457757235,
          "validation_loss": 0.004456290975213051,
          "train_metric": 0.0004882791545242071,
          "validation_metric": 0.0004488212871365249,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4515767379998579
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0052301446266938,
          "train_loss": 0.005085873883217573,
          "validation_loss": 0.0035638725385069847,
          "train_metric": 0.0005122305010445416,
          "validation_metric": 0.00035894018947146833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.469677770999624
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.006386241017025895,
          "train_loss": 0.0048872679471969604,
          "validation_loss": 0.004000648856163025,
          "train_metric": 0.0004922277294099331,
          "validation_metric": 0.00040293068741448224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4871277809997991
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.006183876030263491,
          "train_loss": 0.004212959669530392,
          "validation_loss": 0.003745495341718197,
          "train_metric": 0.0004243138537276536,
          "validation_metric": 0.0003772325289901346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5048069929998746
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.010386545312940143,
          "train_loss": 0.011510458774864674,
          "validation_loss": 0.008765856735408306,
          "train_metric": 0.0011592913651838899,
          "validation_metric": 0.0008828651043586433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.521833811999386
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.009570008638547733,
          "train_loss": 0.005513221491128206,
          "validation_loss": 0.005294816102832556,
          "train_metric": 0.0005552715738303959,
          "validation_metric": 0.0005332744913175702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5385596679989249
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00800592482846696,
          "train_loss": 0.004637943115085363,
          "validation_loss": 0.003704836592078209,
          "train_metric": 0.0004671167116612196,
          "validation_metric": 0.0003731375327333808,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.555863452000267
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.010345806564146187,
          "train_loss": 0.004818933550268412,
          "validation_loss": 0.0040273810736835,
          "train_metric": 0.000485345401102677,
          "validation_metric": 0.00040562308277003467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5730278089995409
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.006773580753360875,
          "train_loss": 0.008461461402475834,
          "validation_loss": 0.006420161575078964,
          "train_metric": 0.0008522073039785028,
          "validation_metric": 0.0006466152262873948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5898225189994264
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.008412933253566734,
          "train_loss": 0.01966680958867073,
          "validation_loss": 0.020628316327929497,
          "train_metric": 0.001980768982321024,
          "validation_metric": 0.0020776083692908287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6073734410001634
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006622763343330007,
          "train_loss": 0.005296480376273394,
          "validation_loss": 0.0032732016406953335,
          "train_metric": 0.000533442129381001,
          "validation_metric": 0.000329664908349514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6262738939985866
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.00481734248023713,
          "train_loss": 0.0037940419279038906,
          "validation_loss": 0.0033788802102208138,
          "train_metric": 0.00038212203071452677,
          "validation_metric": 0.00034030844108201563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.643115630999091
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005591738052316941,
          "train_loss": 0.0038845324888825417,
          "validation_loss": 0.002849757671356201,
          "train_metric": 0.00039123589522205293,
          "validation_metric": 0.00028701714472845197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.660615548998976
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.00399681778071681,
          "train_loss": 0.003882030490785837,
          "validation_loss": 0.00281687849201262,
          "train_metric": 0.000390983943361789,
          "validation_metric": 0.0002837057108990848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6779340009998123
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.004303078592784004,
          "train_loss": 0.006650542374700308,
          "validation_loss": 0.0074018314480781555,
          "train_metric": 0.0006698181969113648,
          "validation_metric": 0.0007454853039234877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6953292460002558
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.00619044776976807,
          "train_loss": 0.011541551910340786,
          "validation_loss": 0.007090327329933643,
          "train_metric": 0.0011624228209257126,
          "validation_metric": 0.000714111840352416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.712221239999053
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.006389095302438363,
          "train_loss": 0.0053476328030228615,
          "validation_loss": 0.005610426422208548,
          "train_metric": 0.0005385939730331302,
          "validation_metric": 0.0005650616367347538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7293934249992162
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.007835877622710541,
          "train_loss": 0.0072725145146250725,
          "validation_loss": 0.004942962434142828,
          "train_metric": 0.0007324611651711166,
          "validation_metric": 0.0004978370270691812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7471246859986422
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.005457720857521053,
          "train_loss": 0.003692010184749961,
          "validation_loss": 0.0032752929255366325,
          "train_metric": 0.00037184575921855867,
          "validation_metric": 0.00032987550366669893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.764673423998829
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.004270849007298239,
          "train_loss": 0.005358549766242504,
          "validation_loss": 0.0029968852177262306,
          "train_metric": 0.0005396935157477856,
          "validation_metric": 0.0003018353309016675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.781593237999914
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004601326523697935,
          "train_loss": 0.004343017935752869,
          "validation_loss": 0.004557962529361248,
          "train_metric": 0.0004374128475319594,
          "validation_metric": 0.00045906126615591347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7989263279996521
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.003860376167722279,
          "train_loss": 0.0036365212872624397,
          "validation_loss": 0.0025321184657514095,
          "train_metric": 0.0003662570961751044,
          "validation_metric": 0.0002550257195252925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8169601079989661
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.003579523388907546,
          "train_loss": 0.003188599832355976,
          "validation_loss": 0.0024696774780750275,
          "train_metric": 0.0003211440925952047,
          "validation_metric": 0.0002487369056325406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.83548465199965
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0036682048812508583,
          "train_loss": 0.003049431601539254,
          "validation_loss": 0.0023456646595150232,
          "train_metric": 0.0003071275423280895,
          "validation_metric": 0.00023624679306522012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8541473209988908
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004737983595987316,
          "train_loss": 0.0031831907108426094,
          "validation_loss": 0.0029399837367236614,
          "train_metric": 0.0003205992979928851,
          "validation_metric": 0.00029610440833494067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8717350829992938
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004149462241912261,
          "train_loss": 0.00368437054567039,
          "validation_loss": 0.0036361932288855314,
          "train_metric": 0.0003710762830451131,
          "validation_metric": 0.000366224063327536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8889889509991917
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004521022419794463,
          "train_loss": 0.002844530623406172,
          "validation_loss": 0.00218679360114038,
          "train_metric": 0.0002864906855393201,
          "validation_metric": 0.0002202458563260734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9066578669990122
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.00492143070732709,
          "train_loss": 0.004249703139066696,
          "validation_loss": 0.004382847808301449,
          "train_metric": 0.00042801458039321005,
          "validation_metric": 0.0004414243740029633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9238509650003834
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.006722190537402639,
          "train_loss": 0.0033907657489180565,
          "validation_loss": 0.0030881226994097233,
          "train_metric": 0.00034150551073253155,
          "validation_metric": 0.0003110244288109243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9412339059999795
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004005443912319606,
          "train_loss": 0.003039856208488345,
          "validation_loss": 0.00197694287635386,
          "train_metric": 0.00030616321600973606,
          "validation_metric": 0.0001991104509215802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9591730920001282
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0031582197261741385,
          "train_loss": 0.003130811732262373,
          "validation_loss": 0.0020158407278358936,
          "train_metric": 0.00031532393768429756,
          "validation_metric": 0.0002030281029874459,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9776433309998538
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.003663124196464196,
          "train_loss": 0.007061227690428495,
          "validation_loss": 0.0043159364722669125,
          "train_metric": 0.000711181084625423,
          "validation_metric": 0.0004346853238530457,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9949187069996697
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.003852410172839882,
          "train_loss": 0.0033565638586878777,
          "validation_loss": 0.003589787520468235,
          "train_metric": 0.0003380608686711639,
          "validation_metric": 0.00036155025009065866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0121504469989304
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0030413188142119907,
          "train_loss": 0.003686188254505396,
          "validation_loss": 0.003963727504014969,
          "train_metric": 0.00037125934613868594,
          "validation_metric": 0.00039921212010085583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0290727989995503
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.00317310440732399,
          "train_loss": 0.0027985419146716595,
          "validation_loss": 0.0018981071189045906,
          "train_metric": 0.0002818588982336223,
          "validation_metric": 0.00019117041665595025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.047346370998639
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0035875934845535085,
          "train_loss": 0.0032164694275707006,
          "validation_loss": 0.0018945119809359312,
          "train_metric": 0.0003239510697312653,
          "validation_metric": 0.00019080829224549234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.065970613999525
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003647388572062482,
          "train_loss": 0.00645049661397934,
          "validation_loss": 0.0070405681617558,
          "train_metric": 0.0006496704299934208,
          "validation_metric": 0.0007091001607477665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.083719176998784
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.004685109637648566,
          "train_loss": 0.005963361356407404,
          "validation_loss": 0.004336930811405182,
          "train_metric": 0.000600607949309051,
          "validation_metric": 0.0004367997753433883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1012944009999046
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0037433152974699624,
          "train_loss": 0.002743094228208065,
          "validation_loss": 0.0025983399245887995,
          "train_metric": 0.0002762744261417538,
          "validation_metric": 0.00026169532793574035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2144565809994674
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0029381671229202766,
          "train_loss": 0.003608223283663392,
          "validation_loss": 0.0039051538333296776,
          "train_metric": 0.0003634070453699678,
          "validation_metric": 0.00039331280277110636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.231854580999425
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.004887055696599418,
          "train_loss": 0.004055550321936607,
          "validation_loss": 0.0024473804514855146,
          "train_metric": 0.00040846012416295707,
          "validation_metric": 0.0002464911958668381,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2486880689993995
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.005119693450978957,
          "train_loss": 0.012568050995469093,
          "validation_loss": 0.01251265499740839,
          "train_metric": 0.001265808241441846,
          "validation_metric": 0.0012602286878973246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2660004289991775
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.00807270266523119,
          "train_loss": 0.004095022100955248,
          "validation_loss": 0.002711828565225005,
          "train_metric": 0.000412435649195686,
          "validation_metric": 0.00027312542079016566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2831717349999963
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004297466053685639,
          "train_loss": 0.00515243923291564,
          "validation_loss": 0.005313125438988209,
          "train_metric": 0.0005189348012208939,
          "validation_metric": 0.0005351185100153089,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3005493029995705
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.003859263742924668,
          "train_loss": 0.003916557878255844,
          "validation_loss": 0.0027050008065998554,
          "train_metric": 0.00039446132723242044,
          "validation_metric": 0.0002724377845879644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.317666200999156
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0028639588308578823,
          "train_loss": 0.002508661476895213,
          "validation_loss": 0.002321206498891115,
          "train_metric": 0.0002526631869841367,
          "validation_metric": 0.0002337834011996165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.334868293999534
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0024528959875169676,
          "train_loss": 0.00322182709351182,
          "validation_loss": 0.0032900236546993256,
          "train_metric": 0.0003244906256441027,
          "validation_metric": 0.0003313591005280614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3525174299993523
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.003631263753050007,
          "train_loss": 0.0024984776973724365,
          "validation_loss": 0.0023069744929671288,
          "train_metric": 0.0002516375097911805,
          "validation_metric": 0.0002323500084457919,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3702368839985866
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.003072737508773571,
          "train_loss": 0.0029748373199254274,
          "validation_loss": 0.0016847255174070597,
          "train_metric": 0.00029961473774164915,
          "validation_metric": 0.00016967936244327575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.388539112000217
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.004389602330775233,
          "train_loss": 0.004248362500220537,
          "validation_loss": 0.0024658264592289925,
          "train_metric": 0.00042787950951606035,
          "validation_metric": 0.00024834900978021324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4061895129998447
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.003914701526809949,
          "train_loss": 0.005522224586457014,
          "validation_loss": 0.005985216237604618,
          "train_metric": 0.0005561781581491232,
          "validation_metric": 0.0006028090137988329,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4229538349991344
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.004959646052157041,
          "train_loss": 0.002941087819635868,
          "validation_loss": 0.0016800971934571862,
          "train_metric": 0.00029621561407111585,
          "validation_metric": 0.00016921322094276547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4484472699987236
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.004658034122257959,
          "train_loss": 0.0023091572802513838,
          "validation_loss": 0.0014317299937829375,
          "train_metric": 0.00023256988788489252,
          "validation_metric": 0.00014419859508052468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4727828839986614
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.003149490243231412,
          "train_loss": 0.004634230863302946,
          "validation_loss": 0.003292625769972801,
          "train_metric": 0.000466742756543681,
          "validation_metric": 0.0003316211805213243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4902436779993877
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.004229694859532174,
          "train_loss": 0.0032592136412858963,
          "validation_loss": 0.0034259045496582985,
          "train_metric": 0.0003282560792285949,
          "validation_metric": 0.0003450445656199008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5073385590003454
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0026186527247773483,
          "train_loss": 0.0024466263130307198,
          "validation_loss": 0.0014038749504834414,
          "train_metric": 0.00024641529307700694,
          "validation_metric": 0.00014139313134364784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.525168117999783
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0027523912467586342,
          "train_loss": 0.002808426506817341,
          "validation_loss": 0.0021690477151423693,
          "train_metric": 0.00028285442385822535,
          "validation_metric": 0.00021845853189006448,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.542239153999617
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.002243004102638224,
          "train_loss": 0.0027828305028378963,
          "validation_loss": 0.0029413970187306404,
          "train_metric": 0.00028027649386785924,
          "validation_metric": 0.0002962467260658741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.558825906999118
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.002406240319032804,
          "train_loss": 0.002711806446313858,
          "validation_loss": 0.0016163700493052602,
          "train_metric": 0.0002731231797952205,
          "validation_metric": 0.00016279486590065062,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.575851886998862
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0028406734090822283,
          "train_loss": 0.0035543811973184347,
          "validation_loss": 0.003748746123164892,
          "train_metric": 0.0003579842741601169,
          "validation_metric": 0.00037755994708277285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5932763740001974
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.003398874214326497,
          "train_loss": 0.0033690810669213533,
          "validation_loss": 0.001998237567022443,
          "train_metric": 0.00033932150108739734,
          "validation_metric": 0.00020125519949942827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6106114229987725
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0024290328547067475,
          "train_loss": 0.0019899788312613964,
          "validation_loss": 0.0018136983271688223,
          "train_metric": 0.0002004233538173139,
          "validation_metric": 0.00018266904226038605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6275909099986166
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0034146856087318156,
          "train_loss": 0.004198167007416487,
          "validation_loss": 0.0019356291741132736,
          "train_metric": 0.0004228239704389125,
          "validation_metric": 0.00019494944717735052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.644929119000153
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0031049733552208636,
          "train_loss": 0.0032714635599404573,
          "validation_loss": 0.003386788535863161,
          "train_metric": 0.00032948984880931675,
          "validation_metric": 0.00034110492561012506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6618355409991636
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0033541370066814125,
          "train_loss": 0.0040321359410882,
          "validation_loss": 0.0023473077453672886,
          "train_metric": 0.0004061019863002002,
          "validation_metric": 0.00023641224834136665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.679067787999884
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.003443765006522881,
          "train_loss": 0.0018467801855877042,
          "validation_loss": 0.0012399136321619153,
          "train_metric": 0.0001860009360825643,
          "validation_metric": 0.00012487957428675145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.697157343998697
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.00231674399037729,
          "train_loss": 0.0020868652500212193,
          "validation_loss": 0.0016565527766942978,
          "train_metric": 0.00021018141706008464,
          "validation_metric": 0.0001668419427005574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.714399525999397
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0023420199868269265,
          "train_loss": 0.0023540265392512083,
          "validation_loss": 0.001360827125608921,
          "train_metric": 0.0002370889560552314,
          "validation_metric": 0.00013705754827242345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7315120240000397
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0024314656548085622,
          "train_loss": 0.0023699786979705095,
          "validation_loss": 0.0017740923212841153,
          "train_metric": 0.00023869557480793446,
          "validation_metric": 0.00017868007125798613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.748942285999874
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.00472701461694669,
          "train_loss": 0.0037188176065683365,
          "validation_loss": 0.0034039246384054422,
          "train_metric": 0.00037454566336236894,
          "validation_metric": 0.00034283081186003983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.765938589000143
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.002485167064151028,
          "train_loss": 0.002888184040784836,
          "validation_loss": 0.002870743628591299,
          "train_metric": 0.00029088734299875796,
          "validation_metric": 0.0002891307813115418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7831070899992483
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0017899643826240208,
          "train_loss": 0.0020022583194077015,
          "validation_loss": 0.0011737896129488945,
          "train_metric": 0.00020166015019640326,
          "validation_metric": 0.00011821979569504037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8013676299997314
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.002547937532654032,
          "train_loss": 0.00259727262891829,
          "validation_loss": 0.0014689300442114472,
          "train_metric": 0.0002615877892822027,
          "validation_metric": 0.00014794524759054184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8194480469992413
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.002217136987383128,
          "train_loss": 0.002082356484606862,
          "validation_loss": 0.0020781406201422215,
          "train_metric": 0.000209727295441553,
          "validation_metric": 0.00020930272876285017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.838045697999405
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.002739379913691664,
          "train_loss": 0.007332931272685528,
          "validation_loss": 0.007556992582976818,
          "train_metric": 0.0007385460776276886,
          "validation_metric": 0.0007611126638948917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.855279809999047
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0036985025653848425,
          "train_loss": 0.002317548729479313,
          "validation_loss": 0.0018475604010745883,
          "train_metric": 0.0002334150776732713,
          "validation_metric": 0.0001860795309767127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8728506459992786
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.002029907871474279,
          "train_loss": 0.0020154155790805817,
          "validation_loss": 0.0020404793322086334,
          "train_metric": 0.00020298526214901358,
          "validation_metric": 0.00020550961198750883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8898828820001654
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.003925524400983704,
          "train_loss": 0.00891520082950592,
          "validation_loss": 0.006053251214325428,
          "train_metric": 0.0008979063131846488,
          "validation_metric": 0.0006096612196415663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9068101679986285
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.004498672184126917,
          "train_loss": 0.0039010322652757168,
          "validation_loss": 0.0037637916393578053,
          "train_metric": 0.00039289769483730197,
          "validation_metric": 0.0003790753544308245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9248083509992284
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0028760708282788983,
          "train_loss": 0.0025275149382650852,
          "validation_loss": 0.0010913514997810125,
          "train_metric": 0.0002545620664022863,
          "validation_metric": 0.00010991693852702156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9428648960001738
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.002225643962447066,
          "train_loss": 0.002388985129073262,
          "validation_loss": 0.001560379983857274,
          "train_metric": 0.0002406098210485652,
          "validation_metric": 0.00015715578047093004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9601393419998203
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0020446465096028987,
          "train_loss": 0.0029221847653388977,
          "validation_loss": 0.001770121743902564,
          "train_metric": 0.00029431175789795816,
          "validation_metric": 0.00017828019917942584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9770098379995034
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0038632741998299025,
          "train_loss": 0.0027961856685578823,
          "validation_loss": 0.0011716235894709826,
          "train_metric": 0.0002816215856000781,
          "validation_metric": 0.00011800164793385193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9938845750002656
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.004816876680706628,
          "train_loss": 0.0019410583190619946,
          "validation_loss": 0.002114409813657403,
          "train_metric": 0.0001954962790478021,
          "validation_metric": 0.00021295562328305095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.011248445000092
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.003910122333763866,
          "train_loss": 0.004306190647184849,
          "validation_loss": 0.002802788745611906,
          "train_metric": 0.00043370373896323144,
          "validation_metric": 0.00028228663722984493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0285375349994865
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0027195307848160155,
          "train_loss": 0.0017317576566711068,
          "validation_loss": 0.0013708341866731644,
          "train_metric": 0.0001744162873364985,
          "validation_metric": 0.0001380653993692249,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.045430975000272
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.003207453824870754,
          "train_loss": 0.002164095174521208,
          "validation_loss": 0.0021214361768215895,
          "train_metric": 0.00021795973589178175,
          "validation_metric": 0.00021366328292060643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0624067440003273
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.003918651786079863,
          "train_loss": 0.0020208247005939484,
          "validation_loss": 0.002069042529910803,
          "train_metric": 0.0002035300713032484,
          "validation_metric": 0.0002083863946609199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0795308830001886
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0021954519870632794,
          "train_loss": 0.0018800152465701103,
          "validation_loss": 0.0015246705152094364,
          "train_metric": 0.00018934826948679984,
          "validation_metric": 0.00015355921641457826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0974013630002446
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0019665814288600814,
          "train_loss": 0.0016171927563846111,
          "validation_loss": 0.0012961354805156589,
          "train_metric": 0.00016287776816170663,
          "validation_metric": 0.0001305420300923288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1154775519989926
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.00161150743952021,
          "train_loss": 0.002287121256813407,
          "validation_loss": 0.00246921693906188,
          "train_metric": 0.00023035050253383815,
          "validation_metric": 0.0002486904850229621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1326430030003394
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0017082940812542802,
          "train_loss": 0.0021731348242610693,
          "validation_loss": 0.001238305470906198,
          "train_metric": 0.00021887019101995975,
          "validation_metric": 0.00012471758236642927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1498103419999097
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.002043557837168919,
          "train_loss": 0.0015215405728667974,
          "validation_loss": 0.001011758460663259,
          "train_metric": 0.00015324399282690138,
          "validation_metric": 0.00010190060856984928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.168510988998605
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.004596661367031629,
          "train_loss": 0.0026526018045842648,
          "validation_loss": 0.001960652181878686,
          "train_metric": 0.00026716035790741444,
          "validation_metric": 0.00019746969337575138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.186673703999986
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.003794072323216824,
          "train_loss": 0.0026669783983379602,
          "validation_loss": 0.0027387577574700117,
          "train_metric": 0.00026860827347263694,
          "validation_metric": 0.0002758376649580896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2042007299987745
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.002611827978398651,
          "train_loss": 0.004849310498684645,
          "validation_loss": 0.004868631716817617,
          "train_metric": 0.0004884048248641193,
          "validation_metric": 0.0004903507069684565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.22192423100023
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.002765119177638553,
          "train_loss": 0.0019056504825130105,
          "validation_loss": 0.0015005883760750294,
          "train_metric": 0.0001919301284942776,
          "validation_metric": 0.00015113376139197499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2394485989989334
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0018815971270669252,
          "train_loss": 0.0015451423823833466,
          "validation_loss": 0.0013780237641185522,
          "train_metric": 0.0001556210918352008,
          "validation_metric": 0.00013878951722290367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.256810509999923
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.002175769366658642,
          "train_loss": 0.0013069378910586238,
          "validation_loss": 0.0011554474476724863,
          "train_metric": 0.0001316299894824624,
          "validation_metric": 0.00011637243733275682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2742301099988254
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.002257746826217044,
          "train_loss": 0.005709926597774029,
          "validation_loss": 0.0035622697323560715,
          "train_metric": 0.000575082958675921,
          "validation_metric": 0.0003587787796277553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.291716288000316
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.006059623003238812,
          "train_loss": 0.003841787576675415,
          "validation_loss": 0.004132368601858616,
          "train_metric": 0.0003869307693094015,
          "validation_metric": 0.0004161970573477447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3097569669989753
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.00384208435571054,
          "train_loss": 0.0026160096749663353,
          "validation_loss": 0.0026538046076893806,
          "train_metric": 0.00026347493985667825,
          "validation_metric": 0.0002672814589459449,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3273220539995236
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.002404776503681205,
          "train_loss": 0.0020878647919744253,
          "validation_loss": 0.0016682478599250317,
          "train_metric": 0.0002102821017615497,
          "validation_metric": 0.00016801980382297188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3451779219994933
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.002175517573050456,
          "train_loss": 0.002940001431852579,
          "validation_loss": 0.002057925332337618,
          "train_metric": 0.0002961061545647681,
          "validation_metric": 0.00020726671209558845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3625187409998034
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0029844017735740636,
          "train_loss": 0.0019907380919903517,
          "validation_loss": 0.0009516011923551559,
          "train_metric": 0.00020049986778758466,
          "validation_metric": 9.584178769728169e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3806649029993423
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0019968493506894447,
          "train_loss": 0.001351445447653532,
          "validation_loss": 0.0012597269378602505,
          "train_metric": 0.0001361126487608999,
          "validation_metric": 0.0001268750784220174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3982608480000636
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.003029394614713965,
          "train_loss": 0.006292493548244238,
          "validation_loss": 0.0040296227671206,
          "train_metric": 0.0006337568629533052,
          "validation_metric": 0.00040584884118288755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4151023539998278
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.002784796211926732,
          "train_loss": 0.00174218718893826,
          "validation_loss": 0.0016326381592079997,
          "train_metric": 0.00017546670278534293,
          "validation_metric": 0.00016443332424387336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.431739525000012
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0018773191877698991,
          "train_loss": 0.005995539948344231,
          "validation_loss": 0.0033505475148558617,
          "train_metric": 0.0006038487190380692,
          "validation_metric": 0.00033745483960956335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4491166129992052
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0024753718862484675,
          "train_loss": 0.0012533690314739943,
          "validation_loss": 0.0009302046964876354,
          "train_metric": 0.00012623473594430834,
          "validation_metric": 9.368680184707046e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.46765534899896
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0028552782869155635,
          "train_loss": 0.0015949177322909236,
          "validation_loss": 0.0009652096778154373,
          "train_metric": 0.0001606342993909493,
          "validation_metric": 9.721238893689588e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4856389429987757
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0014103056983003626,
          "train_loss": 0.0016435282304883003,
          "validation_loss": 0.001664471346884966,
          "train_metric": 0.00016553013119846582,
          "validation_metric": 0.0001676394895184785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5030763630002184
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0013538160601456184,
          "train_loss": 0.0016003268538042903,
          "validation_loss": 0.0014328963588923216,
          "train_metric": 0.00016117906488943845,
          "validation_metric": 0.0001443160726921633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.520379468000101
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0017558653289597714,
          "train_loss": 0.0014195495750755072,
          "validation_loss": 0.0010017710737884045,
          "train_metric": 0.0001429718395229429,
          "validation_metric": 0.00010089471470564604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.53748032600015
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0021536759777518455,
          "train_loss": 0.0011931414483115077,
          "validation_loss": 0.0010134619660675526,
          "train_metric": 0.00012016883556498215,
          "validation_metric": 0.00010207218292634934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5543033049998485
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0014799896325712325,
          "train_loss": 0.0014199354918673635,
          "validation_loss": 0.0014631298836320639,
          "train_metric": 0.00014301069313660264,
          "validation_metric": 0.0001473610900575295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5714047839992418
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.001443434955945122,
          "train_loss": 0.001995242666453123,
          "validation_loss": 0.0021557346917688847,
          "train_metric": 0.00020095353829674423,
          "validation_metric": 0.0002171176893170923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5887450970003556
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0016731325522414409,
          "train_loss": 0.001615283195860684,
          "validation_loss": 0.00159472378436476,
          "train_metric": 0.00016268539184238762,
          "validation_metric": 0.00016061474161688238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6057653759999084
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.003032748390978668,
          "train_loss": 0.003972040489315987,
          "validation_loss": 0.004155698232352734,
          "train_metric": 0.0004000493499916047,
          "validation_metric": 0.0004185466968920082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.622673672000019
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.002420920121949166,
          "train_loss": 0.0013684856239706278,
          "validation_loss": 0.0013153471518307924,
          "train_metric": 0.00013782887253910303,
          "validation_metric": 0.00013247696915641427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.735682349999479
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0028095110628783004,
          "train_loss": 0.003099853405728936,
          "validation_loss": 0.0023865096736699343,
          "train_metric": 0.0003122058988083154,
          "validation_metric": 0.0002403605612926185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.836641869998857
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0016748273556004278,
          "train_loss": 0.00303438026458025,
          "validation_loss": 0.0033034570515155792,
          "train_metric": 0.00030561169842258096,
          "validation_metric": 0.00033271205029450357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9178402719990117
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.001744826524372911,
          "train_loss": 0.001670488971285522,
          "validation_loss": 0.0010841622715815902,
          "train_metric": 0.0001682455331319943,
          "validation_metric": 0.00010919285705313087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.999289901999873
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.0014571568626706721,
          "train_loss": 0.0023368988186120987,
          "validation_loss": 0.0013589492300525308,
          "train_metric": 0.00023536384105682373,
          "validation_metric": 0.00013686837337445468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.109761181000067
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0015762245820951648,
          "train_loss": 0.0012190386187285185,
          "validation_loss": 0.0009693956235423684,
          "train_metric": 0.00012277709902264178,
          "validation_metric": 9.763397974893451e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.188404225998966
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0020300281612435356,
          "train_loss": 0.0029402896761894226,
          "validation_loss": 0.0023377640172839165,
          "train_metric": 0.00029613522929139435,
          "validation_metric": 0.0002354510797886178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.257413088998874
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0034153245160268852,
          "train_loss": 0.004209646489471197,
          "validation_loss": 0.004733462352305651,
          "train_metric": 0.00042398020741529763,
          "validation_metric": 0.0004767369828186929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.323668285998792
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.002427854989946354,
          "train_loss": 0.0017985539743676782,
          "validation_loss": 0.001044916221871972,
          "train_metric": 0.00018114375416189432,
          "validation_metric": 0.00010524014214752242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.39032137599861
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.002642508454300696,
          "train_loss": 0.0020244570914655924,
          "validation_loss": 0.0019404473714530468,
          "train_metric": 0.00020389592100400478,
          "validation_metric": 0.00019543473899830133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.46078117199977
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0017912469174916623,
          "train_loss": 0.0015998497838154435,
          "validation_loss": 0.0014276173897087574,
          "train_metric": 0.00016113099991343915,
          "validation_metric": 0.00014378438936546445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.525971125000069
        }
      ],
      "best": {
        "epoch": 178,
        "updates": 2848
      },
      "best_validation_metric": 9.368680184707046e-05,
      "config": {
        "label": "Residual",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": true,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009679194276750026
    },
    "5bc849f25b25_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3057408444583416,
          "train_loss": 0.18185120820999146,
          "validation_loss": 0.14306208491325378,
          "train_metric": 0.018315389752388,
          "validation_metric": 0.014408689923584461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.015337101000113762
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.17782066157087684,
          "train_loss": 0.15651699900627136,
          "validation_loss": 0.1210668683052063,
          "train_metric": 0.015763819217681885,
          "validation_metric": 0.012193411588668823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02758567800083256
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13965064752846956,
          "train_loss": 0.12703154981136322,
          "validation_loss": 0.0908212810754776,
          "train_metric": 0.012794151902198792,
          "validation_metric": 0.009147187694907188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04035183800078812
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.11224897252395749,
          "train_loss": 0.09400814771652222,
          "validation_loss": 0.07732841372489929,
          "train_metric": 0.009468155913054943,
          "validation_metric": 0.007788234390318394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05258896099985577
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08333829755429178,
          "train_loss": 0.07792487740516663,
          "validation_loss": 0.06564121693372726,
          "train_metric": 0.00784830842167139,
          "validation_metric": 0.006611143704503775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06452740399981849
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.06732771103270352,
          "train_loss": 0.05962647870182991,
          "validation_loss": 0.05468873679637909,
          "train_metric": 0.0060053602792322636,
          "validation_metric": 0.005508049391210079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0772076230005041
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06030448339879513,
          "train_loss": 0.05254873260855675,
          "validation_loss": 0.04422566667199135,
          "train_metric": 0.0052925157360732555,
          "validation_metric": 0.0044542476534843445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08961183599967626
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.05208336468786001,
          "train_loss": 0.051716726273298264,
          "validation_loss": 0.043897129595279694,
          "train_metric": 0.0052087195217609406,
          "validation_metric": 0.004421158228069544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10209760600082518
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04792236024513841,
          "train_loss": 0.04666754603385925,
          "validation_loss": 0.04033757746219635,
          "train_metric": 0.004700183402746916,
          "validation_metric": 0.004062652587890625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1143921010007034
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.04577650607097894,
          "train_loss": 0.04491885006427765,
          "validation_loss": 0.03882687911391258,
          "train_metric": 0.004524062387645245,
          "validation_metric": 0.003910501021891832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12710436500128708
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.03988124500028789,
          "train_loss": 0.035419266670942307,
          "validation_loss": 0.030092889443039894,
          "train_metric": 0.003567298874258995,
          "validation_metric": 0.003030845895409584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13944224800070515
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.035342264105565846,
          "train_loss": 0.03380543738603592,
          "validation_loss": 0.027796130627393723,
          "train_metric": 0.0034047598019242287,
          "validation_metric": 0.002799524459987879,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15177151399984723
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.031071249628439546,
          "train_loss": 0.028387706726789474,
          "validation_loss": 0.023825038224458694,
          "train_metric": 0.0028591062873601913,
          "validation_metric": 0.0023995705414563417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16436487600003602
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.028705363220069557,
          "train_loss": 0.0284382663667202,
          "validation_loss": 0.023395106196403503,
          "train_metric": 0.002864197827875614,
          "validation_metric": 0.002356269396841526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1773063180007739
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.028297286247834563,
          "train_loss": 0.025952495634555817,
          "validation_loss": 0.020937001332640648,
          "train_metric": 0.002613840391859412,
          "validation_metric": 0.0021086980123072863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19002850700053386
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.026325745624490082,
          "train_loss": 0.02542385645210743,
          "validation_loss": 0.022547651082277298,
          "train_metric": 0.0025605978444218636,
          "validation_metric": 0.0022709169425070286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20209134400101902
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.024779850558843464,
          "train_loss": 0.022923527285456657,
          "validation_loss": 0.018705839291214943,
          "train_metric": 0.002308773575350642,
          "validation_metric": 0.0018839837284758687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21429391099991335
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.024960207869298756,
          "train_loss": 0.02051803655922413,
          "validation_loss": 0.014279721304774284,
          "train_metric": 0.002066501649096608,
          "validation_metric": 0.0014382012886926532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22677877000023727
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.020603418059181422,
          "train_loss": 0.02013441175222397,
          "validation_loss": 0.01693541184067726,
          "train_metric": 0.0020278641022741795,
          "validation_metric": 0.0017056724755093455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.238372640000307
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.017503823881270364,
          "train_loss": 0.016560573130846024,
          "validation_loss": 0.013192232698202133,
          "train_metric": 0.0016679201507940888,
          "validation_metric": 0.0013286733301356435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2502121680008713
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.01637994049815461,
          "train_loss": 0.014604705385863781,
          "validation_loss": 0.011832764372229576,
          "train_metric": 0.0014709322713315487,
          "validation_metric": 0.0011917526135221124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2623988699997426
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.014717741025378928,
          "train_loss": 0.013816207647323608,
          "validation_loss": 0.00999525561928749,
          "train_metric": 0.0013915178133174777,
          "validation_metric": 0.0010066855465993285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.274868795000657
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.013663370249560103,
          "train_loss": 0.014533179812133312,
          "validation_loss": 0.011753438040614128,
          "train_metric": 0.001463728491216898,
          "validation_metric": 0.0011837632628157735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2865172599995276
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.016000006697140634,
          "train_loss": 0.017465751618146896,
          "validation_loss": 0.01469956710934639,
          "train_metric": 0.0017590865027159452,
          "validation_metric": 0.0014804864767938852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29831661800017173
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.020609272993169725,
          "train_loss": 0.021301569417119026,
          "validation_loss": 0.016968676820397377,
          "train_metric": 0.0021454161033034325,
          "validation_metric": 0.0017090230248868465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3099644760004594
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.01341663976199925,
          "train_loss": 0.010900997556746006,
          "validation_loss": 0.008084922097623348,
          "train_metric": 0.0010979085927829146,
          "validation_metric": 0.0008142837323248386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3218376340009854
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.013192492217058316,
          "train_loss": 0.011644205078482628,
          "validation_loss": 0.008945869281888008,
          "train_metric": 0.0011727615492418408,
          "validation_metric": 0.0009009951609186828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3336762800008728
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.012018522626021877,
          "train_loss": 0.010243506170809269,
          "validation_loss": 0.007861747406423092,
          "train_metric": 0.0010316884145140648,
          "validation_metric": 0.0007918063784018159,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34569137200014666
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.012864989432273433,
          "train_loss": 0.013847118243575096,
          "validation_loss": 0.011478676460683346,
          "train_metric": 0.0013946308754384518,
          "validation_metric": 0.0011560902930796146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3571893740008818
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.012840379844419658,
          "train_loss": 0.018173998221755028,
          "validation_loss": 0.014235413633286953,
          "train_metric": 0.0018304184777662158,
          "validation_metric": 0.0014337387401610613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3686225530000229
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.013136615976691246,
          "train_loss": 0.013094225898385048,
          "validation_loss": 0.009667407721281052,
          "train_metric": 0.0013188024749979377,
          "validation_metric": 0.000973665970377624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3801374940012465
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.014086473936913535,
          "train_loss": 0.018723998218774796,
          "validation_loss": 0.017207995057106018,
          "train_metric": 0.001885812496766448,
          "validation_metric": 0.001733126468025148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39183506700101134
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01546470815083012,
          "train_loss": 0.015004885382950306,
          "validation_loss": 0.012785745784640312,
          "train_metric": 0.0015112372348085046,
          "validation_metric": 0.0012877336703240871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40368639000007533
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.013454599829856306,
          "train_loss": 0.009090457111597061,
          "validation_loss": 0.006461574230343103,
          "train_metric": 0.0009155576699413359,
          "validation_metric": 0.0006507861544378102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41595857700121996
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.010267586330883205,
          "train_loss": 0.011370032094419003,
          "validation_loss": 0.007344658020883799,
          "train_metric": 0.0011451481841504574,
          "validation_metric": 0.0007397269946523011,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42782877900026506
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.009417531473445706,
          "train_loss": 0.009502162225544453,
          "validation_loss": 0.007631679531186819,
          "train_metric": 0.00095702288672328,
          "validation_metric": 0.0007686347817070782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4395053820007888
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.011668808467220515,
          "train_loss": 0.010406026616692543,
          "validation_loss": 0.008246649987995625,
          "train_metric": 0.001048056990839541,
          "validation_metric": 0.0008305724477395415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45145693099948403
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.011381455231457949,
          "train_loss": 0.012136897072196007,
          "validation_loss": 0.009510010480880737,
          "train_metric": 0.0012223839294165373,
          "validation_metric": 0.0009578133467584848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46311407100074575
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.012793073736247607,
          "train_loss": 0.00904192216694355,
          "validation_loss": 0.006943841464817524,
          "train_metric": 0.000910669332370162,
          "validation_metric": 0.0006993583519943058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47509355600050185
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.009283731487812474,
          "train_loss": 0.007819491438567638,
          "validation_loss": 0.005947722587734461,
          "train_metric": 0.0007875505252741277,
          "validation_metric": 0.0005990328500047326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4875392650010326
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.007997443128260784,
          "train_loss": 0.009070136584341526,
          "validation_loss": 0.006412943359464407,
          "train_metric": 0.0009135109721682966,
          "validation_metric": 0.0006458881543949246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4994668270010152
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.00927292145206593,
          "train_loss": 0.007015549577772617,
          "validation_loss": 0.004979750141501427,
          "train_metric": 0.0007065804675221443,
          "validation_metric": 0.000501542235724628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5117680500006827
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009256921330234036,
          "train_loss": 0.010151050053536892,
          "validation_loss": 0.008510993793606758,
          "train_metric": 0.001022376585751772,
          "validation_metric": 0.0008571961079724133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5239174310008821
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01022659832960926,
          "train_loss": 0.008883602917194366,
          "validation_loss": 0.005962051451206207,
          "train_metric": 0.0008947239839471877,
          "validation_metric": 0.0006004759925417602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5358651839997037
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.008463179481623229,
          "train_loss": 0.008255747146904469,
          "validation_loss": 0.005772091913968325,
          "train_metric": 0.0008314886945299804,
          "validation_metric": 0.0005813440075144172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5475120049995894
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.007438105851178989,
          "train_loss": 0.006874631159007549,
          "validation_loss": 0.004806286655366421,
          "train_metric": 0.0006923876353539526,
          "validation_metric": 0.0004840715555474162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5598627960007434
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.006828691664850339,
          "train_loss": 0.0072225662879645824,
          "validation_loss": 0.005266937427222729,
          "train_metric": 0.0007274303934536874,
          "validation_metric": 0.0005304665537551045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5718061070001568
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.007311208915780298,
          "train_loss": 0.007752222009003162,
          "validation_loss": 0.006069664843380451,
          "train_metric": 0.0007807754445821047,
          "validation_metric": 0.0006113143754191697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5835708650010929
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.008883359972969629,
          "train_loss": 0.006585492752492428,
          "validation_loss": 0.005326066631823778,
          "train_metric": 0.0006632668082602322,
          "validation_metric": 0.0005364218959584832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5964669730001333
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.007252365699969232,
          "train_loss": 0.007961304858326912,
          "validation_loss": 0.005588250234723091,
          "train_metric": 0.0008018334629014134,
          "validation_metric": 0.0005628280341625214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.614668354000969
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.006298056279774755,
          "train_loss": 0.006259843707084656,
          "validation_loss": 0.005208888091146946,
          "train_metric": 0.0006304685375653207,
          "validation_metric": 0.0005246201762929559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.630842336000569
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.006815289307269268,
          "train_loss": 0.006204939913004637,
          "validation_loss": 0.004302111454308033,
          "train_metric": 0.0006249388097785413,
          "validation_metric": 0.0004332929092925042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.648138108999774
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.006616720085730776,
          "train_loss": 0.006390729919075966,
          "validation_loss": 0.004949348978698254,
          "train_metric": 0.0006436508847400546,
          "validation_metric": 0.0004984802799299359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6634757980009454
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.006861820787889883,
          "train_loss": 0.005533527582883835,
          "validation_loss": 0.004149790853261948,
          "train_metric": 0.0005573166417889297,
          "validation_metric": 0.0004179516981821507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6804479679994984
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.0057468981394777074,
          "train_loss": 0.006055423989892006,
          "validation_loss": 0.003916261717677116,
          "train_metric": 0.00060988002223894,
          "validation_metric": 0.0003944315540138632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6960791000001336
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.005683069466613233,
          "train_loss": 0.005231151822954416,
          "validation_loss": 0.003850997658446431,
          "train_metric": 0.0005268624518066645,
          "validation_metric": 0.00038785842480137944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7084713710009964
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.005967860444798134,
          "train_loss": 0.005784656852483749,
          "validation_loss": 0.003646858735010028,
          "train_metric": 0.000582609442062676,
          "validation_metric": 0.00036729828570969403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7203502350002964
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.005565942330576945,
          "train_loss": 0.004933000076562166,
          "validation_loss": 0.0035538645461201668,
          "train_metric": 0.0004968336434103549,
          "validation_metric": 0.00035793223651126027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7332874200001243
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.006480748736066744,
          "train_loss": 0.005111610982567072,
          "validation_loss": 0.0038722525350749493,
          "train_metric": 0.000514822662808001,
          "validation_metric": 0.0003899990697391331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7451992950009299
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.005288970292895101,
          "train_loss": 0.006706181913614273,
          "validation_loss": 0.004314645193517208,
          "train_metric": 0.000675422081258148,
          "validation_metric": 0.00043455531704239547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7569629379995604
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0057978276163339615,
          "train_loss": 0.005068046040832996,
          "validation_loss": 0.003546724794432521,
          "train_metric": 0.0005104350275360048,
          "validation_metric": 0.00035721316817216575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7695359839999583
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.005810117145301774,
          "train_loss": 0.005381204653531313,
          "validation_loss": 0.003783063031733036,
          "train_metric": 0.0005419752560555935,
          "validation_metric": 0.0003810162888839841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7957229990006454
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.005235293596342672,
          "train_loss": 0.004641404841095209,
          "validation_loss": 0.0034183552488684654,
          "train_metric": 0.00046746531734243035,
          "validation_metric": 0.0003442841989453882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8138520009997592
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.005583508784184232,
          "train_loss": 0.00531682837754488,
          "validation_loss": 0.0043704695999622345,
          "train_metric": 0.0005354915047064424,
          "validation_metric": 0.0004401776532176882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8320652420006809
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.006284631570451893,
          "train_loss": 0.004827606026083231,
          "validation_loss": 0.0038336729630827904,
          "train_metric": 0.0004862188652623445,
          "validation_metric": 0.0003861134755425155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8499771030001284
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005527699759113602,
          "train_loss": 0.0048644812777638435,
          "validation_loss": 0.002998816780745983,
          "train_metric": 0.0004899327759630978,
          "validation_metric": 0.0003020298609044403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8685947169997235
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.005861759746039752,
          "train_loss": 0.004774818196892738,
          "validation_loss": 0.003193967044353485,
          "train_metric": 0.00048090217751450837,
          "validation_metric": 0.0003216846671421081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0206731279995438
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.005350327308406122,
          "train_loss": 0.006637380458414555,
          "validation_loss": 0.005715301260352135,
          "train_metric": 0.0006684926920570433,
          "validation_metric": 0.0005756242317147553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.038839345001179
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.006015184102579951,
          "train_loss": 0.005435135215520859,
          "validation_loss": 0.004723444581031799,
          "train_metric": 0.0005474069621413946,
          "validation_metric": 0.00047572804032824934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0614321070006554
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0053273004305083305,
          "train_loss": 0.004579838365316391,
          "validation_loss": 0.003218955360352993,
          "train_metric": 0.0004612645134329796,
          "validation_metric": 0.0003242013917770237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0800838779996411
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.005068351645604707,
          "train_loss": 0.005395839922130108,
          "validation_loss": 0.003962008282542229,
          "train_metric": 0.0005434492486529052,
          "validation_metric": 0.00039903895230963826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0988700779998908
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.005400688154622912,
          "train_loss": 0.004358076024800539,
          "validation_loss": 0.0032358989119529724,
          "train_metric": 0.00043892947724089026,
          "validation_metric": 0.00032590789487585425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1195635510011925
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0059378365040174685,
          "train_loss": 0.006800681818276644,
          "validation_loss": 0.006642906926572323,
          "train_metric": 0.0006849398487247527,
          "validation_metric": 0.0006690493319183588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.142270821999773
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005233767253230326,
          "train_loss": 0.007784011773765087,
          "validation_loss": 0.006885326933115721,
          "train_metric": 0.0007839772151783109,
          "validation_metric": 0.0006934648845344782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1636956260008446
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0062410623067989945,
          "train_loss": 0.004759213887155056,
          "validation_loss": 0.003570742905139923,
          "train_metric": 0.00047933062887750566,
          "validation_metric": 0.00035963216214440763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1794464220001828
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.006019235515850596,
          "train_loss": 0.004811122547835112,
          "validation_loss": 0.0033644712530076504,
          "train_metric": 0.00048455866635777056,
          "validation_metric": 0.00033885720768012106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1932209440001316
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.005993802231387235,
          "train_loss": 0.006095379125326872,
          "validation_loss": 0.004258971661329269,
          "train_metric": 0.0006139042670838535,
          "validation_metric": 0.0004289480857551098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2155465970008663
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.004641653926228173,
          "train_loss": 0.00420210650190711,
          "validation_loss": 0.0029149192851036787,
          "train_metric": 0.0004232207720633596,
          "validation_metric": 0.0002935800002887845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2317767980002827
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.004445495163963642,
          "train_loss": 0.005444948561489582,
          "validation_loss": 0.003588505554944277,
          "train_metric": 0.0005483952118083835,
          "validation_metric": 0.0003614210581872612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2449367999997776
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.005780828432762064,
          "train_loss": 0.005872534122318029,
          "validation_loss": 0.003947665449231863,
          "train_metric": 0.0005914600915275514,
          "validation_metric": 0.0003975944418925792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2570563530007348
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0054164044704521075,
          "train_loss": 0.004863618407398462,
          "validation_loss": 0.004155816975980997,
          "train_metric": 0.000489845871925354,
          "validation_metric": 0.0004185587167739868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.269165798999893
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.005822308041388169,
          "train_loss": 0.005616763606667519,
          "validation_loss": 0.003736957209184766,
          "train_metric": 0.00056569988373667,
          "validation_metric": 0.0003763726563192904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2812105749999319
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.006408159533748403,
          "train_loss": 0.006463616155087948,
          "validation_loss": 0.0046817949041724205,
          "train_metric": 0.0006509916856884956,
          "validation_metric": 0.00047153333434835076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2941499259995908
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.006352205775328912,
          "train_loss": 0.004994518589228392,
          "validation_loss": 0.004389088600873947,
          "train_metric": 0.0005030296160839498,
          "validation_metric": 0.000442052842117846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.30679707300078
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005285533829010092,
          "train_loss": 0.004425332881510258,
          "validation_loss": 0.0030121933668851852,
          "train_metric": 0.0004457033355720341,
          "validation_metric": 0.00030337704811245203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.318641010000647
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005119116234709509,
          "train_loss": 0.0055526550859212875,
          "validation_loss": 0.0046606832183897495,
          "train_metric": 0.0005592430825345218,
          "validation_metric": 0.0004694069502875209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3305315930010693
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0054295011213980615,
          "train_loss": 0.00611022999510169,
          "validation_loss": 0.004299456719309092,
          "train_metric": 0.0006153999711386859,
          "validation_metric": 0.00043302556150592864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.342448532999697
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0045953455301059876,
          "train_loss": 0.003767765825614333,
          "validation_loss": 0.002512366743758321,
          "train_metric": 0.000379475619411096,
          "validation_metric": 0.00025303635629825294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3548688560003939
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.004676159980590455,
          "train_loss": 0.005913136061280966,
          "validation_loss": 0.004925603978335857,
          "train_metric": 0.0005955494125373662,
          "validation_metric": 0.000496088876388967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3675758660010615
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004823824354389217,
          "train_loss": 0.004803627263754606,
          "validation_loss": 0.004472620319575071,
          "train_metric": 0.0004838037129957229,
          "validation_metric": 0.0004504659154918045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3798529630003031
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004768033511936665,
          "train_loss": 0.003656575921922922,
          "validation_loss": 0.0026814662851393223,
          "train_metric": 0.00036827693111263216,
          "validation_metric": 0.00027006748132407665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3921667779995914
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004073190742929,
          "train_loss": 0.006146005354821682,
          "validation_loss": 0.0042499578557908535,
          "train_metric": 0.0006190031417645514,
          "validation_metric": 0.000428040191764012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.404005080999923
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0055284610207309015,
          "train_loss": 0.003937103319913149,
          "validation_loss": 0.002437623217701912,
          "train_metric": 0.0003965306677855551,
          "validation_metric": 0.00024550847592763603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4167984619998606
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0049750418329495005,
          "train_loss": 0.0035211576614528894,
          "validation_loss": 0.0024981757160276175,
          "train_metric": 0.0003546381485648453,
          "validation_metric": 0.0002516071544960141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.429334481999831
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004926310612063389,
          "train_loss": 0.0056104459799826145,
          "validation_loss": 0.005162070970982313,
          "train_metric": 0.000565063557587564,
          "validation_metric": 0.0005199048318900168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4416616830003477
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.00457811496744398,
          "train_loss": 0.0035083452239632607,
          "validation_loss": 0.0027148877270519733,
          "train_metric": 0.0003533477138262242,
          "validation_metric": 0.00027343357214704156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.453457730000082
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0038975166535237804,
          "train_loss": 0.003266651416197419,
          "validation_loss": 0.0022719413973391056,
          "train_metric": 0.00032900518272072077,
          "validation_metric": 0.00022882162011228502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4656885050007986
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.004353592841653153,
          "train_loss": 0.005987572483718395,
          "validation_loss": 0.004357532598078251,
          "train_metric": 0.000603046384640038,
          "validation_metric": 0.0004388746456243098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4782672630008165
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.007027880245004781,
          "train_loss": 0.003532577771693468,
          "validation_loss": 0.0028709189500659704,
          "train_metric": 0.000355788302840665,
          "validation_metric": 0.0002891484182327986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4902760650002165
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004618098930222914,
          "train_loss": 0.006787808611989021,
          "validation_loss": 0.0052382126450538635,
          "train_metric": 0.0006836432730779052,
          "validation_metric": 0.0005275736330077052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5026117249999515
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005917414498981088,
          "train_loss": 0.007802060339599848,
          "validation_loss": 0.006218149326741695,
          "train_metric": 0.0007857950404286385,
          "validation_metric": 0.0006262692622840405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5147773170010623
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.004891372169367969,
          "train_loss": 0.0038094751071184874,
          "validation_loss": 0.002656390890479088,
          "train_metric": 0.00038367637898772955,
          "validation_metric": 0.00026754196733236313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5268691800010856
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0047379937095684,
          "train_loss": 0.004543782211840153,
          "validation_loss": 0.0030680310446769,
          "train_metric": 0.0004576331120915711,
          "validation_metric": 0.0003090008394792676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.538128370000777
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0046207040213630535,
          "train_loss": 0.004189496394246817,
          "validation_loss": 0.003529977984726429,
          "train_metric": 0.00042195073910988867,
          "validation_metric": 0.00035552645567804575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5502890970001317
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003932131854526233,
          "train_loss": 0.0034876123536378145,
          "validation_loss": 0.0021193150896579027,
          "train_metric": 0.00035125951399095356,
          "validation_metric": 0.00021344967535696924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5626175489996967
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.00361641696508741,
          "train_loss": 0.004738669376820326,
          "validation_loss": 0.003137103980407119,
          "train_metric": 0.00047726143384352326,
          "validation_metric": 0.00031595758628100157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5752398660006293
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.00421824926888803,
          "train_loss": 0.004284790251404047,
          "validation_loss": 0.0030047274194657803,
          "train_metric": 0.0004315483383834362,
          "validation_metric": 0.00030262517975643277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.58712303700122
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.006816689900006168,
          "train_loss": 0.0038810905534774065,
          "validation_loss": 0.0026062377728521824,
          "train_metric": 0.00039088918128982186,
          "validation_metric": 0.00026249076472595334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.598729684999853
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0051575257602962665,
          "train_loss": 0.004518231842666864,
          "validation_loss": 0.0030579494778066874,
          "train_metric": 0.00045505978050641716,
          "validation_metric": 0.00030798546504229307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.610786993000147
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.004335119767347351,
          "train_loss": 0.006003900431096554,
          "validation_loss": 0.004198165610432625,
          "train_metric": 0.0006046909256838262,
          "validation_metric": 0.00042282382491976023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6223852210005134
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004690614689025097,
          "train_loss": 0.003927458077669144,
          "validation_loss": 0.0025539712514728308,
          "train_metric": 0.00039555918192490935,
          "validation_metric": 0.0002572266384959221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6338429790012015
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004483993470785208,
          "train_loss": 0.004046859219670296,
          "validation_loss": 0.002704824088141322,
          "train_metric": 0.00040758485556580126,
          "validation_metric": 0.00027241994393989444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6458692260002863
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0034093073481926695,
          "train_loss": 0.0052793254144489765,
          "validation_loss": 0.00395508948713541,
          "train_metric": 0.0005317143513821065,
          "validation_metric": 0.0003983421192970127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6573275350001495
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.00535646773641929,
          "train_loss": 0.0038928892463445663,
          "validation_loss": 0.003726498456671834,
          "train_metric": 0.0003920775488950312,
          "validation_metric": 0.0003753192722797394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6687650349995238
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.004436049741343595,
          "train_loss": 0.004203645046800375,
          "validation_loss": 0.002665060805156827,
          "train_metric": 0.0004233756917528808,
          "validation_metric": 0.00026841519866138697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.680371198001012
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0036649763278546743,
          "train_loss": 0.003873492358252406,
          "validation_loss": 0.002534786704927683,
          "train_metric": 0.00039012389606796205,
          "validation_metric": 0.00025529443519189954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6917744559996208
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0035584890210884623,
          "train_loss": 0.0032731688115745783,
          "validation_loss": 0.0022228413727134466,
          "train_metric": 0.0003296615614090115,
          "validation_metric": 0.00022387641365639865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7034468199999537
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003205988909030566,
          "train_loss": 0.003316253889352083,
          "validation_loss": 0.0025222990661859512,
          "train_metric": 0.00033400088432244956,
          "validation_metric": 0.00025403674226254225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7153947389997484
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0034973064975929447,
          "train_loss": 0.0040602730587124825,
          "validation_loss": 0.0025129697751253843,
          "train_metric": 0.00040893579716794193,
          "validation_metric": 0.0002530971250962466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7269100390003587
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.003767910020542331,
          "train_loss": 0.003192337229847908,
          "validation_loss": 0.0021984667982906103,
          "train_metric": 0.00032152055064216256,
          "validation_metric": 0.00022142150555737317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7382243129995913
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.003892539389198646,
          "train_loss": 0.003293877001851797,
          "validation_loss": 0.002867072820663452,
          "train_metric": 0.0003317472292110324,
          "validation_metric": 0.0002887610753532499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.750607244001003
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.003834067603747826,
          "train_loss": 0.002939149970188737,
          "validation_loss": 0.0020248882938176394,
          "train_metric": 0.000296020443784073,
          "validation_metric": 0.00020393935847096145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7629952959996444
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0035093145997961983,
          "train_loss": 0.002978731645271182,
          "validation_loss": 0.0021320576779544353,
          "train_metric": 0.00030000697006471455,
          "validation_metric": 0.00021473303786478937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7752464759996656
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0035317220572324004,
          "train_loss": 0.0031696909572929144,
          "validation_loss": 0.002115786075592041,
          "train_metric": 0.00031923962524160743,
          "validation_metric": 0.00021309423027560115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7871302599996852
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.004334396435297094,
          "train_loss": 0.005304331891238689,
          "validation_loss": 0.0037582286167889833,
          "train_metric": 0.0005342328222468495,
          "validation_metric": 0.0003785150474868715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7991630739998072
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.004663044048356824,
          "train_loss": 0.004026136826723814,
          "validation_loss": 0.002643594052642584,
          "train_metric": 0.000405497761676088,
          "validation_metric": 0.0002662531042005867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.811482123001042
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.003928199119400233,
          "train_loss": 0.0038833373691886663,
          "validation_loss": 0.0028767099138349295,
          "train_metric": 0.0003911155217792839,
          "validation_metric": 0.000289731688098982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.823209346999647
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0036898496909998357,
          "train_loss": 0.00392697099596262,
          "validation_loss": 0.0031566289253532887,
          "train_metric": 0.00039551020017825067,
          "validation_metric": 0.0003179241030011326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8348556740002095
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.003174459816364106,
          "train_loss": 0.004343569278717041,
          "validation_loss": 0.0036135362461209297,
          "train_metric": 0.0004374684067443013,
          "validation_metric": 0.0003639420901890844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8466873090001172
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.005352681226213463,
          "train_loss": 0.0056690638884902,
          "validation_loss": 0.005004751496016979,
          "train_metric": 0.0005709673278033733,
          "validation_metric": 0.0005040602991357446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8585856150002655
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.004641719024220947,
          "train_loss": 0.004019936081022024,
          "validation_loss": 0.002818363020196557,
          "train_metric": 0.00040487328078597784,
          "validation_metric": 0.0002838551881723106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8702785839996068
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.003382593291462399,
          "train_loss": 0.0026573622599244118,
          "validation_loss": 0.0017692161491140723,
          "train_metric": 0.00026763981441035867,
          "validation_metric": 0.00017818894411902875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.882425621000948
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0030600178506574593,
          "train_loss": 0.0038041071966290474,
          "validation_loss": 0.00328077282756567,
          "train_metric": 0.0003831357753369957,
          "validation_metric": 0.00033042742870748043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.894566807000956
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0038748957158531994,
          "train_loss": 0.00813765823841095,
          "validation_loss": 0.007632839027792215,
          "train_metric": 0.0008195951231755316,
          "validation_metric": 0.0007687516044825315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.906398856001033
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.004792598090716638,
          "train_loss": 0.0041766539216041565,
          "validation_loss": 0.0035274235997349024,
          "train_metric": 0.0004206573066767305,
          "validation_metric": 0.0003552691778168082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.91826147699976
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0058836762473220006,
          "train_loss": 0.006190415471792221,
          "validation_loss": 0.004424794111400843,
          "train_metric": 0.0006234760512597859,
          "validation_metric": 0.00044564908603206277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9300525740000012
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.005927729813265614,
          "train_loss": 0.0027383507695049047,
          "validation_loss": 0.0018062472809106112,
          "train_metric": 0.0002757967449724674,
          "validation_metric": 0.0001819186145439744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9417834819996642
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.003709530341438949,
          "train_loss": 0.0033649129327386618,
          "validation_loss": 0.002530371304601431,
          "train_metric": 0.00033890173654071987,
          "validation_metric": 0.00025484972866252065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9534309780010517
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0035946535354014486,
          "train_loss": 0.00367038999684155,
          "validation_loss": 0.002438151743263006,
          "train_metric": 0.00036966826883144677,
          "validation_metric": 0.0002455617068335414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9653977230009332
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.00319260818650946,
          "train_loss": 0.003999529406428337,
          "validation_loss": 0.003420056775212288,
          "train_metric": 0.0004028179100714624,
          "validation_metric": 0.0003444555914029479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9774409830006334
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0031393547033076175,
          "train_loss": 0.0030524858739227057,
          "validation_loss": 0.0022543163504451513,
          "train_metric": 0.0003074351989198476,
          "validation_metric": 0.00022704644652549177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9891128819999722
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.003377741333679296,
          "train_loss": 0.004474192392081022,
          "validation_loss": 0.003678934182971716,
          "train_metric": 0.00045062426943331957,
          "validation_metric": 0.000370528781786561,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.000772321000113
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.004240290247253142,
          "train_loss": 0.007878913544118404,
          "validation_loss": 0.006281642708927393,
          "train_metric": 0.0007935353205539286,
          "validation_metric": 0.0006326640723273158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.01267371799986
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.00338081249356037,
          "train_loss": 0.0028165800031274557,
          "validation_loss": 0.002256578765809536,
          "train_metric": 0.00028367561753839254,
          "validation_metric": 0.00022727435862179846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.024911366001106
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0036752172964042984,
          "train_loss": 0.005854913964867592,
          "validation_loss": 0.005957808345556259,
          "train_metric": 0.0005896854563616216,
          "validation_metric": 0.0006000486901029944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.037112903999514
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0037044844139018096,
          "train_loss": 0.003184410510584712,
          "validation_loss": 0.002440069802105427,
          "train_metric": 0.000320722145261243,
          "validation_metric": 0.00024575486895628273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0490329260010185
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.004059064565808512,
          "train_loss": 0.005489780101925135,
          "validation_loss": 0.004053638316690922,
          "train_metric": 0.000552910496480763,
          "validation_metric": 0.0004082676023244858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.061225082999954
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.004405483501614071,
          "train_loss": 0.0033456431701779366,
          "validation_loss": 0.002252972684800625,
          "train_metric": 0.0003369609476067126,
          "validation_metric": 0.0002269111282657832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0729018229994836
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.003769657967495732,
          "train_loss": 0.0029887137934565544,
          "validation_loss": 0.0024262680672109127,
          "train_metric": 0.00030101227457635105,
          "validation_metric": 0.0002443648409098387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.084778880000158
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0029857454064767808,
          "train_loss": 0.003486915025860071,
          "validation_loss": 0.0022663972340524197,
          "train_metric": 0.00035118928644806147,
          "validation_metric": 0.00022826323402114213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.096866346000752
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0026376548630651087,
          "train_loss": 0.0028745378367602825,
          "validation_loss": 0.002202028641477227,
          "train_metric": 0.00028951294370926917,
          "validation_metric": 0.0002217802539234981,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1084792640012893
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.002853973321180092,
          "train_loss": 0.002630020258948207,
          "validation_loss": 0.0016821697354316711,
          "train_metric": 0.0002648860099725425,
          "validation_metric": 0.0001694219681667164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.120684433999486
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0027943956229137257,
          "train_loss": 0.0023969581816345453,
          "validation_loss": 0.0019306334434077144,
          "train_metric": 0.0002414128539385274,
          "validation_metric": 0.00019444632926024497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.132889382999565
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.003112567042990122,
          "train_loss": 0.003348172176629305,
          "validation_loss": 0.0021480254363268614,
          "train_metric": 0.000337215606123209,
          "validation_metric": 0.00021634124277625233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.144628559000921
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0034175651308032684,
          "train_loss": 0.003435567021369934,
          "validation_loss": 0.0034098424948751926,
          "train_metric": 0.00034601771039888263,
          "validation_metric": 0.00034342685830779374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1563771199998882
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.003548424290784169,
          "train_loss": 0.0026300777681171894,
          "validation_loss": 0.0024048432242125273,
          "train_metric": 0.00026489183073863387,
          "validation_metric": 0.00024220702471211553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.168063520000942
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0031863577823969536,
          "train_loss": 0.002945939777418971,
          "validation_loss": 0.0026810592971742153,
          "train_metric": 0.00029670423828065395,
          "validation_metric": 0.0002700264740269631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1799492629997985
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0026648578023014124,
          "train_loss": 0.0032045706175267696,
          "validation_loss": 0.0022559084463864565,
          "train_metric": 0.0003227526322007179,
          "validation_metric": 0.00022720683773513883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.192177012000684
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.003376838238182245,
          "train_loss": 0.0028649596497416496,
          "validation_loss": 0.0018867834005504847,
          "train_metric": 0.00028854829724878073,
          "validation_metric": 0.00019002992485184222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.204065789001106
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.002891153362725163,
          "train_loss": 0.008384707383811474,
          "validation_loss": 0.007588061969727278,
          "train_metric": 0.0008444770937785506,
          "validation_metric": 0.0007642418495379388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2159951550002006
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0045323612648644485,
          "train_loss": 0.004999445751309395,
          "validation_loss": 0.00527253607288003,
          "train_metric": 0.0005035258363932371,
          "validation_metric": 0.0005310304695740342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.228340356999979
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.00411058340978343,
          "train_loss": 0.002716198330745101,
          "validation_loss": 0.0023050825111567974,
          "train_metric": 0.0002735655289143324,
          "validation_metric": 0.00023215945111587644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.240317022000454
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0033877945606946014,
          "train_loss": 0.0027822647243738174,
          "validation_loss": 0.0023476150818169117,
          "train_metric": 0.00028021950856782496,
          "validation_metric": 0.0002364432148169726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.252141433000361
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.003429299766139593,
          "train_loss": 0.002710760571062565,
          "validation_loss": 0.0018133610719814897,
          "train_metric": 0.00027301788213662803,
          "validation_metric": 0.00018263507809024304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2644353340001544
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0034073698698193766,
          "train_loss": 0.004460243508219719,
          "validation_loss": 0.003793308511376381,
          "train_metric": 0.0004492193984333426,
          "validation_metric": 0.0003820481360889971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2763080700005958
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0035942096292274073,
          "train_loss": 0.003387169912457466,
          "validation_loss": 0.0020354860462248325,
          "train_metric": 0.00034114340087398887,
          "validation_metric": 0.00020500668324530125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2882490669999243
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0025179784643114544,
          "train_loss": 0.002252730308100581,
          "validation_loss": 0.0014835622860118747,
          "train_metric": 0.00022688675380777568,
          "validation_metric": 0.00014941896370146424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3010663439999917
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0031313605213654228,
          "train_loss": 0.004451652988791466,
          "validation_loss": 0.004341877531260252,
          "train_metric": 0.0004483542579691857,
          "validation_metric": 0.0004372979747131467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3132881750007073
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0031882455077720806,
          "train_loss": 0.002210333477705717,
          "validation_loss": 0.0017155923414975405,
          "train_metric": 0.0002226166834589094,
          "validation_metric": 0.00017278817540500313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3258214050001698
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0025819979346124455,
          "train_loss": 0.003362989751622081,
          "validation_loss": 0.002619760809466243,
          "train_metric": 0.00033870802144519985,
          "validation_metric": 0.00026385264936834574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3378531360012857
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0031695523066446185,
          "train_loss": 0.0028710647020488977,
          "validation_loss": 0.0019611769821494818,
          "train_metric": 0.0002891631156671792,
          "validation_metric": 0.00019752256048377603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3496208240012493
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0034672471156227402,
          "train_loss": 0.0025314530357718468,
          "validation_loss": 0.0015392169589176774,
          "train_metric": 0.00025495869340375066,
          "validation_metric": 0.00015502430323977023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.362165169999571
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0035358480745344423,
          "train_loss": 0.002534402534365654,
          "validation_loss": 0.0017995801754295826,
          "train_metric": 0.00025525575620122254,
          "validation_metric": 0.00018124713096767664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.374190701000771
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.005338521441444755,
          "train_loss": 0.002814941806718707,
          "validation_loss": 0.002184823388233781,
          "train_metric": 0.0002835106570273638,
          "validation_metric": 0.00022004744096193463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.386036203000913
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0029864490861655213,
          "train_loss": 0.003539434401318431,
          "validation_loss": 0.002709232969209552,
          "train_metric": 0.0003564788494259119,
          "validation_metric": 0.0002728640101850033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3982688699998107
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.004030684165627463,
          "train_loss": 0.0036999029107391834,
          "validation_loss": 0.0028419620357453823,
          "train_metric": 0.0003726407012436539,
          "validation_metric": 0.00028623195248655975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4103566049998335
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.003549753448169213,
          "train_loss": 0.0036475001834332943,
          "validation_loss": 0.002652571303769946,
          "train_metric": 0.0003673628671094775,
          "validation_metric": 0.0002671573020052165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4225789140000416
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0034516766318120062,
          "train_loss": 0.003762959036976099,
          "validation_loss": 0.002342422027140856,
          "train_metric": 0.00037899144808761775,
          "validation_metric": 0.0002359201607760042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4342882720011403
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0027032552752643824,
          "train_loss": 0.002131497021764517,
          "validation_loss": 0.0015696206828579307,
          "train_metric": 0.00021467656188178807,
          "validation_metric": 0.00015808644820936024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4461950050008454
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.00265920768288197,
          "train_loss": 0.0031132781878113747,
          "validation_loss": 0.002886026632040739,
          "train_metric": 0.0003135579463560134,
          "validation_metric": 0.00029067002469673753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.458332280000832
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0033956680417759344,
          "train_loss": 0.002263477537781,
          "validation_loss": 0.0017613241216167808,
          "train_metric": 0.0002279691689182073,
          "validation_metric": 0.00017739410395734012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.47021182699973
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0029981656261952594,
          "train_loss": 0.0021254231687635183,
          "validation_loss": 0.0015009372727945447,
          "train_metric": 0.00021406484302133322,
          "validation_metric": 0.00015116887516342103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.481889267000952
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.002204551343311323,
          "train_loss": 0.0034755647648125887,
          "validation_loss": 0.0032438719645142555,
          "train_metric": 0.0003500461461953819,
          "validation_metric": 0.00032671092776581645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.494100289000926
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.003833627182757482,
          "train_loss": 0.00557368341833353,
          "validation_loss": 0.004663162864744663,
          "train_metric": 0.0005613609682768583,
          "validation_metric": 0.0004696566320490092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.50552862800032
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.006104557709477376,
          "train_loss": 0.0063149346970021725,
          "validation_loss": 0.005661233328282833,
          "train_metric": 0.0006360170664265752,
          "validation_metric": 0.0005701786722056568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.517248047999601
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.007628354505868629,
          "train_loss": 0.004314083606004715,
          "validation_loss": 0.004297105595469475,
          "train_metric": 0.0004344986518844962,
          "validation_metric": 0.00043278877274133265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.529558291000285
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.006081807863665745,
          "train_loss": 0.0043970756232738495,
          "validation_loss": 0.0033758399076759815,
          "train_metric": 0.0004428573593031615,
          "validation_metric": 0.00034000223968178034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5414888440009236
        }
      ],
      "best": {
        "epoch": 167,
        "updates": 2672
      },
      "best_validation_metric": 0.00014941896370146424,
      "config": {
        "label": "Tanh",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012223704990773633
    },
    "a7bb18d2c173_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7289919313043356,
          "train_loss": 0.2777801752090454,
          "validation_loss": 0.228195920586586,
          "train_metric": 0.027977004647254944,
          "validation_metric": 0.02298305742442608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.019601037000029464
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.2019943268969655,
          "train_loss": 0.13836921751499176,
          "validation_loss": 0.11704222857952118,
          "train_metric": 0.013936040922999382,
          "validation_metric": 0.011788065545260906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.038679276000038953
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.1311952737160027,
          "train_loss": 0.10032099485397339,
          "validation_loss": 0.08393031358718872,
          "train_metric": 0.010103963315486908,
          "validation_metric": 0.008453154936432838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05800905499927467
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10741799045354128,
          "train_loss": 0.08218062669038773,
          "validation_loss": 0.0676356628537178,
          "train_metric": 0.008276931941509247,
          "validation_metric": 0.0068120164796710014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07630367700039642
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08750398340635002,
          "train_loss": 0.06645867228507996,
          "validation_loss": 0.05322501063346863,
          "train_metric": 0.006693474482744932,
          "validation_metric": 0.005360627546906471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09511480500077596
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.07259275927208364,
          "train_loss": 0.057584524154663086,
          "validation_loss": 0.04773184657096863,
          "train_metric": 0.0057997023686766624,
          "validation_metric": 0.0048073772341012955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11378505200082145
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06560769595671445,
          "train_loss": 0.05122227594256401,
          "validation_loss": 0.04228980466723442,
          "train_metric": 0.005158919841051102,
          "validation_metric": 0.004259274806827307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13235059500038915
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.062013103044591844,
          "train_loss": 0.04715650528669357,
          "validation_loss": 0.038969215005636215,
          "train_metric": 0.00474943034350872,
          "validation_metric": 0.003924836404621601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15124407899929793
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0586589208105579,
          "train_loss": 0.050989095121622086,
          "validation_loss": 0.043174754828214645,
          "train_metric": 0.005135434679687023,
          "validation_metric": 0.004348402842879295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16947970100045495
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.06392525497358292,
          "train_loss": 0.05687223747372627,
          "validation_loss": 0.04932605102658272,
          "train_metric": 0.005727963987737894,
          "validation_metric": 0.004967939108610153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1875466410001536
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.05662829475477338,
          "train_loss": 0.051668234169483185,
          "validation_loss": 0.04496077075600624,
          "train_metric": 0.005203835666179657,
          "validation_metric": 0.0045282840728759766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20695926599910308
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.055891335709020495,
          "train_loss": 0.045070186257362366,
          "validation_loss": 0.039089370518922806,
          "train_metric": 0.0045393044129014015,
          "validation_metric": 0.003936938010156155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22702412699982233
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.05181926186196506,
          "train_loss": 0.03629867359995842,
          "validation_loss": 0.029910337179899216,
          "train_metric": 0.0036558695137500763,
          "validation_metric": 0.0030124597251415253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2534318979996897
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.04680821963120252,
          "train_loss": 0.03517628461122513,
          "validation_loss": 0.029551757499575615,
          "train_metric": 0.0035428269766271114,
          "validation_metric": 0.0029763446655124426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2727631920006388
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.04642454953864217,
          "train_loss": 0.031789712607860565,
          "validation_loss": 0.025689929723739624,
          "train_metric": 0.003201743122190237,
          "validation_metric": 0.0025873957201838493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29244659500000125
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.044966342276893556,
          "train_loss": 0.02884243056178093,
          "validation_loss": 0.02382916584610939,
          "train_metric": 0.0029049040749669075,
          "validation_metric": 0.0023999863769859076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3112509910006338
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.04668564849998802,
          "train_loss": 0.027492348104715347,
          "validation_loss": 0.022660356014966965,
          "train_metric": 0.0027689288835972548,
          "validation_metric": 0.0022822683677077293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3303829870001209
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.04032665048725903,
          "train_loss": 0.027071718126535416,
          "validation_loss": 0.022455893456935883,
          "train_metric": 0.0027265644166618586,
          "validation_metric": 0.002261675661429763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3483539930002735
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.03721896244678646,
          "train_loss": 0.025471488013863564,
          "validation_loss": 0.02198648266494274,
          "train_metric": 0.0025653948541730642,
          "validation_metric": 0.0022143980022519827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.367898291999154
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.03935298282885924,
          "train_loss": 0.034415993839502335,
          "validation_loss": 0.02723383530974388,
          "train_metric": 0.003466252703219652,
          "validation_metric": 0.0027428921312093735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38577603200064914
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.03853065246948972,
          "train_loss": 0.02315734513103962,
          "validation_loss": 0.02040201984345913,
          "train_metric": 0.0023323227651417255,
          "validation_metric": 0.0020548165775835514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40411572300035914
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.03625843382906169,
          "train_loss": 0.02296084724366665,
          "validation_loss": 0.019133741036057472,
          "train_metric": 0.0023125321604311466,
          "validation_metric": 0.0019270802149549127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4225742100006755
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.03464341163635254,
          "train_loss": 0.02157323993742466,
          "validation_loss": 0.01746632531285286,
          "train_metric": 0.0021727776620537043,
          "validation_metric": 0.0017591441283002496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44063333200028865
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.03874533448833972,
          "train_loss": 0.021601153537631035,
          "validation_loss": 0.01760752499103546,
          "train_metric": 0.0021755890920758247,
          "validation_metric": 0.0017733651911839843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45843828699980804
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.028915220871567726,
          "train_loss": 0.018392978236079216,
          "validation_loss": 0.015309855341911316,
          "train_metric": 0.0018524733604863286,
          "validation_metric": 0.0015419526025652885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4764881579994835
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.03202313289511949,
          "train_loss": 0.018804414197802544,
          "validation_loss": 0.015327099710702896,
          "train_metric": 0.0018939118599519134,
          "validation_metric": 0.0015436895191669464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4959092450008029
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.02947720792144537,
          "train_loss": 0.016878286376595497,
          "validation_loss": 0.013709775172173977,
          "train_metric": 0.0016999191138893366,
          "validation_metric": 0.0013807982904836535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5143687370000407
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.028820125153288245,
          "train_loss": 0.016045892611145973,
          "validation_loss": 0.013997022993862629,
          "train_metric": 0.001616083551198244,
          "validation_metric": 0.0014097290113568306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5326868209995155
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.030251346179284155,
          "train_loss": 0.021729450672864914,
          "validation_loss": 0.01639833301305771,
          "train_metric": 0.002188510727137327,
          "validation_metric": 0.0016515799798071384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5498733709991939
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.02910318912472576,
          "train_loss": 0.016825692728161812,
          "validation_loss": 0.01310402899980545,
          "train_metric": 0.0016946223331615329,
          "validation_metric": 0.001319789793342352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5680846640007076
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.02845583384623751,
          "train_loss": 0.018129175528883934,
          "validation_loss": 0.01500378455966711,
          "train_metric": 0.001825904124416411,
          "validation_metric": 0.0015111262910068035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.585333241000626
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.030989460821729153,
          "train_loss": 0.014804827980697155,
          "validation_loss": 0.012360291555523872,
          "train_metric": 0.0014910879544913769,
          "validation_metric": 0.0012448834022507071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6036119729997154
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.02693056734278798,
          "train_loss": 0.015570315532386303,
          "validation_loss": 0.013285323046147823,
          "train_metric": 0.0015681851655244827,
          "validation_metric": 0.0013380489544942975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.621957191000547
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.026236666017211974,
          "train_loss": 0.01774716190993786,
          "validation_loss": 0.014239327982068062,
          "train_metric": 0.0017874293262138963,
          "validation_metric": 0.0014341329224407673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6394985120005003
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.029043307527899742,
          "train_loss": 0.01581514999270439,
          "validation_loss": 0.01243089884519577,
          "train_metric": 0.0015928440261632204,
          "validation_metric": 0.0012519946321845055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.656932689000314
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.029311845486517996,
          "train_loss": 0.01490138191729784,
          "validation_loss": 0.012832557782530785,
          "train_metric": 0.0015008127084001899,
          "validation_metric": 0.0012924481416121125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.674743497000236
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.02712525927927345,
          "train_loss": 0.01414641086012125,
          "validation_loss": 0.012048013508319855,
          "train_metric": 0.0014247746439650655,
          "validation_metric": 0.0012134317075833678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6934624120003718
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.021914949116762727,
          "train_loss": 0.0134446294978261,
          "validation_loss": 0.011598327197134495,
          "train_metric": 0.0013540936633944511,
          "validation_metric": 0.001168140908703208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.712286234000203
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.02432641328778118,
          "train_loss": 0.014958390966057777,
          "validation_loss": 0.011897426098585129,
          "train_metric": 0.0015065540792420506,
          "validation_metric": 0.0011982652358710766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7300811580007576
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.026866638392675668,
          "train_loss": 0.013413868844509125,
          "validation_loss": 0.010356653481721878,
          "train_metric": 0.0013509958516806364,
          "validation_metric": 0.0010430841939523816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7482433320001292
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.02631628717062995,
          "train_loss": 0.011885719373822212,
          "validation_loss": 0.009980037808418274,
          "train_metric": 0.0011970860650762916,
          "validation_metric": 0.001005152938887477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7663732080000045
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.024661167117301375,
          "train_loss": 0.01256015244871378,
          "validation_loss": 0.01124526746571064,
          "train_metric": 0.0012650126591324806,
          "validation_metric": 0.0011325820814818144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7838967459993
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.02727387734921649,
          "train_loss": 0.01424884982407093,
          "validation_loss": 0.011644882149994373,
          "train_metric": 0.0014350919518619776,
          "validation_metric": 0.0011728297686204314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.801723481999943
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.026935845206025988,
          "train_loss": 0.014068319462239742,
          "validation_loss": 0.01117042638361454,
          "train_metric": 0.0014169096248224378,
          "validation_metric": 0.0011250443058088422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8191769589993783
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.026023496233392507,
          "train_loss": 0.01722739078104496,
          "validation_loss": 0.012030339799821377,
          "train_metric": 0.001735079800710082,
          "validation_metric": 0.0012116518337279558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8369564519998676
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.027643650653772056,
          "train_loss": 0.016534997150301933,
          "validation_loss": 0.011877214536070824,
          "train_metric": 0.001665344461798668,
          "validation_metric": 0.0011962293647229671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8550463989995478
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.024333584122359753,
          "train_loss": 0.011561717838048935,
          "validation_loss": 0.0088661452755332,
          "train_metric": 0.0011644538026303053,
          "validation_metric": 0.0008929655887186527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8740896420003992
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.023406369029544294,
          "train_loss": 0.010144304484128952,
          "validation_loss": 0.00872817449271679,
          "train_metric": 0.00102169718593359,
          "validation_metric": 0.0008790696738287807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8930044480002834
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.023102430335711688,
          "train_loss": 0.013495366089046001,
          "validation_loss": 0.01061217300593853,
          "train_metric": 0.0013592037139460444,
          "validation_metric": 0.0010688191978260875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.910987686000226
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.022617280716076493,
          "train_loss": 0.011991484090685844,
          "validation_loss": 0.00908802542835474,
          "train_metric": 0.001207738183438778,
          "validation_metric": 0.0009153125574812293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9290861150002456
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.02505373692838475,
          "train_loss": 0.016572101041674614,
          "validation_loss": 0.011399833485484123,
          "train_metric": 0.0016690813936293125,
          "validation_metric": 0.0011481496039777994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.946989553000094
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.02567390038166195,
          "train_loss": 0.01042377669364214,
          "validation_loss": 0.008263282477855682,
          "train_metric": 0.0010498446645215154,
          "validation_metric": 0.0008322474895976484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9657418499991763
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.027640147833153605,
          "train_loss": 0.020625600591301918,
          "validation_loss": 0.015625974163413048,
          "train_metric": 0.002077335026115179,
          "validation_metric": 0.0015737909125164151,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9834565239998483
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.024174668302293867,
          "train_loss": 0.013222146779298782,
          "validation_loss": 0.011149827390909195,
          "train_metric": 0.0013316861586645246,
          "validation_metric": 0.0011229697847738862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.001550810999106
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.02204036939656362,
          "train_loss": 0.010507611557841301,
          "validation_loss": 0.009330801665782928,
          "train_metric": 0.0010582881513983011,
          "validation_metric": 0.0009397641406394541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.019176764999429
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.021406468586064875,
          "train_loss": 0.013546832837164402,
          "validation_loss": 0.01130546722561121,
          "train_metric": 0.0013643874553963542,
          "validation_metric": 0.0011386452242732048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0369141079991095
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.022530790127348155,
          "train_loss": 0.009077356196939945,
          "validation_loss": 0.0077413711696863174,
          "train_metric": 0.0009142380440607667,
          "validation_metric": 0.0007796824793331325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0559941840001557
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.02324520645197481,
          "train_loss": 0.009412314742803574,
          "validation_loss": 0.0080995112657547,
          "train_metric": 0.0009479739237576723,
          "validation_metric": 0.0008157531265169382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0744741300004534
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.020196903264150023,
          "train_loss": 0.010063853114843369,
          "validation_loss": 0.007303474470973015,
          "train_metric": 0.0010135945631191134,
          "validation_metric": 0.0007355791749432683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0930084249994252
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.019791924743913114,
          "train_loss": 0.01025882177054882,
          "validation_loss": 0.0076626078225672245,
          "train_metric": 0.0010332308011129498,
          "validation_metric": 0.0007717497064732015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1109105449995695
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.021556858089752495,
          "train_loss": 0.009640303440392017,
          "validation_loss": 0.007180613465607166,
          "train_metric": 0.0009709360310807824,
          "validation_metric": 0.0007232051575556397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1335726489996887
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.02105795475654304,
          "train_loss": 0.009602770209312439,
          "validation_loss": 0.007597131188958883,
          "train_metric": 0.00096715590916574,
          "validation_metric": 0.000765155185945332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1520218069999828
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.02093443158082664,
          "train_loss": 0.0090787960216403,
          "validation_loss": 0.006801752373576164,
          "train_metric": 0.0009143830975517631,
          "validation_metric": 0.0006850475911051035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1702974710005947
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.02030666306382045,
          "train_loss": 0.010088932700455189,
          "validation_loss": 0.007247159257531166,
          "train_metric": 0.0010161201935261488,
          "validation_metric": 0.0007299073040485382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1880098260007799
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.019531450583599508,
          "train_loss": 0.007837418466806412,
          "validation_loss": 0.006240725051611662,
          "train_metric": 0.0007893560687080026,
          "validation_metric": 0.0006285429699346423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.207660197000223
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.021013393066823483,
          "train_loss": 0.008740782737731934,
          "validation_loss": 0.008105468936264515,
          "train_metric": 0.0008803396485745907,
          "validation_metric": 0.0008163531892932951,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2256230050006707
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.017937281052581966,
          "train_loss": 0.00763807725161314,
          "validation_loss": 0.006069784983992577,
          "train_metric": 0.0007692790823057294,
          "validation_metric": 0.0006113264826126397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2442122909997124
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.018298963317647576,
          "train_loss": 0.01080680638551712,
          "validation_loss": 0.00742223858833313,
          "train_metric": 0.0010884220246225595,
          "validation_metric": 0.0007475407910533249,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2622395590005908
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.019683609949424863,
          "train_loss": 0.008038625121116638,
          "validation_loss": 0.006383425556123257,
          "train_metric": 0.0008096209494397044,
          "validation_metric": 0.0006429153145290911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2799785539991717
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.018066456133965403,
          "train_loss": 0.007259288802742958,
          "validation_loss": 0.005732705816626549,
          "train_metric": 0.0007311290828511119,
          "validation_metric": 0.0005773772136308253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2987629399995058
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.019515486375894397,
          "train_loss": 0.007699925918132067,
          "validation_loss": 0.005584304220974445,
          "train_metric": 0.0007755083497613668,
          "validation_metric": 0.0005624305922538042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3173287480003637
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.01787458616308868,
          "train_loss": 0.011057010851800442,
          "validation_loss": 0.006723417900502682,
          "train_metric": 0.0011136215180158615,
          "validation_metric": 0.0006771580665372312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.335383723000632
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.02180626941844821,
          "train_loss": 0.010956108570098877,
          "validation_loss": 0.0067098443396389484,
          "train_metric": 0.001103459158912301,
          "validation_metric": 0.0006757910014130175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3526800599993294
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.01840860361699015,
          "train_loss": 0.007662320975214243,
          "validation_loss": 0.00630244892090559,
          "train_metric": 0.0007717208936810493,
          "validation_metric": 0.0006347596645355225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3697041359992
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.01668613002402708,
          "train_loss": 0.008338955231010914,
          "validation_loss": 0.006775938905775547,
          "train_metric": 0.0008398689678870142,
          "validation_metric": 0.0006824476877227426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.387388237999403
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.01945177186280489,
          "train_loss": 0.009538927115499973,
          "validation_loss": 0.008998853154480457,
          "train_metric": 0.0009607258252799511,
          "validation_metric": 0.0009063315810635686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4071066699998482
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.018977137398906052,
          "train_loss": 0.006903631146997213,
          "validation_loss": 0.00543606374412775,
          "train_metric": 0.0006953083793632686,
          "validation_metric": 0.0005475004436448216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.425832786000683
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.019605003588367254,
          "train_loss": 0.008795623667538166,
          "validation_loss": 0.0066111949272453785,
          "train_metric": 0.0008858630317263305,
          "validation_metric": 0.0006658554193563759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.444032846000482
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.016580469033215195,
          "train_loss": 0.00872507132589817,
          "validation_loss": 0.005698170512914658,
          "train_metric": 0.0008787572151049972,
          "validation_metric": 0.0005738988984376192,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4617090379997535
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.018103825219441205,
          "train_loss": 0.007456119637936354,
          "validation_loss": 0.005424278788268566,
          "train_metric": 0.000750953156966716,
          "validation_metric": 0.0005463134730234742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4803943780007103
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.01871454482898116,
          "train_loss": 0.009934679605066776,
          "validation_loss": 0.005811847280710936,
          "train_metric": 0.0010005844524130225,
          "validation_metric": 0.0005853478796780109,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4985637230001885
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.019465037737973034,
          "train_loss": 0.007299418561160564,
          "validation_loss": 0.005681312642991543,
          "train_metric": 0.0007351707899942994,
          "validation_metric": 0.0005722010391764343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5159355159994448
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.018788114946801215,
          "train_loss": 0.006409605965018272,
          "validation_loss": 0.0053343940526247025,
          "train_metric": 0.0006455520633608103,
          "validation_metric": 0.0005372606101445854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5348619060005149
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.016899113601539284,
          "train_loss": 0.0074028619565069675,
          "validation_loss": 0.0051732915453612804,
          "train_metric": 0.0007455892045982182,
          "validation_metric": 0.0005210349336266518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5538774329997977
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.021278974425513297,
          "train_loss": 0.012638125568628311,
          "validation_loss": 0.01164862047880888,
          "train_metric": 0.001272865803912282,
          "validation_metric": 0.0011732064886018634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5728750199996284
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.019644636660814285,
          "train_loss": 0.011931583285331726,
          "validation_loss": 0.008506420068442822,
          "train_metric": 0.0012017053086310625,
          "validation_metric": 0.0008567354525439441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5912001069991675
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.022068485559429973,
          "train_loss": 0.012266822159290314,
          "validation_loss": 0.007077292539179325,
          "train_metric": 0.001235469477251172,
          "validation_metric": 0.0007127989083528519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.610222949000672
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.02047435031272471,
          "train_loss": 0.009292913600802422,
          "validation_loss": 0.005733969155699015,
          "train_metric": 0.0009359482210129499,
          "validation_metric": 0.0005775044555775821,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.627926920000391
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.019306515634525567,
          "train_loss": 0.0062831672839820385,
          "validation_loss": 0.0058004846796393394,
          "train_metric": 0.0006328175659291446,
          "validation_metric": 0.0005842035752721131,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6464534679998906
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.01865815673954785,
          "train_loss": 0.008200012147426605,
          "validation_loss": 0.006499440409243107,
          "train_metric": 0.0008258752641268075,
          "validation_metric": 0.0006545998621731997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6657965680005873
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.017181594448629767,
          "train_loss": 0.0066528404131531715,
          "validation_loss": 0.004430283326655626,
          "train_metric": 0.0006700497469864786,
          "validation_metric": 0.0004462019132915884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.684788266000396
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.018069337646011263,
          "train_loss": 0.009827774949371815,
          "validation_loss": 0.006115525495260954,
          "train_metric": 0.000989817432127893,
          "validation_metric": 0.0006159332115203142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7035938280005212
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.01879103557439521,
          "train_loss": 0.018561437726020813,
          "validation_loss": 0.011640934273600578,
          "train_metric": 0.0018694400787353516,
          "validation_metric": 0.0011724323267117143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7280010800004675
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.019976833660621196,
          "train_loss": 0.008559084497392178,
          "validation_loss": 0.005654994864016771,
          "train_metric": 0.0008620396256446838,
          "validation_metric": 0.0005695504369214177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.746336958000029
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.017195260210428387,
          "train_loss": 0.012580852024257183,
          "validation_loss": 0.0131684560328722,
          "train_metric": 0.0012670973083004355,
          "validation_metric": 0.0013262786669656634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.763918783999543
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.01903111004503444,
          "train_loss": 0.006611810997128487,
          "validation_loss": 0.0058419471606612206,
          "train_metric": 0.0006659173523075879,
          "validation_metric": 0.0005883795092813671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7814030200006528
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.017295465746428818,
          "train_loss": 0.00592816574499011,
          "validation_loss": 0.005759972147643566,
          "train_metric": 0.0005970631027594209,
          "validation_metric": 0.0005801233346574008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7989561590002268
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.017221618443727493,
          "train_loss": 0.009105207398533821,
          "validation_loss": 0.006353578995913267,
          "train_metric": 0.0009170431294478476,
          "validation_metric": 0.0006399091216735542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8343789329992433
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.01703184563666582,
          "train_loss": 0.006973897572606802,
          "validation_loss": 0.005452718120068312,
          "train_metric": 0.0007023853831924498,
          "validation_metric": 0.000549177813809365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8545692150000832
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.016947621596045792,
          "train_loss": 0.00719565711915493,
          "validation_loss": 0.004526908043771982,
          "train_metric": 0.0007247201865538955,
          "validation_metric": 0.00045593359391205013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8703842939994502
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.01844858581898734,
          "train_loss": 0.008429322391748428,
          "validation_loss": 0.007068062666803598,
          "train_metric": 0.0008489704923704267,
          "validation_metric": 0.0007118693902157247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8924604900003033
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.01863090042024851,
          "train_loss": 0.007956110872328281,
          "validation_loss": 0.005781175103038549,
          "train_metric": 0.0008013102924451232,
          "validation_metric": 0.0005822587409056723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9133601779994933
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.019440190633758903,
          "train_loss": 0.01673116721212864,
          "validation_loss": 0.015258179046213627,
          "train_metric": 0.001685102004557848,
          "validation_metric": 0.00153674790635705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9289962839993677
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.021661274367943406,
          "train_loss": 0.006357107777148485,
          "validation_loss": 0.004624702502042055,
          "train_metric": 0.0006402645958587527,
          "validation_metric": 0.00046578311594203115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9468461239994213
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.016578868438955396,
          "train_loss": 0.0057359118945896626,
          "validation_loss": 0.00511936517432332,
          "train_metric": 0.0005777000915259123,
          "validation_metric": 0.000515603693202138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9711607730005198
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.015052577218739316,
          "train_loss": 0.007158917374908924,
          "validation_loss": 0.0049996087327599525,
          "train_metric": 0.0007210199837572873,
          "validation_metric": 0.0005035422509536147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9879510840000876
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.015871884068474174,
          "train_loss": 0.012206700630486012,
          "validation_loss": 0.00986306183040142,
          "train_metric": 0.0012294142507016659,
          "validation_metric": 0.0009933713590726256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0013408020004135
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.019469881430268288,
          "train_loss": 0.005782227497547865,
          "validation_loss": 0.004364361986517906,
          "train_metric": 0.0005823648534715176,
          "validation_metric": 0.0004395625728648156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0144560570006433
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.017301712068729103,
          "train_loss": 0.005560537800192833,
          "validation_loss": 0.004470979794859886,
          "train_metric": 0.0005600369768217206,
          "validation_metric": 0.00045030072215013206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0291985749991
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.017352883296553046,
          "train_loss": 0.011703734286129475,
          "validation_loss": 0.007391922641545534,
          "train_metric": 0.0011787572875618935,
          "validation_metric": 0.0007444874499924481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.042730875000416
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.016634171246550977,
          "train_loss": 0.006071698386222124,
          "validation_loss": 0.005393782164901495,
          "train_metric": 0.0006115192081779242,
          "validation_metric": 0.0005432419711723924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0581914840004174
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.017549286945722997,
          "train_loss": 0.00603463314473629,
          "validation_loss": 0.004012263845652342,
          "train_metric": 0.000607786001637578,
          "validation_metric": 0.00040410051587969065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0718459660001827
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.017830200027674437,
          "train_loss": 0.0066477651707828045,
          "validation_loss": 0.005131111014634371,
          "train_metric": 0.0006695386837236583,
          "validation_metric": 0.0005167867057025433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.085194691000652
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0182557042571716,
          "train_loss": 0.005529055837541819,
          "validation_loss": 0.004940563812851906,
          "train_metric": 0.0005568662891164422,
          "validation_metric": 0.0004975955234840512,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0974958169990714
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.016952190781012177,
          "train_loss": 0.007626613602042198,
          "validation_loss": 0.007987825199961662,
          "train_metric": 0.0007681247079744935,
          "validation_metric": 0.0008045045542530715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1108811630001583
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.014420553983654827,
          "train_loss": 0.007310741115361452,
          "validation_loss": 0.005573198199272156,
          "train_metric": 0.0007363110198639333,
          "validation_metric": 0.000561312073841691,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1238465679998626
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.01599524141056463,
          "train_loss": 0.007114170119166374,
          "validation_loss": 0.00422412995249033,
          "train_metric": 0.0007165131974034011,
          "validation_metric": 0.0004254389205016196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1370010690006893
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.014913299237377942,
          "train_loss": 0.005638767033815384,
          "validation_loss": 0.003669400466606021,
          "train_metric": 0.0005679159658029675,
          "validation_metric": 0.00036956858821213245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.151272799999788
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.018144745408790186,
          "train_loss": 0.0053519876673817635,
          "validation_loss": 0.005202109459787607,
          "train_metric": 0.0005390326259657741,
          "validation_metric": 0.0005239374586381018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.164069305999874
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.016974193509668112,
          "train_loss": 0.00619234936311841,
          "validation_loss": 0.006244798190891743,
          "train_metric": 0.0006236707558855414,
          "validation_metric": 0.0006289531593210995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.191340310000669
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.018064319505356252,
          "train_loss": 0.018050266429781914,
          "validation_loss": 0.013582068495452404,
          "train_metric": 0.0018179569160565734,
          "validation_metric": 0.0013679363764822483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2145224750001944
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.020046867022756487,
          "train_loss": 0.006790748797357082,
          "validation_loss": 0.0036667422391474247,
          "train_metric": 0.0006839394336566329,
          "validation_metric": 0.0003693008329719305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2413766289992054
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.018917885900009423,
          "train_loss": 0.00823942944407463,
          "validation_loss": 0.005194006487727165,
          "train_metric": 0.0008298450848087668,
          "validation_metric": 0.0005231213290244341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2660806879994198
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.01639147539390251,
          "train_loss": 0.011266800574958324,
          "validation_loss": 0.009115681052207947,
          "train_metric": 0.0011347507825121284,
          "validation_metric": 0.0009180980850942433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.279598821000036
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.01662985677830875,
          "train_loss": 0.004842035472393036,
          "validation_loss": 0.0035049240104854107,
          "train_metric": 0.00048767210682854056,
          "validation_metric": 0.000353003095369786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.295355144000496
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.014359217719174922,
          "train_loss": 0.005324976518750191,
          "validation_loss": 0.004433084279298782,
          "train_metric": 0.0005363120581023395,
          "validation_metric": 0.0004464839876163751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3094234870004584
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.017273081291932613,
          "train_loss": 0.008421655744314194,
          "validation_loss": 0.005488785449415445,
          "train_metric": 0.0008481983677484095,
          "validation_metric": 0.0005528103793039918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.326186147000044
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.017822682624682784,
          "train_loss": 0.004909242037683725,
          "validation_loss": 0.004091611597687006,
          "train_metric": 0.000494440901093185,
          "validation_metric": 0.0004120921657886356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3402429269990535
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.015334415365941823,
          "train_loss": 0.0065027265809476376,
          "validation_loss": 0.004265469498932362,
          "train_metric": 0.0006549308309331536,
          "validation_metric": 0.000429602456279099,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3558296730006987
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.016915058280574158,
          "train_loss": 0.008866759017109871,
          "validation_loss": 0.004506326280534267,
          "train_metric": 0.0008930276962928474,
          "validation_metric": 0.00045386067358776927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3703779139996186
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.017618968908209354,
          "train_loss": 0.004726205952465534,
          "validation_loss": 0.003843548009172082,
          "train_metric": 0.0004760061274282634,
          "validation_metric": 0.00038710812805220485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3845693679995748
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.01602339744567871,
          "train_loss": 0.007086019963026047,
          "validation_loss": 0.00608662236481905,
          "train_metric": 0.000713677960447967,
          "validation_metric": 0.0006130224210210145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.397270022000157
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.015200823952909559,
          "train_loss": 0.005369020625948906,
          "validation_loss": 0.003106713993474841,
          "train_metric": 0.0005407481803558767,
          "validation_metric": 0.0003128968528471887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4121693029992457
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.015454323787707835,
          "train_loss": 0.006941275671124458,
          "validation_loss": 0.004232142120599747,
          "train_metric": 0.0006990998517721891,
          "validation_metric": 0.000426245853304863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.428134300000238
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.015008258167654276,
          "train_loss": 0.004408265929669142,
          "validation_loss": 0.003436828264966607,
          "train_metric": 0.00044398443424142897,
          "validation_metric": 0.0003461447777226567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.441552005999256
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.017664609709754586,
          "train_loss": 0.0055632819421589375,
          "validation_loss": 0.0054558040574193,
          "train_metric": 0.0005603134050033987,
          "validation_metric": 0.000549488642718643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4554800480000267
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.015864765795413405,
          "train_loss": 0.009670838713645935,
          "validation_loss": 0.006022611167281866,
          "train_metric": 0.0009740115492604673,
          "validation_metric": 0.0006065753404982388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4719317689996387
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.019040947139728814,
          "train_loss": 0.00609600031748414,
          "validation_loss": 0.00611207215115428,
          "train_metric": 0.0006139667821116745,
          "validation_metric": 0.000615585595369339,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.494833304999702
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.016749725909903646,
          "train_loss": 0.0062386528588831425,
          "validation_loss": 0.004447597544640303,
          "train_metric": 0.0006283342372626066,
          "validation_metric": 0.000447945756604895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.513343837999855
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.016304465156281367,
          "train_loss": 0.004955274984240532,
          "validation_loss": 0.004576437175273895,
          "train_metric": 0.0004990771994926035,
          "validation_metric": 0.00046092201955616474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.577144318000137
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.017841664550360292,
          "train_loss": 0.00475727953016758,
          "validation_loss": 0.004149561747908592,
          "train_metric": 0.0004791357205249369,
          "validation_metric": 0.00041792873525992036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.613046950000353
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.014802760211750865,
          "train_loss": 0.004767913371324539,
          "validation_loss": 0.0036692689172923565,
          "train_metric": 0.0004802067414857447,
          "validation_metric": 0.00036955525865778327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6480172789997596
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.014020418166182935,
          "train_loss": 0.004499323666095734,
          "validation_loss": 0.004226243123412132,
          "train_metric": 0.0004531554295681417,
          "validation_metric": 0.00042565169860608876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6796422820007137
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.016049997590016574,
          "train_loss": 0.005794101860374212,
          "validation_loss": 0.0029216203838586807,
          "train_metric": 0.0005835607880726457,
          "validation_metric": 0.00029425485990941525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7121916899995995
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.014419444574741647,
          "train_loss": 0.007175138220191002,
          "validation_loss": 0.004519188776612282,
          "train_metric": 0.0007226535817608237,
          "validation_metric": 0.0004551562014967203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7311967969999387
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.016121945285703987,
          "train_loss": 0.00918841827660799,
          "validation_loss": 0.004332445096224546,
          "train_metric": 0.0009254237520508468,
          "validation_metric": 0.0004363480256870389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7607664440001827
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.01592617854475975,
          "train_loss": 0.005504071246832609,
          "validation_loss": 0.0036384558770805597,
          "train_metric": 0.0005543498555198312,
          "validation_metric": 0.00036645197542384267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7797967080005037
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.015000268525909632,
          "train_loss": 0.004971137270331383,
          "validation_loss": 0.005475563928484917,
          "train_metric": 0.0005006747087463737,
          "validation_metric": 0.0005514787044376135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7985898019996966
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.013514646503608674,
          "train_loss": 0.004472264554351568,
          "validation_loss": 0.0031378522980958223,
          "train_metric": 0.0004504301177803427,
          "validation_metric": 0.0003160329652018845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8209669740008394
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0146804066025652,
          "train_loss": 0.0057885898277163506,
          "validation_loss": 0.003305652178823948,
          "train_metric": 0.000583005603402853,
          "validation_metric": 0.0003329331520944834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.842250151999906
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.013302329345606267,
          "train_loss": 0.007909530773758888,
          "validation_loss": 0.0045564924366772175,
          "train_metric": 0.0007966189878061414,
          "validation_metric": 0.0004589132440742105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8586886920002144
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.01940641406690702,
          "train_loss": 0.012036203406751156,
          "validation_loss": 0.007200996391475201,
          "train_metric": 0.001212242292240262,
          "validation_metric": 0.000725257967133075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.872842696000589
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.024024564016144723,
          "train_loss": 0.007734923157840967,
          "validation_loss": 0.004106452688574791,
          "train_metric": 0.0007790331728756428,
          "validation_metric": 0.0004135868512094021,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.885623095000483
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.015633541566785425,
          "train_loss": 0.004343742039054632,
          "validation_loss": 0.0034587783738970757,
          "train_metric": 0.000437485781731084,
          "validation_metric": 0.0003483555337879807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.897751036000045
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.013649746018927544,
          "train_loss": 0.004424244165420532,
          "validation_loss": 0.0044363755732774734,
          "train_metric": 0.0004455936432350427,
          "validation_metric": 0.00044681550934910774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9120872219991725
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.014646199881099164,
          "train_loss": 0.005990093108266592,
          "validation_loss": 0.004853857681155205,
          "train_metric": 0.0006033002864569426,
          "validation_metric": 0.0004888628027401865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9282557370006543
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.014597769943065941,
          "train_loss": 0.004406401421874762,
          "validation_loss": 0.003944097552448511,
          "train_metric": 0.0004437965399120003,
          "validation_metric": 0.000397235038690269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.942712018999373
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.013954876223579049,
          "train_loss": 0.003962237387895584,
          "validation_loss": 0.0036142910830676556,
          "train_metric": 0.00039906203164719045,
          "validation_metric": 0.0003640181676018983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.958405187999233
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.013834504294209182,
          "train_loss": 0.003653448773548007,
          "validation_loss": 0.00305464887060225,
          "train_metric": 0.0003679619694594294,
          "validation_metric": 0.00030765304109081626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9734815709998657
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.015113310189917684,
          "train_loss": 0.006407046224921942,
          "validation_loss": 0.0034614531323313713,
          "train_metric": 0.0006452941452153027,
          "validation_metric": 0.00034862488973885775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9874407650004287
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.01438681106083095,
          "train_loss": 0.004624280147254467,
          "validation_loss": 0.003704034024849534,
          "train_metric": 0.0004657405079342425,
          "validation_metric": 0.00037305676960386336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0009159239998553
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.014431917807087302,
          "train_loss": 0.007337099872529507,
          "validation_loss": 0.003445536596700549,
          "train_metric": 0.0007389657548628747,
          "validation_metric": 0.00034702179254963994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.014015151999047
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.017032452684361488,
          "train_loss": 0.010377952829003334,
          "validation_loss": 0.006912250071763992,
          "train_metric": 0.0010452293790876865,
          "validation_metric": 0.0006961764884181321,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0278593029997865
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.017030167451594025,
          "train_loss": 0.0043317535892128944,
          "validation_loss": 0.0038515287451446056,
          "train_metric": 0.0004362783511169255,
          "validation_metric": 0.00038791188853792846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0424350689991115
        }
      ],
      "best": {
        "epoch": 144,
        "updates": 2304
      },
      "best_validation_metric": 0.00029425485990941525,
      "config": {
        "label": "Dropout 0.1",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.1,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01715385845544422
    },
    "92630a9c55d2_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 1,
          "samples_seen": 1024,
          "online_loss": 1.0258803367614746,
          "train_loss": 0.9930850267410278,
          "validation_loss": 0.7092403173446655,
          "train_metric": 0.10001988708972931,
          "validation_metric": 0.07143208384513855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.005988419999994221
        },
        {
          "epoch": 2,
          "updates": 2,
          "samples_seen": 2048,
          "online_loss": 0.9930850267410278,
          "train_loss": 0.9605929851531982,
          "validation_loss": 0.6880066990852356,
          "train_metric": 0.09674739837646484,
          "validation_metric": 0.06929351389408112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01184934500088275
        },
        {
          "epoch": 3,
          "updates": 3,
          "samples_seen": 3072,
          "online_loss": 0.9605928659439087,
          "train_loss": 0.9267858266830444,
          "validation_loss": 0.6659043431282043,
          "train_metric": 0.09334246814250946,
          "validation_metric": 0.06706744432449341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01772717600033502
        },
        {
          "epoch": 4,
          "updates": 4,
          "samples_seen": 4096,
          "online_loss": 0.9267858266830444,
          "train_loss": 0.8905212879180908,
          "validation_loss": 0.6421554684638977,
          "train_metric": 0.08969003707170486,
          "validation_metric": 0.06467554718255997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.023468453000532463
        },
        {
          "epoch": 5,
          "updates": 5,
          "samples_seen": 5120,
          "online_loss": 0.8905212879180908,
          "train_loss": 0.8509505987167358,
          "validation_loss": 0.6161958575248718,
          "train_metric": 0.08570463955402374,
          "validation_metric": 0.062060996890068054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02939556900128082
        },
        {
          "epoch": 6,
          "updates": 6,
          "samples_seen": 6144,
          "online_loss": 0.8509507179260254,
          "train_loss": 0.8074913620948792,
          "validation_loss": 0.587623655796051,
          "train_metric": 0.08132757246494293,
          "validation_metric": 0.05918330326676369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.034993397001017
        },
        {
          "epoch": 7,
          "updates": 7,
          "samples_seen": 7168,
          "online_loss": 0.8074913620948792,
          "train_loss": 0.7598076462745667,
          "validation_loss": 0.5561920404434204,
          "train_metric": 0.07652503252029419,
          "validation_metric": 0.05601762980222702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04064549300164799
        },
        {
          "epoch": 8,
          "updates": 8,
          "samples_seen": 8192,
          "online_loss": 0.7598075866699219,
          "train_loss": 0.7078191637992859,
          "validation_loss": 0.5218236446380615,
          "train_metric": 0.07128895819187164,
          "validation_metric": 0.05255616456270218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04640992900021956
        },
        {
          "epoch": 9,
          "updates": 9,
          "samples_seen": 9216,
          "online_loss": 0.7078192234039307,
          "train_loss": 0.6517389416694641,
          "validation_loss": 0.4846305251121521,
          "train_metric": 0.06564076244831085,
          "validation_metric": 0.048810213804244995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05213681800159975
        },
        {
          "epoch": 10,
          "updates": 10,
          "samples_seen": 10240,
          "online_loss": 0.6517389416694641,
          "train_loss": 0.5921094417572021,
          "validation_loss": 0.4449300765991211,
          "train_metric": 0.05963509902358055,
          "validation_metric": 0.04481172934174538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05799119300172606
        },
        {
          "epoch": 11,
          "updates": 11,
          "samples_seen": 11264,
          "online_loss": 0.5921094417572021,
          "train_loss": 0.5298457741737366,
          "validation_loss": 0.40326598286628723,
          "train_metric": 0.05336412787437439,
          "validation_metric": 0.04061547666788101,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06373000600069645
        },
        {
          "epoch": 12,
          "updates": 12,
          "samples_seen": 12288,
          "online_loss": 0.5298458337783813,
          "train_loss": 0.4663173258304596,
          "validation_loss": 0.36046072840690613,
          "train_metric": 0.04696577414870262,
          "validation_metric": 0.0363042838871479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06929700400178262
        },
        {
          "epoch": 13,
          "updates": 13,
          "samples_seen": 13312,
          "online_loss": 0.4663173258304596,
          "train_loss": 0.4034489691257477,
          "validation_loss": 0.31767868995666504,
          "train_metric": 0.0406339056789875,
          "validation_metric": 0.03199543431401253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07507266300126503
        },
        {
          "epoch": 14,
          "updates": 14,
          "samples_seen": 14336,
          "online_loss": 0.40344899892807007,
          "train_loss": 0.343758225440979,
          "validation_loss": 0.27644097805023193,
          "train_metric": 0.03462206944823265,
          "validation_metric": 0.027842121198773384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08082416900106182
        },
        {
          "epoch": 15,
          "updates": 15,
          "samples_seen": 15360,
          "online_loss": 0.343758225440979,
          "train_loss": 0.2902417480945587,
          "validation_loss": 0.23856022953987122,
          "train_metric": 0.0292320866137743,
          "validation_metric": 0.024026911705732346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08664289100124734
        },
        {
          "epoch": 16,
          "updates": 16,
          "samples_seen": 16384,
          "online_loss": 0.2902417480945587,
          "train_loss": 0.2460261732339859,
          "validation_loss": 0.20596328377723694,
          "train_metric": 0.024778854101896286,
          "validation_metric": 0.02074386551976204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17820527300136746
        },
        {
          "epoch": 17,
          "updates": 17,
          "samples_seen": 17408,
          "online_loss": 0.2460261732339859,
          "train_loss": 0.21371406316757202,
          "validation_loss": 0.18037617206573486,
          "train_metric": 0.021524500101804733,
          "validation_metric": 0.01816682517528534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18399549900095735
        },
        {
          "epoch": 18,
          "updates": 18,
          "samples_seen": 18432,
          "online_loss": 0.21371407806873322,
          "train_loss": 0.19444365799427032,
          "validation_loss": 0.16286668181419373,
          "train_metric": 0.019583653658628464,
          "validation_metric": 0.016403336077928543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18970795000132057
        },
        {
          "epoch": 19,
          "updates": 19,
          "samples_seen": 19456,
          "online_loss": 0.19444367289543152,
          "train_loss": 0.1870141327381134,
          "validation_loss": 0.15339042246341705,
          "train_metric": 0.01883537508547306,
          "validation_metric": 0.0154489204287529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19561709000117844
        },
        {
          "epoch": 20,
          "updates": 20,
          "samples_seen": 20480,
          "online_loss": 0.18701410293579102,
          "train_loss": 0.1878947913646698,
          "validation_loss": 0.15067341923713684,
          "train_metric": 0.01892407238483429,
          "validation_metric": 0.015175273641943932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20123453800079005
        },
        {
          "epoch": 21,
          "updates": 21,
          "samples_seen": 21504,
          "online_loss": 0.1878947764635086,
          "train_loss": 0.19239771366119385,
          "validation_loss": 0.15249933302402496,
          "train_metric": 0.01937759481370449,
          "validation_metric": 0.015359176322817802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20635724700150604
        },
        {
          "epoch": 22,
          "updates": 22,
          "samples_seen": 22528,
          "online_loss": 0.19239771366119385,
          "train_loss": 0.19602666795253754,
          "validation_loss": 0.1560187041759491,
          "train_metric": 0.019743088632822037,
          "validation_metric": 0.015713632106781006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21162866600025154
        },
        {
          "epoch": 23,
          "updates": 23,
          "samples_seen": 23552,
          "online_loss": 0.19602665305137634,
          "train_loss": 0.19552338123321533,
          "validation_loss": 0.15826033055782318,
          "train_metric": 0.019692396745085716,
          "validation_metric": 0.01593940332531929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21684722500140197
        },
        {
          "epoch": 24,
          "updates": 24,
          "samples_seen": 24576,
          "online_loss": 0.19552338123321533,
          "train_loss": 0.18979987502098083,
          "validation_loss": 0.15719343721866608,
          "train_metric": 0.01911594718694687,
          "validation_metric": 0.015831949189305305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22192133400130842
        },
        {
          "epoch": 25,
          "updates": 25,
          "samples_seen": 25600,
          "online_loss": 0.18979986011981964,
          "train_loss": 0.17997907102108002,
          "validation_loss": 0.15244463086128235,
          "train_metric": 0.0181268323212862,
          "validation_metric": 0.015353664755821228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2271248060005746
        },
        {
          "epoch": 26,
          "updates": 26,
          "samples_seen": 26624,
          "online_loss": 0.17997907102108002,
          "train_loss": 0.168293759226799,
          "validation_loss": 0.14496231079101562,
          "train_metric": 0.016949929296970367,
          "validation_metric": 0.014600072987377644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23271503700016183
        },
        {
          "epoch": 27,
          "updates": 27,
          "samples_seen": 27648,
          "online_loss": 0.168293759226799,
          "train_loss": 0.1568431854248047,
          "validation_loss": 0.13618676364421844,
          "train_metric": 0.01579667069017887,
          "validation_metric": 0.013716231100261211,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23850246100118966
        },
        {
          "epoch": 28,
          "updates": 28,
          "samples_seen": 28672,
          "online_loss": 0.1568431854248047,
          "train_loss": 0.1469542235136032,
          "validation_loss": 0.12741924822330475,
          "train_metric": 0.014800691977143288,
          "validation_metric": 0.01283320039510727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2445274830006383
        },
        {
          "epoch": 29,
          "updates": 29,
          "samples_seen": 29696,
          "online_loss": 0.1469542235136032,
          "train_loss": 0.13911910355091095,
          "validation_loss": 0.11954168230295181,
          "train_metric": 0.014011567458510399,
          "validation_metric": 0.01203980017453432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2501660030011408
        },
        {
          "epoch": 30,
          "updates": 30,
          "samples_seen": 30720,
          "online_loss": 0.13911908864974976,
          "train_loss": 0.13322202861309052,
          "validation_loss": 0.11298336833715439,
          "train_metric": 0.013417634181678295,
          "validation_metric": 0.011379270814359188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25575066500096
        },
        {
          "epoch": 31,
          "updates": 31,
          "samples_seen": 31744,
          "online_loss": 0.13322202861309052,
          "train_loss": 0.12883873283863068,
          "validation_loss": 0.10782220959663391,
          "train_metric": 0.012976166792213917,
          "validation_metric": 0.010859457775950432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2615513290002127
        },
        {
          "epoch": 32,
          "updates": 32,
          "samples_seen": 32768,
          "online_loss": 0.12883874773979187,
          "train_loss": 0.12548963725566864,
          "validation_loss": 0.10392986983060837,
          "train_metric": 0.01263885572552681,
          "validation_metric": 0.010467435233294964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2673477530006494
        },
        {
          "epoch": 33,
          "updates": 33,
          "samples_seen": 33792,
          "online_loss": 0.12548963725566864,
          "train_loss": 0.12279529124498367,
          "validation_loss": 0.10109636932611465,
          "train_metric": 0.012367491610348225,
          "validation_metric": 0.01018205564469099,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27323774300020887
        },
        {
          "epoch": 34,
          "updates": 34,
          "samples_seen": 34816,
          "online_loss": 0.12279529869556427,
          "train_loss": 0.12052901089191437,
          "validation_loss": 0.0991029366850853,
          "train_metric": 0.012139239348471165,
          "validation_metric": 0.00998128391802311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27919134500007203
        },
        {
          "epoch": 35,
          "updates": 35,
          "samples_seen": 35840,
          "online_loss": 0.12052901834249496,
          "train_loss": 0.11858896911144257,
          "validation_loss": 0.09774117916822433,
          "train_metric": 0.011943846940994263,
          "validation_metric": 0.00984413456171751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28496990100029507
        },
        {
          "epoch": 36,
          "updates": 36,
          "samples_seen": 36864,
          "online_loss": 0.11858897656202316,
          "train_loss": 0.11692400276660919,
          "validation_loss": 0.09679676592350006,
          "train_metric": 0.011776156723499298,
          "validation_metric": 0.009749014861881733,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29056251300062286
        },
        {
          "epoch": 37,
          "updates": 37,
          "samples_seen": 37888,
          "online_loss": 0.1169239953160286,
          "train_loss": 0.11545105278491974,
          "validation_loss": 0.0960247591137886,
          "train_metric": 0.011627807281911373,
          "validation_metric": 0.009671262465417385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2962838390012621
        },
        {
          "epoch": 38,
          "updates": 38,
          "samples_seen": 38912,
          "online_loss": 0.11545105278491974,
          "train_loss": 0.11400263011455536,
          "validation_loss": 0.09514527022838593,
          "train_metric": 0.011481929570436478,
          "validation_metric": 0.009582682512700558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30195944900151517
        },
        {
          "epoch": 39,
          "updates": 39,
          "samples_seen": 39936,
          "online_loss": 0.11400264501571655,
          "train_loss": 0.11233367770910263,
          "validation_loss": 0.093878373503685,
          "train_metric": 0.011313837952911854,
          "validation_metric": 0.009455086663365364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3077007380015857
        },
        {
          "epoch": 40,
          "updates": 40,
          "samples_seen": 40960,
          "online_loss": 0.11233370006084442,
          "train_loss": 0.11019480228424072,
          "validation_loss": 0.09201671183109283,
          "train_metric": 0.011098417453467846,
          "validation_metric": 0.009267585352063179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31343037400074536
        },
        {
          "epoch": 41,
          "updates": 41,
          "samples_seen": 41984,
          "online_loss": 0.11019480228424072,
          "train_loss": 0.10744266957044601,
          "validation_loss": 0.08950686454772949,
          "train_metric": 0.010821232572197914,
          "validation_metric": 0.009014803916215897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3190202460009459
        },
        {
          "epoch": 42,
          "updates": 42,
          "samples_seen": 43008,
          "online_loss": 0.10744266211986542,
          "train_loss": 0.10413007438182831,
          "validation_loss": 0.08649515360593796,
          "train_metric": 0.010487599298357964,
          "validation_metric": 0.008711474947631359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32454259500082117
        },
        {
          "epoch": 43,
          "updates": 43,
          "samples_seen": 44032,
          "online_loss": 0.1041300818324089,
          "train_loss": 0.10052191466093063,
          "validation_loss": 0.08330486714839935,
          "train_metric": 0.010124199092388153,
          "validation_metric": 0.008390161208808422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3304775310007244
        },
        {
          "epoch": 44,
          "updates": 44,
          "samples_seen": 45056,
          "online_loss": 0.10052193701267242,
          "train_loss": 0.09702043980360031,
          "validation_loss": 0.08034425973892212,
          "train_metric": 0.009771543554961681,
          "validation_metric": 0.008091979660093784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3361281380002765
        },
        {
          "epoch": 45,
          "updates": 45,
          "samples_seen": 46080,
          "online_loss": 0.09702044725418091,
          "train_loss": 0.09403005242347717,
          "validation_loss": 0.07797850668430328,
          "train_metric": 0.009470362216234207,
          "validation_metric": 0.007853709161281586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34203962100036733
        },
        {
          "epoch": 46,
          "updates": 46,
          "samples_seen": 47104,
          "online_loss": 0.09403005242347717,
          "train_loss": 0.09181907027959824,
          "validation_loss": 0.07641331851482391,
          "train_metric": 0.009247681125998497,
          "validation_metric": 0.007696069777011871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.347801548001371
        },
        {
          "epoch": 47,
          "updates": 47,
          "samples_seen": 48128,
          "online_loss": 0.09181907773017883,
          "train_loss": 0.09043284505605698,
          "validation_loss": 0.07563193887472153,
          "train_metric": 0.009108063764870167,
          "validation_metric": 0.0076173716224730015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.353541224001674
        },
        {
          "epoch": 48,
          "updates": 48,
          "samples_seen": 49152,
          "online_loss": 0.09043283760547638,
          "train_loss": 0.08968853205442429,
          "validation_loss": 0.07540559768676758,
          "train_metric": 0.009033100679516792,
          "validation_metric": 0.007594576105475426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35929237800155533
        },
        {
          "epoch": 49,
          "updates": 49,
          "samples_seen": 50176,
          "online_loss": 0.08968853950500488,
          "train_loss": 0.08924926072359085,
          "validation_loss": 0.07537256926298141,
          "train_metric": 0.008988859131932259,
          "validation_metric": 0.007591248955577612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36493706700093753
        },
        {
          "epoch": 50,
          "updates": 50,
          "samples_seen": 51200,
          "online_loss": 0.08924926072359085,
          "train_loss": 0.08874537795782089,
          "validation_loss": 0.07515601813793182,
          "train_metric": 0.008938110433518887,
          "validation_metric": 0.007569437846541405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3706252050014882
        },
        {
          "epoch": 51,
          "updates": 51,
          "samples_seen": 52224,
          "online_loss": 0.08874538540840149,
          "train_loss": 0.08789462596178055,
          "validation_loss": 0.07447745651006699,
          "train_metric": 0.008852425031363964,
          "validation_metric": 0.007501095999032259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3764290940016508
        },
        {
          "epoch": 52,
          "updates": 52,
          "samples_seen": 53248,
          "online_loss": 0.08789463341236115,
          "train_loss": 0.08657757937908173,
          "validation_loss": 0.07322690635919571,
          "train_metric": 0.008719777688384056,
          "validation_metric": 0.0073751467280089855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3822875780006143
        },
        {
          "epoch": 53,
          "updates": 53,
          "samples_seen": 54272,
          "online_loss": 0.08657759428024292,
          "train_loss": 0.08484683930873871,
          "validation_loss": 0.0714704766869545,
          "train_metric": 0.008545463904738426,
          "validation_metric": 0.007198244798928499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3877202080002462
        },
        {
          "epoch": 54,
          "updates": 54,
          "samples_seen": 55296,
          "online_loss": 0.08484683930873871,
          "train_loss": 0.08287624269723892,
          "validation_loss": 0.06940144300460815,
          "train_metric": 0.008346992544829845,
          "validation_metric": 0.006989859510213137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39337068300119427
        },
        {
          "epoch": 55,
          "updates": 55,
          "samples_seen": 56320,
          "online_loss": 0.08287625759840012,
          "train_loss": 0.0808776393532753,
          "validation_loss": 0.06725942343473434,
          "train_metric": 0.00814569927752018,
          "validation_metric": 0.006774122826755047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3989928550017794
        },
        {
          "epoch": 56,
          "updates": 56,
          "samples_seen": 57344,
          "online_loss": 0.0808776393532753,
          "train_loss": 0.07902024686336517,
          "validation_loss": 0.065249502658844,
          "train_metric": 0.007958630099892616,
          "validation_metric": 0.0065716905519366264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4834740860005695
        },
        {
          "epoch": 57,
          "updates": 57,
          "samples_seen": 58368,
          "online_loss": 0.07902024686336517,
          "train_loss": 0.07738320529460907,
          "validation_loss": 0.06348907947540283,
          "train_metric": 0.007793753407895565,
          "validation_metric": 0.006394387222826481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4892245270002604
        },
        {
          "epoch": 58,
          "updates": 58,
          "samples_seen": 59392,
          "online_loss": 0.07738321274518967,
          "train_loss": 0.07595473527908325,
          "validation_loss": 0.061997175216674805,
          "train_metric": 0.007649882696568966,
          "validation_metric": 0.006244128104299307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49509376200148836
        },
        {
          "epoch": 59,
          "updates": 59,
          "samples_seen": 60416,
          "online_loss": 0.07595473527908325,
          "train_loss": 0.07466956228017807,
          "validation_loss": 0.06072187051177025,
          "train_metric": 0.007520444691181183,
          "validation_metric": 0.006115684751421213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5008553370007576
        },
        {
          "epoch": 60,
          "updates": 60,
          "samples_seen": 61440,
          "online_loss": 0.07466956973075867,
          "train_loss": 0.07345877587795258,
          "validation_loss": 0.05958591774106026,
          "train_metric": 0.007398498244583607,
          "validation_metric": 0.006001275032758713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5067036020009255
        },
        {
          "epoch": 61,
          "updates": 61,
          "samples_seen": 62464,
          "online_loss": 0.07345877587795258,
          "train_loss": 0.0722837820649147,
          "validation_loss": 0.05852556973695755,
          "train_metric": 0.007280157878994942,
          "validation_metric": 0.005894480738788843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5127242570015369
        },
        {
          "epoch": 62,
          "updates": 62,
          "samples_seen": 63488,
          "online_loss": 0.0722837895154953,
          "train_loss": 0.07114094495773315,
          "validation_loss": 0.057507358491420746,
          "train_metric": 0.007165056187659502,
          "validation_metric": 0.005791930481791496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5184492790012882
        },
        {
          "epoch": 63,
          "updates": 63,
          "samples_seen": 64512,
          "online_loss": 0.07114094495773315,
          "train_loss": 0.0700426921248436,
          "validation_loss": 0.056522827595472336,
          "train_metric": 0.007054443471133709,
          "validation_metric": 0.005692771635949612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5243919160002406
        },
        {
          "epoch": 64,
          "updates": 64,
          "samples_seen": 65536,
          "online_loss": 0.0700426921248436,
          "train_loss": 0.06899376958608627,
          "validation_loss": 0.055573031306266785,
          "train_metric": 0.006948800757527351,
          "validation_metric": 0.005597111769020557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5303150640011154
        },
        {
          "epoch": 65,
          "updates": 65,
          "samples_seen": 66560,
          "online_loss": 0.06899377703666687,
          "train_loss": 0.0679788887500763,
          "validation_loss": 0.05465605482459068,
          "train_metric": 0.0068465848453342915,
          "validation_metric": 0.005504758562892675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5359448390008765
        },
        {
          "epoch": 66,
          "updates": 66,
          "samples_seen": 67584,
          "online_loss": 0.06797889620065689,
          "train_loss": 0.06696764379739761,
          "validation_loss": 0.053764164447784424,
          "train_metric": 0.00674473587423563,
          "validation_metric": 0.005414930172264576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5415838150001946
        },
        {
          "epoch": 67,
          "updates": 67,
          "samples_seen": 68608,
          "online_loss": 0.06696764379739761,
          "train_loss": 0.06593137234449387,
          "validation_loss": 0.05288952589035034,
          "train_metric": 0.006640366278588772,
          "validation_metric": 0.005326839629560709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5476425600008952
        },
        {
          "epoch": 68,
          "updates": 68,
          "samples_seen": 69632,
          "online_loss": 0.06593137234449387,
          "train_loss": 0.06486069411039352,
          "validation_loss": 0.05203245207667351,
          "train_metric": 0.006532531231641769,
          "validation_metric": 0.005240518134087324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.553378683000119
        },
        {
          "epoch": 69,
          "updates": 69,
          "samples_seen": 70656,
          "online_loss": 0.06486069411039352,
          "train_loss": 0.06377396732568741,
          "validation_loss": 0.05120578035712242,
          "train_metric": 0.0064230808056890965,
          "validation_metric": 0.005157258361577988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5592590900014329
        },
        {
          "epoch": 70,
          "updates": 70,
          "samples_seen": 71680,
          "online_loss": 0.06377396732568741,
          "train_loss": 0.0627121701836586,
          "validation_loss": 0.05043190345168114,
          "train_metric": 0.006316140294075012,
          "validation_metric": 0.005079315975308418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.564829909000764
        },
        {
          "epoch": 71,
          "updates": 71,
          "samples_seen": 72704,
          "online_loss": 0.062712162733078,
          "train_loss": 0.06172265112400055,
          "validation_loss": 0.049733664840459824,
          "train_metric": 0.006216479931026697,
          "validation_metric": 0.005008992273360491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5704642130003776
        },
        {
          "epoch": 72,
          "updates": 72,
          "samples_seen": 73728,
          "online_loss": 0.06172265112400055,
          "train_loss": 0.06083950400352478,
          "validation_loss": 0.04912300035357475,
          "train_metric": 0.006127532105892897,
          "validation_metric": 0.0049474891275167465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5763023160016019
        },
        {
          "epoch": 73,
          "updates": 73,
          "samples_seen": 74752,
          "online_loss": 0.06083950772881508,
          "train_loss": 0.060069505125284195,
          "validation_loss": 0.048593468964099884,
          "train_metric": 0.006049980875104666,
          "validation_metric": 0.004894156940281391,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5822025250017759
        },
        {
          "epoch": 74,
          "updates": 74,
          "samples_seen": 75776,
          "online_loss": 0.060069505125284195,
          "train_loss": 0.05939045548439026,
          "validation_loss": 0.04812036082148552,
          "train_metric": 0.005981588736176491,
          "validation_metric": 0.004846506752073765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5880111170008604
        },
        {
          "epoch": 75,
          "updates": 75,
          "samples_seen": 76800,
          "online_loss": 0.05939045175909996,
          "train_loss": 0.0587625652551651,
          "validation_loss": 0.04766906425356865,
          "train_metric": 0.005918350536376238,
          "validation_metric": 0.004801054019480944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5942807810006343
        },
        {
          "epoch": 76,
          "updates": 76,
          "samples_seen": 77824,
          "online_loss": 0.0587625652551651,
          "train_loss": 0.05814670771360397,
          "validation_loss": 0.0472075492143631,
          "train_metric": 0.005856323521584272,
          "validation_metric": 0.004754571244120598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6001620710012503
        },
        {
          "epoch": 77,
          "updates": 77,
          "samples_seen": 78848,
          "online_loss": 0.05814670771360397,
          "train_loss": 0.05751989036798477,
          "validation_loss": 0.04671688377857208,
          "train_metric": 0.0057931928895413876,
          "validation_metric": 0.004705153405666351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.605991422000443
        },
        {
          "epoch": 78,
          "updates": 78,
          "samples_seen": 79872,
          "online_loss": 0.05751989036798477,
          "train_loss": 0.056880563497543335,
          "validation_loss": 0.046195000410079956,
          "train_metric": 0.005728801712393761,
          "validation_metric": 0.004652590956538916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6119782270016003
        },
        {
          "epoch": 79,
          "updates": 79,
          "samples_seen": 80896,
          "online_loss": 0.05688056722283363,
          "train_loss": 0.056242652237415314,
          "validation_loss": 0.045653123408555984,
          "train_metric": 0.005664553493261337,
          "validation_metric": 0.004598014988005161,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.617593424000006
        },
        {
          "epoch": 80,
          "updates": 80,
          "samples_seen": 81920,
          "online_loss": 0.056242648512125015,
          "train_loss": 0.05562387406826019,
          "validation_loss": 0.04510842263698578,
          "train_metric": 0.005602233111858368,
          "validation_metric": 0.004543154500424862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6231933030012442
        },
        {
          "epoch": 81,
          "updates": 81,
          "samples_seen": 82944,
          "online_loss": 0.055623870342969894,
          "train_loss": 0.055036187171936035,
          "validation_loss": 0.044577740132808685,
          "train_metric": 0.005543044302612543,
          "validation_metric": 0.004489706829190254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6290215630015155
        },
        {
          "epoch": 82,
          "updates": 82,
          "samples_seen": 83968,
          "online_loss": 0.055036190897226334,
          "train_loss": 0.054483119398355484,
          "validation_loss": 0.04407498240470886,
          "train_metric": 0.005487340502440929,
          "validation_metric": 0.004439070820808411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6356382920002943
        },
        {
          "epoch": 83,
          "updates": 83,
          "samples_seen": 84992,
          "online_loss": 0.054483115673065186,
          "train_loss": 0.05396325886249542,
          "validation_loss": 0.04361151158809662,
          "train_metric": 0.005434981547296047,
          "validation_metric": 0.0043923924677073956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.641223919001277
        },
        {
          "epoch": 84,
          "updates": 84,
          "samples_seen": 86016,
          "online_loss": 0.053963251411914825,
          "train_loss": 0.05347517877817154,
          "validation_loss": 0.04319652542471886,
          "train_metric": 0.005385824479162693,
          "validation_metric": 0.004350596107542515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6514206040010322
        },
        {
          "epoch": 85,
          "updates": 85,
          "samples_seen": 87040,
          "online_loss": 0.05347517877817154,
          "train_loss": 0.05301891639828682,
          "validation_loss": 0.04283514991402626,
          "train_metric": 0.005339870695024729,
          "validation_metric": 0.004314199555665255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6569861920015683
        },
        {
          "epoch": 86,
          "updates": 86,
          "samples_seen": 88064,
          "online_loss": 0.05301891267299652,
          "train_loss": 0.05259322375059128,
          "validation_loss": 0.04252493381500244,
          "train_metric": 0.005296996794641018,
          "validation_metric": 0.004282955080270767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6663448710005468
        },
        {
          "epoch": 87,
          "updates": 87,
          "samples_seen": 89088,
          "online_loss": 0.05259322747588158,
          "train_loss": 0.05219176411628723,
          "validation_loss": 0.04225369915366173,
          "train_metric": 0.005256563425064087,
          "validation_metric": 0.004255637992173433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6722872580012336
        },
        {
          "epoch": 88,
          "updates": 88,
          "samples_seen": 90112,
          "online_loss": 0.05219176411628723,
          "train_loss": 0.05180253088474274,
          "validation_loss": 0.04200197011232376,
          "train_metric": 0.00521736079826951,
          "validation_metric": 0.004230284132063389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.678020812001705
        },
        {
          "epoch": 89,
          "updates": 89,
          "samples_seen": 91136,
          "online_loss": 0.05180253088474274,
          "train_loss": 0.05141232907772064,
          "validation_loss": 0.04175002872943878,
          "train_metric": 0.005178062245249748,
          "validation_metric": 0.004204909782856703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6836534240010224
        },
        {
          "epoch": 90,
          "updates": 90,
          "samples_seen": 92160,
          "online_loss": 0.05141233652830124,
          "train_loss": 0.05101370811462402,
          "validation_loss": 0.0414862260222435,
          "train_metric": 0.00513791386038065,
          "validation_metric": 0.004178340546786785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6892802160000429
        },
        {
          "epoch": 91,
          "updates": 91,
          "samples_seen": 93184,
          "online_loss": 0.05101371183991432,
          "train_loss": 0.05060913413763046,
          "validation_loss": 0.04121150076389313,
          "train_metric": 0.005097166169434786,
          "validation_metric": 0.004150671418756247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6951699770015694
        },
        {
          "epoch": 92,
          "updates": 92,
          "samples_seen": 94208,
          "online_loss": 0.05060913786292076,
          "train_loss": 0.05020901560783386,
          "validation_loss": 0.04093712940812111,
          "train_metric": 0.005056868307292461,
          "validation_metric": 0.004123037680983543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.700993197000571
        },
        {
          "epoch": 93,
          "updates": 93,
          "samples_seen": 95232,
          "online_loss": 0.05020901560783386,
          "train_loss": 0.04982450604438782,
          "validation_loss": 0.04067692905664444,
          "train_metric": 0.005018142051994801,
          "validation_metric": 0.0040968311950564384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7834495840015734
        },
        {
          "epoch": 94,
          "updates": 94,
          "samples_seen": 96256,
          "online_loss": 0.049824509769678116,
          "train_loss": 0.04946020990610123,
          "validation_loss": 0.04043892025947571,
          "train_metric": 0.004981451202183962,
          "validation_metric": 0.004072860348969698,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7891779270012194
        },
        {
          "epoch": 95,
          "updates": 95,
          "samples_seen": 97280,
          "online_loss": 0.049460213631391525,
          "train_loss": 0.04911188781261444,
          "validation_loss": 0.04022151231765747,
          "train_metric": 0.004946369677782059,
          "validation_metric": 0.004050963558256626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7949709110016556
        },
        {
          "epoch": 96,
          "updates": 96,
          "samples_seen": 98304,
          "online_loss": 0.04911188781261444,
          "train_loss": 0.04877028986811638,
          "validation_loss": 0.04001576825976372,
          "train_metric": 0.004911964759230614,
          "validation_metric": 0.004030242096632719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8008366340000066
        },
        {
          "epoch": 97,
          "updates": 97,
          "samples_seen": 99328,
          "online_loss": 0.04877028986811638,
          "train_loss": 0.0484275296330452,
          "validation_loss": 0.03981102257966995,
          "train_metric": 0.004877443425357342,
          "validation_metric": 0.004009620286524296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8064671570009523
        },
        {
          "epoch": 98,
          "updates": 98,
          "samples_seen": 100352,
          "online_loss": 0.0484275296330452,
          "train_loss": 0.04808063805103302,
          "validation_loss": 0.03959909453988075,
          "train_metric": 0.004842505790293217,
          "validation_metric": 0.003988275304436684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8123142250005913
        },
        {
          "epoch": 99,
          "updates": 99,
          "samples_seen": 101376,
          "online_loss": 0.04808063805103302,
          "train_loss": 0.04773023724555969,
          "validation_loss": 0.03937488794326782,
          "train_metric": 0.004807214252650738,
          "validation_metric": 0.003965694457292557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8180587030001334
        },
        {
          "epoch": 100,
          "updates": 100,
          "samples_seen": 102400,
          "online_loss": 0.047730229794979095,
          "train_loss": 0.047376878559589386,
          "validation_loss": 0.03913471847772598,
          "train_metric": 0.0047716256231069565,
          "validation_metric": 0.003941505216062069,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8235153410005296
        },
        {
          "epoch": 101,
          "updates": 101,
          "samples_seen": 103424,
          "online_loss": 0.047376878559589386,
          "train_loss": 0.04701949283480644,
          "validation_loss": 0.038875531405210495,
          "train_metric": 0.004735630936920643,
          "validation_metric": 0.003915401175618172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8294063660014217
        },
        {
          "epoch": 102,
          "updates": 102,
          "samples_seen": 104448,
          "online_loss": 0.04701949283480644,
          "train_loss": 0.04665706306695938,
          "validation_loss": 0.038596559315919876,
          "train_metric": 0.004699129145592451,
          "validation_metric": 0.0038873038720339537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8351787540013902
        },
        {
          "epoch": 103,
          "updates": 103,
          "samples_seen": 105472,
          "online_loss": 0.04665707051753998,
          "train_loss": 0.04629116132855415,
          "validation_loss": 0.038301218301057816,
          "train_metric": 0.004662275779992342,
          "validation_metric": 0.003857559058815241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.840766078001252
        },
        {
          "epoch": 104,
          "updates": 104,
          "samples_seen": 106496,
          "online_loss": 0.046291157603263855,
          "train_loss": 0.04592559486627579,
          "validation_loss": 0.037996985018253326,
          "train_metric": 0.004625457338988781,
          "validation_metric": 0.0038269171491265297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8468620840012591
        },
        {
          "epoch": 105,
          "updates": 105,
          "samples_seen": 107520,
          "online_loss": 0.04592559114098549,
          "train_loss": 0.045563213527202606,
          "validation_loss": 0.03769233822822571,
          "train_metric": 0.00458896066993475,
          "validation_metric": 0.0037962342612445354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.852494041000682
        },
        {
          "epoch": 106,
          "updates": 106,
          "samples_seen": 108544,
          "online_loss": 0.045563217252492905,
          "train_loss": 0.04520297423005104,
          "validation_loss": 0.03739339858293533,
          "train_metric": 0.004552678670734167,
          "validation_metric": 0.003766125999391079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8579646080015664
        },
        {
          "epoch": 107,
          "updates": 107,
          "samples_seen": 109568,
          "online_loss": 0.04520297795534134,
          "train_loss": 0.04484056681394577,
          "validation_loss": 0.0371025875210762,
          "train_metric": 0.004516178276389837,
          "validation_metric": 0.0037368368357419968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.863708275001045
        },
        {
          "epoch": 108,
          "updates": 108,
          "samples_seen": 110592,
          "online_loss": 0.04484056681394577,
          "train_loss": 0.04447215795516968,
          "validation_loss": 0.036819763481616974,
          "train_metric": 0.004479072522372007,
          "validation_metric": 0.003708351869136095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8693925400002627
        },
        {
          "epoch": 109,
          "updates": 109,
          "samples_seen": 111616,
          "online_loss": 0.04447215795516968,
          "train_loss": 0.04409727081656456,
          "validation_loss": 0.036543626338243484,
          "train_metric": 0.004441315773874521,
          "validation_metric": 0.0036805400159209967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8750773370011302
        },
        {
          "epoch": 110,
          "updates": 110,
          "samples_seen": 112640,
          "online_loss": 0.04409727454185486,
          "train_loss": 0.04371815547347069,
          "validation_loss": 0.0362718440592289,
          "train_metric": 0.004403132479637861,
          "validation_metric": 0.003653166815638542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8809753800014732
        },
        {
          "epoch": 111,
          "updates": 111,
          "samples_seen": 113664,
          "online_loss": 0.04371815174818039,
          "train_loss": 0.043336883187294006,
          "validation_loss": 0.03600064292550087,
          "train_metric": 0.004364732187241316,
          "validation_metric": 0.0036258529871702194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8864945530003752
        },
        {
          "epoch": 112,
          "updates": 112,
          "samples_seen": 114688,
          "online_loss": 0.043336883187294006,
          "train_loss": 0.04295378923416138,
          "validation_loss": 0.03572562709450722,
          "train_metric": 0.004326148424297571,
          "validation_metric": 0.003598154289647937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8918186660011997
        },
        {
          "epoch": 113,
          "updates": 113,
          "samples_seen": 115712,
          "online_loss": 0.042953792959451675,
          "train_loss": 0.04256860166788101,
          "validation_loss": 0.035443976521492004,
          "train_metric": 0.004287353716790676,
          "validation_metric": 0.0035697873681783676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8974663400003919
        },
        {
          "epoch": 114,
          "updates": 114,
          "samples_seen": 116736,
          "online_loss": 0.042568594217300415,
          "train_loss": 0.04218215495347977,
          "validation_loss": 0.03515595942735672,
          "train_metric": 0.004248432349413633,
          "validation_metric": 0.0035407794639468193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.903225210000528
        },
        {
          "epoch": 115,
          "updates": 115,
          "samples_seen": 117760,
          "online_loss": 0.04218215122818947,
          "train_loss": 0.04179615527391434,
          "validation_loss": 0.03486410155892372,
          "train_metric": 0.004209556151181459,
          "validation_metric": 0.003511384828016162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9090269350017479
        },
        {
          "epoch": 116,
          "updates": 116,
          "samples_seen": 118784,
          "online_loss": 0.041796162724494934,
          "train_loss": 0.04141122102737427,
          "validation_loss": 0.03457057476043701,
          "train_metric": 0.004170786589384079,
          "validation_metric": 0.003481821622699499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.914862121000624
        },
        {
          "epoch": 117,
          "updates": 117,
          "samples_seen": 119808,
          "online_loss": 0.04141122102737427,
          "train_loss": 0.04102613404393196,
          "validation_loss": 0.034275852143764496,
          "train_metric": 0.004132001660764217,
          "validation_metric": 0.0034521385096013546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9204912210007024
        },
        {
          "epoch": 118,
          "updates": 118,
          "samples_seen": 120832,
          "online_loss": 0.04102613776922226,
          "train_loss": 0.04063945636153221,
          "validation_loss": 0.0339793860912323,
          "train_metric": 0.004093057010322809,
          "validation_metric": 0.003422279842197895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.926405081001576
        },
        {
          "epoch": 119,
          "updates": 119,
          "samples_seen": 121856,
          "online_loss": 0.04063945636153221,
          "train_loss": 0.04025142639875412,
          "validation_loss": 0.03368090093135834,
          "train_metric": 0.004053976386785507,
          "validation_metric": 0.0033922172151505947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9321581230014999
        },
        {
          "epoch": 120,
          "updates": 120,
          "samples_seen": 122880,
          "online_loss": 0.04025142639875412,
          "train_loss": 0.03986368328332901,
          "validation_loss": 0.03338056802749634,
          "train_metric": 0.004014924168586731,
          "validation_metric": 0.0033619687892496586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9377637590005179
        },
        {
          "epoch": 121,
          "updates": 121,
          "samples_seen": 123904,
          "online_loss": 0.03986368328332901,
          "train_loss": 0.039477504789829254,
          "validation_loss": 0.03307804837822914,
          "train_metric": 0.003976029343903065,
          "validation_metric": 0.0033314996398985386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9432944010004576
        },
        {
          "epoch": 122,
          "updates": 122,
          "samples_seen": 124928,
          "online_loss": 0.039477501064538956,
          "train_loss": 0.039093002676963806,
          "validation_loss": 0.03277259320020676,
          "train_metric": 0.003937304485589266,
          "validation_metric": 0.003300735494121909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9490573070015671
        },
        {
          "epoch": 123,
          "updates": 123,
          "samples_seen": 125952,
          "online_loss": 0.039093006402254105,
          "train_loss": 0.038710277527570724,
          "validation_loss": 0.032464414834976196,
          "train_metric": 0.0038987575098872185,
          "validation_metric": 0.0032696968410164118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9547080510001251
        },
        {
          "epoch": 124,
          "updates": 124,
          "samples_seen": 126976,
          "online_loss": 0.03871028125286102,
          "train_loss": 0.038330137729644775,
          "validation_loss": 0.03215558081865311,
          "train_metric": 0.00386047107167542,
          "validation_metric": 0.003238592529669404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9604084290003811
        },
        {
          "epoch": 125,
          "updates": 125,
          "samples_seen": 128000,
          "online_loss": 0.038330137729644775,
          "train_loss": 0.0379534550011158,
          "validation_loss": 0.031849201768636703,
          "train_metric": 0.0038225334137678146,
          "validation_metric": 0.0032077347859740257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9658797730007791
        },
        {
          "epoch": 126,
          "updates": 126,
          "samples_seen": 129024,
          "online_loss": 0.0379534587264061,
          "train_loss": 0.03758011758327484,
          "validation_loss": 0.03154782950878143,
          "train_metric": 0.0037849314976483583,
          "validation_metric": 0.0031773820519447327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9714838760010025
        },
        {
          "epoch": 127,
          "updates": 127,
          "samples_seen": 130048,
          "online_loss": 0.03758011758327484,
          "train_loss": 0.037209488451480865,
          "validation_loss": 0.03125258535146713,
          "train_metric": 0.0037476036231964827,
          "validation_metric": 0.003147645853459835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9771243280010822
        },
        {
          "epoch": 128,
          "updates": 128,
          "samples_seen": 131072,
          "online_loss": 0.037209488451480865,
          "train_loss": 0.03684167191386223,
          "validation_loss": 0.030963294208049774,
          "train_metric": 0.0037105584051460028,
          "validation_metric": 0.00311850942671299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.982923297000525
        },
        {
          "epoch": 129,
          "updates": 129,
          "samples_seen": 132096,
          "online_loss": 0.03684167563915253,
          "train_loss": 0.036477573215961456,
          "validation_loss": 0.03067859075963497,
          "train_metric": 0.0036738873459398746,
          "validation_metric": 0.003089835401624441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9886392890002753
        },
        {
          "epoch": 130,
          "updates": 130,
          "samples_seen": 133120,
          "online_loss": 0.036477576941251755,
          "train_loss": 0.036117807030677795,
          "validation_loss": 0.030396156013011932,
          "train_metric": 0.003637653309851885,
          "validation_metric": 0.0030613893177360296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.99425251500179
        },
        {
          "epoch": 131,
          "updates": 131,
          "samples_seen": 134144,
          "online_loss": 0.036117807030677795,
          "train_loss": 0.03576233983039856,
          "validation_loss": 0.03011367656290531,
          "train_metric": 0.0036018523387610912,
          "validation_metric": 0.003032939275726676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0781708270005765
        },
        {
          "epoch": 132,
          "updates": 132,
          "samples_seen": 135168,
          "online_loss": 0.03576234355568886,
          "train_loss": 0.03541123867034912,
          "validation_loss": 0.029830317944288254,
          "train_metric": 0.003566490253433585,
          "validation_metric": 0.0030044002924114466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0903747480006132
        },
        {
          "epoch": 133,
          "updates": 133,
          "samples_seen": 136192,
          "online_loss": 0.03541124239563942,
          "train_loss": 0.03506504371762276,
          "validation_loss": 0.029547058045864105,
          "train_metric": 0.0035316227003932,
          "validation_metric": 0.0029758713208138943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1029643770016264
        },
        {
          "epoch": 134,
          "updates": 134,
          "samples_seen": 137216,
          "online_loss": 0.035065047442913055,
          "train_loss": 0.03472407907247543,
          "validation_loss": 0.029265543445944786,
          "train_metric": 0.0034972820430994034,
          "validation_metric": 0.0029475183691829443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1088058550012647
        },
        {
          "epoch": 135,
          "updates": 135,
          "samples_seen": 138240,
          "online_loss": 0.03472407907247543,
          "train_loss": 0.034388091415166855,
          "validation_loss": 0.028986889868974686,
          "train_metric": 0.0034634429030120373,
          "validation_metric": 0.002919453661888838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.11471698600144
        },
        {
          "epoch": 136,
          "updates": 136,
          "samples_seen": 139264,
          "online_loss": 0.034388087689876556,
          "train_loss": 0.03405686095356941,
          "validation_loss": 0.028711378574371338,
          "train_metric": 0.0034300819970667362,
          "validation_metric": 0.002891704672947526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1204920500003936
        },
        {
          "epoch": 137,
          "updates": 137,
          "samples_seen": 140288,
          "online_loss": 0.034056857228279114,
          "train_loss": 0.033730652183294296,
          "validation_loss": 0.028438625857234,
          "train_metric": 0.0033972279634326696,
          "validation_metric": 0.0028642339166253805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1263357160005398
        },
        {
          "epoch": 138,
          "updates": 138,
          "samples_seen": 141312,
          "online_loss": 0.033730655908584595,
          "train_loss": 0.03340977430343628,
          "validation_loss": 0.028167687356472015,
          "train_metric": 0.0033649103716015816,
          "validation_metric": 0.0028369459323585033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1320319140013453
        },
        {
          "epoch": 139,
          "updates": 139,
          "samples_seen": 142336,
          "online_loss": 0.03340977802872658,
          "train_loss": 0.03309409320354462,
          "validation_loss": 0.027897637337446213,
          "train_metric": 0.0033331154845654964,
          "validation_metric": 0.0028097478207200766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1379980770016118
        },
        {
          "epoch": 140,
          "updates": 140,
          "samples_seen": 143360,
          "online_loss": 0.03309409320354462,
          "train_loss": 0.03278348967432976,
          "validation_loss": 0.027628464624285698,
          "train_metric": 0.00330183282494545,
          "validation_metric": 0.0027826379518955946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1436255450007593
        },
        {
          "epoch": 141,
          "updates": 141,
          "samples_seen": 144384,
          "online_loss": 0.03278348594903946,
          "train_loss": 0.03247813135385513,
          "validation_loss": 0.027361474931240082,
          "train_metric": 0.00327107822522521,
          "validation_metric": 0.0027557476423680782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.149199633000535
        },
        {
          "epoch": 142,
          "updates": 142,
          "samples_seen": 145408,
          "online_loss": 0.03247813135385513,
          "train_loss": 0.03217807412147522,
          "validation_loss": 0.02709847502410412,
          "train_metric": 0.003240857506170869,
          "validation_metric": 0.002729259431362152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1551097670017043
        },
        {
          "epoch": 143,
          "updates": 143,
          "samples_seen": 146432,
          "online_loss": 0.03217807412147522,
          "train_loss": 0.03188301622867584,
          "validation_loss": 0.02684067189693451,
          "train_metric": 0.003211140399798751,
          "validation_metric": 0.002703294390812516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.161009538000144
        },
        {
          "epoch": 144,
          "updates": 144,
          "samples_seen": 147456,
          "online_loss": 0.031883012503385544,
          "train_loss": 0.03159274160861969,
          "validation_loss": 0.026588251814246178,
          "train_metric": 0.0031819052528589964,
          "validation_metric": 0.0026778713800013065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1668387120007537
        },
        {
          "epoch": 145,
          "updates": 145,
          "samples_seen": 148480,
          "online_loss": 0.03159274160861969,
          "train_loss": 0.03130728751420975,
          "validation_loss": 0.026340413838624954,
          "train_metric": 0.0031531555578112602,
          "validation_metric": 0.002652909839525819,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1727971480013366
        },
        {
          "epoch": 146,
          "updates": 146,
          "samples_seen": 149504,
          "online_loss": 0.031307291239500046,
          "train_loss": 0.031026573851704597,
          "validation_loss": 0.026095816865563393,
          "train_metric": 0.003124882932752371,
          "validation_metric": 0.002628275426104665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1787775620014145
        },
        {
          "epoch": 147,
          "updates": 147,
          "samples_seen": 150528,
          "online_loss": 0.031026573851704597,
          "train_loss": 0.030750317499041557,
          "validation_loss": 0.025853363797068596,
          "train_metric": 0.003097059205174446,
          "validation_metric": 0.0026038559153676033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.184404996000012
        },
        {
          "epoch": 148,
          "updates": 148,
          "samples_seen": 151552,
          "online_loss": 0.030750319361686707,
          "train_loss": 0.03047836199402809,
          "validation_loss": 0.02561296336352825,
          "train_metric": 0.003069669473916292,
          "validation_metric": 0.0025796438567340374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1901246950001223
        },
        {
          "epoch": 149,
          "updates": 149,
          "samples_seen": 152576,
          "online_loss": 0.03047836199402809,
          "train_loss": 0.03021063841879368,
          "validation_loss": 0.025375360623002052,
          "train_metric": 0.0030427046585828066,
          "validation_metric": 0.0025557130575180054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.195896047000133
        },
        {
          "epoch": 150,
          "updates": 150,
          "samples_seen": 153600,
          "online_loss": 0.030210640281438828,
          "train_loss": 0.029946928843855858,
          "validation_loss": 0.02514127641916275,
          "train_metric": 0.0030161449685692787,
          "validation_metric": 0.0025321373250335455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2016081350011518
        },
        {
          "epoch": 151,
          "updates": 151,
          "samples_seen": 154624,
          "online_loss": 0.029946930706501007,
          "train_loss": 0.02968693897128105,
          "validation_loss": 0.02491086721420288,
          "train_metric": 0.00298995990306139,
          "validation_metric": 0.0025089315604418516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.207321124000373
        },
        {
          "epoch": 152,
          "updates": 152,
          "samples_seen": 155648,
          "online_loss": 0.029686937108635902,
          "train_loss": 0.029430503025650978,
          "validation_loss": 0.024683713912963867,
          "train_metric": 0.002964132698252797,
          "validation_metric": 0.0024860536213964224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2129700110017438
        },
        {
          "epoch": 153,
          "updates": 153,
          "samples_seen": 156672,
          "online_loss": 0.029430504888296127,
          "train_loss": 0.029177503660321236,
          "validation_loss": 0.02445906586945057,
          "train_metric": 0.002938651479780674,
          "validation_metric": 0.002463427372276783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2188021450001543
        },
        {
          "epoch": 154,
          "updates": 154,
          "samples_seen": 157696,
          "online_loss": 0.029177501797676086,
          "train_loss": 0.028927674517035484,
          "validation_loss": 0.024236397817730904,
          "train_metric": 0.0029134894721210003,
          "validation_metric": 0.0024410011246800423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2245869700000185
        },
        {
          "epoch": 155,
          "updates": 155,
          "samples_seen": 158720,
          "online_loss": 0.028927674517035484,
          "train_loss": 0.028680797666311264,
          "validation_loss": 0.024015776813030243,
          "train_metric": 0.0028886250220239162,
          "validation_metric": 0.0024187813978642225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2305132870005764
        },
        {
          "epoch": 156,
          "updates": 156,
          "samples_seen": 159744,
          "online_loss": 0.028680801391601562,
          "train_loss": 0.02843674272298813,
          "validation_loss": 0.02379787340760231,
          "train_metric": 0.0028640448581427336,
          "validation_metric": 0.0023968347813934088,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2363497580008698
        },
        {
          "epoch": 157,
          "updates": 157,
          "samples_seen": 160768,
          "online_loss": 0.028436744585633278,
          "train_loss": 0.028195317834615707,
          "validation_loss": 0.023583384230732918,
          "train_metric": 0.0028397291898727417,
          "validation_metric": 0.002375232055783272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2422537290003675
        },
        {
          "epoch": 158,
          "updates": 158,
          "samples_seen": 161792,
          "online_loss": 0.028195319697260857,
          "train_loss": 0.027956288307905197,
          "validation_loss": 0.023372555151581764,
          "train_metric": 0.002815654966980219,
          "validation_metric": 0.0023539981339126825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.247894151001674
        },
        {
          "epoch": 159,
          "updates": 159,
          "samples_seen": 162816,
          "online_loss": 0.027956292033195496,
          "train_loss": 0.027719510719180107,
          "validation_loss": 0.02316519059240818,
          "train_metric": 0.0027918072883039713,
          "validation_metric": 0.0023331132251769304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2535437130009086
        },
        {
          "epoch": 160,
          "updates": 160,
          "samples_seen": 163840,
          "online_loss": 0.027719508856534958,
          "train_loss": 0.02748485654592514,
          "validation_loss": 0.022960782051086426,
          "train_metric": 0.002768174046650529,
          "validation_metric": 0.002312525874003768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2592593130011664
        },
        {
          "epoch": 161,
          "updates": 161,
          "samples_seen": 164864,
          "online_loss": 0.02748485654592514,
          "train_loss": 0.027252160012722015,
          "validation_loss": 0.022758863866329193,
          "train_metric": 0.0027447377797216177,
          "validation_metric": 0.0022921895142644644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2647636430010607
        },
        {
          "epoch": 162,
          "updates": 162,
          "samples_seen": 165888,
          "online_loss": 0.027252160012722015,
          "train_loss": 0.02702128142118454,
          "validation_loss": 0.02255934290587902,
          "train_metric": 0.0027214845176786184,
          "validation_metric": 0.00227209459990263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2704353820008691
        },
        {
          "epoch": 163,
          "updates": 163,
          "samples_seen": 166912,
          "online_loss": 0.02702128142118454,
          "train_loss": 0.026792125776410103,
          "validation_loss": 0.022362379357218742,
          "train_metric": 0.0026984047144651413,
          "validation_metric": 0.002252256963402033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.276371250000011
        },
        {
          "epoch": 164,
          "updates": 164,
          "samples_seen": 167936,
          "online_loss": 0.026792127639055252,
          "train_loss": 0.026564572006464005,
          "validation_loss": 0.022168081253767014,
          "train_metric": 0.0026754867285490036,
          "validation_metric": 0.0022326880134642124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.282103344999996
        },
        {
          "epoch": 165,
          "updates": 165,
          "samples_seen": 168960,
          "online_loss": 0.026564572006464005,
          "train_loss": 0.02633850835263729,
          "validation_loss": 0.021976284682750702,
          "train_metric": 0.0026527182199060917,
          "validation_metric": 0.0022133709862828255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2877433280009427
        },
        {
          "epoch": 166,
          "updates": 166,
          "samples_seen": 169984,
          "online_loss": 0.02633850835263729,
          "train_loss": 0.026113850995898247,
          "validation_loss": 0.021786633878946304,
          "train_metric": 0.0026300917379558086,
          "validation_metric": 0.0021942697931081057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2935059320006985
        },
        {
          "epoch": 167,
          "updates": 167,
          "samples_seen": 171008,
          "online_loss": 0.026113852858543396,
          "train_loss": 0.025890545919537544,
          "validation_loss": 0.021598778665065765,
          "train_metric": 0.0026076012291014194,
          "validation_metric": 0.002175349975004792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.299249661000431
        },
        {
          "epoch": 168,
          "updates": 168,
          "samples_seen": 172032,
          "online_loss": 0.025890545919537544,
          "train_loss": 0.02566850185394287,
          "validation_loss": 0.021412523463368416,
          "train_metric": 0.0025852371472865343,
          "validation_metric": 0.0021565908100456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3857621120005206
        },
        {
          "epoch": 169,
          "updates": 169,
          "samples_seen": 173056,
          "online_loss": 0.025668499991297722,
          "train_loss": 0.02544766664505005,
          "validation_loss": 0.021227912977337837,
          "train_metric": 0.0025629953015595675,
          "validation_metric": 0.0021379978861659765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3913547870015464
        },
        {
          "epoch": 170,
          "updates": 170,
          "samples_seen": 174080,
          "online_loss": 0.02544766291975975,
          "train_loss": 0.025228001177310944,
          "validation_loss": 0.02104506827890873,
          "train_metric": 0.0025408719666302204,
          "validation_metric": 0.0021195823792368174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.39716984900042
        },
        {
          "epoch": 171,
          "updates": 171,
          "samples_seen": 175104,
          "online_loss": 0.025228003039956093,
          "train_loss": 0.02500946819782257,
          "validation_loss": 0.020864052698016167,
          "train_metric": 0.002518862020224333,
          "validation_metric": 0.0021013508085161448,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4027061620017776
        },
        {
          "epoch": 172,
          "updates": 172,
          "samples_seen": 176128,
          "online_loss": 0.02500947006046772,
          "train_loss": 0.02479201927781105,
          "validation_loss": 0.020684804767370224,
          "train_metric": 0.002496961737051606,
          "validation_metric": 0.0020832980517297983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.408122146000096
        },
        {
          "epoch": 173,
          "updates": 173,
          "samples_seen": 177152,
          "online_loss": 0.02479202300310135,
          "train_loss": 0.024575665593147278,
          "validation_loss": 0.02050723321735859,
          "train_metric": 0.0024751711171120405,
          "validation_metric": 0.0020654136314988136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4138971410011436
        },
        {
          "epoch": 174,
          "updates": 174,
          "samples_seen": 178176,
          "online_loss": 0.024575665593147278,
          "train_loss": 0.024360371753573418,
          "validation_loss": 0.020331302657723427,
          "train_metric": 0.002453487366437912,
          "validation_metric": 0.0020476942881941795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4196881760017277
        },
        {
          "epoch": 175,
          "updates": 175,
          "samples_seen": 179200,
          "online_loss": 0.024360371753573418,
          "train_loss": 0.02414610981941223,
          "validation_loss": 0.020156988874077797,
          "train_metric": 0.0024319079238921404,
          "validation_metric": 0.002030138159170747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4251389290002408
        },
        {
          "epoch": 176,
          "updates": 176,
          "samples_seen": 180224,
          "online_loss": 0.02414610981941223,
          "train_loss": 0.023932896554470062,
          "validation_loss": 0.019984275102615356,
          "train_metric": 0.0024104334879666567,
          "validation_metric": 0.002012742916122079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4308941520012013
        },
        {
          "epoch": 177,
          "updates": 177,
          "samples_seen": 181248,
          "online_loss": 0.023932896554470062,
          "train_loss": 0.023720718920230865,
          "validation_loss": 0.019813038408756256,
          "train_metric": 0.0023890638258308172,
          "validation_metric": 0.001995496451854706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4365337470007944
        },
        {
          "epoch": 178,
          "updates": 178,
          "samples_seen": 182272,
          "online_loss": 0.023720718920230865,
          "train_loss": 0.023509567603468895,
          "validation_loss": 0.019643103703856468,
          "train_metric": 0.0023677973076701164,
          "validation_metric": 0.001978381536900997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4422305440002674
        },
        {
          "epoch": 179,
          "updates": 179,
          "samples_seen": 183296,
          "online_loss": 0.023509565740823746,
          "train_loss": 0.023299453780055046,
          "validation_loss": 0.01947432942688465,
          "train_metric": 0.0023466357961297035,
          "validation_metric": 0.0019613830372691154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4480705490004766
        },
        {
          "epoch": 180,
          "updates": 180,
          "samples_seen": 184320,
          "online_loss": 0.023299451917409897,
          "train_loss": 0.023090394213795662,
          "validation_loss": 0.019306605681777,
          "train_metric": 0.0023255799897015095,
          "validation_metric": 0.00194449070841074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4539013500016154
        },
        {
          "epoch": 181,
          "updates": 181,
          "samples_seen": 185344,
          "online_loss": 0.023090394213795662,
          "train_loss": 0.022882379591464996,
          "validation_loss": 0.019139841198921204,
          "train_metric": 0.0023046296555548906,
          "validation_metric": 0.0019276948878541589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4596386790017277
        },
        {
          "epoch": 182,
          "updates": 182,
          "samples_seen": 186368,
          "online_loss": 0.022882383316755295,
          "train_loss": 0.022675421088933945,
          "validation_loss": 0.0189739391207695,
          "train_metric": 0.002283785492181778,
          "validation_metric": 0.001910985796712339,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4653859350000857
        },
        {
          "epoch": 183,
          "updates": 183,
          "samples_seen": 187392,
          "online_loss": 0.022675422951579094,
          "train_loss": 0.02246953547000885,
          "validation_loss": 0.018808746710419655,
          "train_metric": 0.0022630493622273207,
          "validation_metric": 0.0018943483009934425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4710688920004031
        },
        {
          "epoch": 184,
          "updates": 184,
          "samples_seen": 188416,
          "online_loss": 0.02246953547000885,
          "train_loss": 0.02226472645998001,
          "validation_loss": 0.01864410936832428,
          "train_metric": 0.002242421731352806,
          "validation_metric": 0.0018777662189677358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.476842279000266
        },
        {
          "epoch": 185,
          "updates": 185,
          "samples_seen": 189440,
          "online_loss": 0.022264724597334862,
          "train_loss": 0.022061005234718323,
          "validation_loss": 0.018479984253644943,
          "train_metric": 0.0022219037637114525,
          "validation_metric": 0.0018612362910062075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4827081729999918
        },
        {
          "epoch": 186,
          "updates": 186,
          "samples_seen": 190464,
          "online_loss": 0.022061007097363472,
          "train_loss": 0.021858401596546173,
          "validation_loss": 0.01831640861928463,
          "train_metric": 0.0022014982532709837,
          "validation_metric": 0.0018447617767378688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.488620191001246
        },
        {
          "epoch": 187,
          "updates": 187,
          "samples_seen": 191488,
          "online_loss": 0.021858401596546173,
          "train_loss": 0.02165691927075386,
          "validation_loss": 0.018153509125113487,
          "train_metric": 0.0021812058985233307,
          "validation_metric": 0.0018283551326021552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4943601230006607
        },
        {
          "epoch": 188,
          "updates": 188,
          "samples_seen": 192512,
          "online_loss": 0.02165692113339901,
          "train_loss": 0.021456578746438026,
          "validation_loss": 0.01799130067229271,
          "train_metric": 0.0021610280964523554,
          "validation_metric": 0.0018120178719982505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.500131123000756
        },
        {
          "epoch": 189,
          "updates": 189,
          "samples_seen": 193536,
          "online_loss": 0.021456578746438026,
          "train_loss": 0.02125740237534046,
          "validation_loss": 0.017829734832048416,
          "train_metric": 0.002140967408195138,
          "validation_metric": 0.0017957454547286034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5058465440015425
        },
        {
          "epoch": 190,
          "updates": 190,
          "samples_seen": 194560,
          "online_loss": 0.021257400512695312,
          "train_loss": 0.021059410646557808,
          "validation_loss": 0.017668716609477997,
          "train_metric": 0.0021210266277194023,
          "validation_metric": 0.0017795284511521459,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5115522510004666
        },
        {
          "epoch": 191,
          "updates": 191,
          "samples_seen": 195584,
          "online_loss": 0.021059412509202957,
          "train_loss": 0.020862611010670662,
          "validation_loss": 0.017508231103420258,
          "train_metric": 0.0021012055221945047,
          "validation_metric": 0.001763364765793085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5173670610001864
        },
        {
          "epoch": 192,
          "updates": 192,
          "samples_seen": 196608,
          "online_loss": 0.020862609148025513,
          "train_loss": 0.020667031407356262,
          "validation_loss": 0.01734834909439087,
          "train_metric": 0.0020815078169107437,
          "validation_metric": 0.0017472621984779835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5229280880012084
        },
        {
          "epoch": 193,
          "updates": 193,
          "samples_seen": 197632,
          "online_loss": 0.02066703327000141,
          "train_loss": 0.020472707226872444,
          "validation_loss": 0.01718917302787304,
          "train_metric": 0.0020619358401745558,
          "validation_metric": 0.0017312304116785526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.528657305001616
        },
        {
          "epoch": 194,
          "updates": 194,
          "samples_seen": 198656,
          "online_loss": 0.020472705364227295,
          "train_loss": 0.020279642194509506,
          "validation_loss": 0.01703072339296341,
          "train_metric": 0.002042490988969803,
          "validation_metric": 0.0017152719665318727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5343373460000294
        },
        {
          "epoch": 195,
          "updates": 195,
          "samples_seen": 199680,
          "online_loss": 0.020279640331864357,
          "train_loss": 0.02008787728846073,
          "validation_loss": 0.016872890293598175,
          "train_metric": 0.002023177221417427,
          "validation_metric": 0.0016993754543364048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5399255540014565
        },
        {
          "epoch": 196,
          "updates": 196,
          "samples_seen": 200704,
          "online_loss": 0.020087875425815582,
          "train_loss": 0.019897427409887314,
          "validation_loss": 0.016715584322810173,
          "train_metric": 0.0020039959345012903,
          "validation_metric": 0.0016835323767736554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5458170150013757
        },
        {
          "epoch": 197,
          "updates": 197,
          "samples_seen": 201728,
          "online_loss": 0.019897425547242165,
          "train_loss": 0.01970832422375679,
          "validation_loss": 0.01655876822769642,
          "train_metric": 0.001984950387850404,
          "validation_metric": 0.001667738426476717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5515801920009835
        },
        {
          "epoch": 198,
          "updates": 198,
          "samples_seen": 202752,
          "online_loss": 0.01970832608640194,
          "train_loss": 0.0195205919444561,
          "validation_loss": 0.01640251837670803,
          "train_metric": 0.0019660424441099167,
          "validation_metric": 0.001652001403272152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5572676710016822
        },
        {
          "epoch": 199,
          "updates": 199,
          "samples_seen": 203776,
          "online_loss": 0.01952059008181095,
          "train_loss": 0.019334254786372185,
          "validation_loss": 0.016246991232037544,
          "train_metric": 0.0019472752464935184,
          "validation_metric": 0.0016363373724743724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5630301160017552
        },
        {
          "epoch": 200,
          "updates": 200,
          "samples_seen": 204800,
          "online_loss": 0.019334256649017334,
          "train_loss": 0.019149335101246834,
          "validation_loss": 0.01609225571155548,
          "train_metric": 0.0019286508904770017,
          "validation_metric": 0.001620753202587366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5687842050010659
        },
        {
          "epoch": 201,
          "updates": 201,
          "samples_seen": 205824,
          "online_loss": 0.019149336963891983,
          "train_loss": 0.01896587200462818,
          "validation_loss": 0.015938322991132736,
          "train_metric": 0.0019101729849353433,
          "validation_metric": 0.0016052494756877422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5745388900013495
        },
        {
          "epoch": 202,
          "updates": 202,
          "samples_seen": 206848,
          "online_loss": 0.018965870141983032,
          "train_loss": 0.018783878535032272,
          "validation_loss": 0.015785153955221176,
          "train_metric": 0.0018918435089290142,
          "validation_metric": 0.0015898229321464896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.580287444001442
        },
        {
          "epoch": 203,
          "updates": 203,
          "samples_seen": 207872,
          "online_loss": 0.018783878535032272,
          "train_loss": 0.018603375181555748,
          "validation_loss": 0.015632793307304382,
          "train_metric": 0.0018736638594418764,
          "validation_metric": 0.0015744775300845504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5859639310001512
        },
        {
          "epoch": 204,
          "updates": 204,
          "samples_seen": 208896,
          "online_loss": 0.018603377044200897,
          "train_loss": 0.018424399197101593,
          "validation_loss": 0.015481379814445972,
          "train_metric": 0.001855637994594872,
          "validation_metric": 0.001559227705001831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5917531330014754
        },
        {
          "epoch": 205,
          "updates": 205,
          "samples_seen": 209920,
          "online_loss": 0.018424399197101593,
          "train_loss": 0.018246965482831,
          "validation_loss": 0.015331093221902847,
          "train_metric": 0.0018377677770331502,
          "validation_metric": 0.0015440917341038585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5974697880010353
        },
        {
          "epoch": 206,
          "updates": 206,
          "samples_seen": 210944,
          "online_loss": 0.01824696734547615,
          "train_loss": 0.018071098253130913,
          "validation_loss": 0.015181987546384335,
          "train_metric": 0.0018200549529865384,
          "validation_metric": 0.0015290741575881839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6032973500005028
        },
        {
          "epoch": 207,
          "updates": 207,
          "samples_seen": 211968,
          "online_loss": 0.018071098253130913,
          "train_loss": 0.017896821722388268,
          "validation_loss": 0.015034045092761517,
          "train_metric": 0.0018025024328380823,
          "validation_metric": 0.0015141740441322327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6089033290008956
        },
        {
          "epoch": 208,
          "updates": 208,
          "samples_seen": 212992,
          "online_loss": 0.017896821722388268,
          "train_loss": 0.01772415079176426,
          "validation_loss": 0.014887221157550812,
          "train_metric": 0.0017851116135716438,
          "validation_metric": 0.0014993863878771663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.700951310000164
        },
        {
          "epoch": 209,
          "updates": 209,
          "samples_seen": 214016,
          "online_loss": 0.01772415079176426,
          "train_loss": 0.01755310781300068,
          "validation_loss": 0.014741534367203712,
          "train_metric": 0.0017678847070783377,
          "validation_metric": 0.0014847132842987776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7069755590000568
        },
        {
          "epoch": 210,
          "updates": 210,
          "samples_seen": 215040,
          "online_loss": 0.017553109675645828,
          "train_loss": 0.017383717000484467,
          "validation_loss": 0.014597069472074509,
          "train_metric": 0.0017508243909105659,
          "validation_metric": 0.0014701634645462036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.712673950001772
        },
        {
          "epoch": 211,
          "updates": 211,
          "samples_seen": 216064,
          "online_loss": 0.017383717000484467,
          "train_loss": 0.017215993255376816,
          "validation_loss": 0.01445392332971096,
          "train_metric": 0.0017339318292215466,
          "validation_metric": 0.0014557462418451905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7184469320000062
        },
        {
          "epoch": 212,
          "updates": 212,
          "samples_seen": 217088,
          "online_loss": 0.017215993255376816,
          "train_loss": 0.017049945890903473,
          "validation_loss": 0.014312129467725754,
          "train_metric": 0.0017172080697491765,
          "validation_metric": 0.0014414653414860368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.724031460000333
        },
        {
          "epoch": 213,
          "updates": 213,
          "samples_seen": 218112,
          "online_loss": 0.017049945890903473,
          "train_loss": 0.016885602846741676,
          "validation_loss": 0.014171640388667583,
          "train_metric": 0.001700656022876501,
          "validation_metric": 0.001427315641194582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7301830090000294
        },
        {
          "epoch": 214,
          "updates": 214,
          "samples_seen": 219136,
          "online_loss": 0.016885604709386826,
          "train_loss": 0.01672298088669777,
          "validation_loss": 0.014032428152859211,
          "train_metric": 0.0016842774348333478,
          "validation_metric": 0.0014132946962490678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7361247240005468
        },
        {
          "epoch": 215,
          "updates": 215,
          "samples_seen": 220160,
          "online_loss": 0.01672298088669777,
          "train_loss": 0.01656208373606205,
          "validation_loss": 0.013894537463784218,
          "train_metric": 0.0016680725384503603,
          "validation_metric": 0.0013994069304317236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7418143220002094
        },
        {
          "epoch": 216,
          "updates": 216,
          "samples_seen": 221184,
          "online_loss": 0.01656208373606205,
          "train_loss": 0.01640293374657631,
          "validation_loss": 0.013758068904280663,
          "train_metric": 0.0016520433127880096,
          "validation_metric": 0.0013856622390449047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.747621638000055
        },
        {
          "epoch": 217,
          "updates": 217,
          "samples_seen": 222208,
          "online_loss": 0.01640293188393116,
          "train_loss": 0.016245543956756592,
          "validation_loss": 0.013623112812638283,
          "train_metric": 0.0016361918533220887,
          "validation_metric": 0.001372070051729679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7535126320017298
        },
        {
          "epoch": 218,
          "updates": 218,
          "samples_seen": 223232,
          "online_loss": 0.01624554581940174,
          "train_loss": 0.016089918091893196,
          "validation_loss": 0.013489669188857079,
          "train_metric": 0.001620517810806632,
          "validation_metric": 0.0013586300192400813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7593313400011539
        },
        {
          "epoch": 219,
          "updates": 219,
          "samples_seen": 224256,
          "online_loss": 0.016089918091893196,
          "train_loss": 0.01593608781695366,
          "validation_loss": 0.013357715681195259,
          "train_metric": 0.0016050240956246853,
          "validation_metric": 0.001345340278930962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7650198380015354
        },
        {
          "epoch": 220,
          "updates": 220,
          "samples_seen": 225280,
          "online_loss": 0.01593608595430851,
          "train_loss": 0.015784041956067085,
          "validation_loss": 0.013227248564362526,
          "train_metric": 0.0015897109406068921,
          "validation_metric": 0.0013322001323103905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7707740270016075
        },
        {
          "epoch": 221,
          "updates": 221,
          "samples_seen": 226304,
          "online_loss": 0.015784041956067085,
          "train_loss": 0.015633797273039818,
          "validation_loss": 0.01309829019010067,
          "train_metric": 0.0015745788114145398,
          "validation_metric": 0.0013192116748541594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7766738530008297
        },
        {
          "epoch": 222,
          "updates": 222,
          "samples_seen": 227328,
          "online_loss": 0.015633797273039818,
          "train_loss": 0.015485364943742752,
          "validation_loss": 0.012970906682312489,
          "train_metric": 0.0015596292214468122,
          "validation_metric": 0.0013063822407275438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7823339220012713
        },
        {
          "epoch": 223,
          "updates": 223,
          "samples_seen": 228352,
          "online_loss": 0.015485365875065327,
          "train_loss": 0.015338747762143612,
          "validation_loss": 0.01284513995051384,
          "train_metric": 0.0015448624035343528,
          "validation_metric": 0.0012937153223901987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7877839020002284
        },
        {
          "epoch": 224,
          "updates": 224,
          "samples_seen": 229376,
          "online_loss": 0.015338745899498463,
          "train_loss": 0.01519395224750042,
          "validation_loss": 0.01272093877196312,
          "train_metric": 0.0015302790561690927,
          "validation_metric": 0.001281206263229251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.793330998001693
        },
        {
          "epoch": 225,
          "updates": 225,
          "samples_seen": 230400,
          "online_loss": 0.01519395224750042,
          "train_loss": 0.015050990507006645,
          "validation_loss": 0.012598264031112194,
          "train_metric": 0.0015158806927502155,
          "validation_metric": 0.0012688509887084365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7991624680016685
        },
        {
          "epoch": 226,
          "updates": 226,
          "samples_seen": 231424,
          "online_loss": 0.01505099143832922,
          "train_loss": 0.014909853227436543,
          "validation_loss": 0.012477121315896511,
          "train_metric": 0.0015016660327091813,
          "validation_metric": 0.0012566500809043646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.804939726000157
        },
        {
          "epoch": 227,
          "updates": 227,
          "samples_seen": 232448,
          "online_loss": 0.014909852296113968,
          "train_loss": 0.014770550653338432,
          "validation_loss": 0.012357545085251331,
          "train_metric": 0.0014876356581225991,
          "validation_metric": 0.0012446066830307245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8107630080012314
        },
        {
          "epoch": 228,
          "updates": 228,
          "samples_seen": 233472,
          "online_loss": 0.014770549722015858,
          "train_loss": 0.014633073471486568,
          "validation_loss": 0.012239589355885983,
          "train_metric": 0.0014737894525751472,
          "validation_metric": 0.0012327264994382858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8165550120011176
        },
        {
          "epoch": 229,
          "updates": 229,
          "samples_seen": 234496,
          "online_loss": 0.014633072540163994,
          "train_loss": 0.01449743565171957,
          "validation_loss": 0.01212325133383274,
          "train_metric": 0.0014601288130506873,
          "validation_metric": 0.0012210095301270485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8222116960005224
        },
        {
          "epoch": 230,
          "updates": 230,
          "samples_seen": 235520,
          "online_loss": 0.014497434720396996,
          "train_loss": 0.014363635331392288,
          "validation_loss": 0.012008504010736942,
          "train_metric": 0.0014466525753960013,
          "validation_metric": 0.0012094525154680014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.827947388001121
        },
        {
          "epoch": 231,
          "updates": 231,
          "samples_seen": 236544,
          "online_loss": 0.014363635331392288,
          "train_loss": 0.014231644570827484,
          "validation_loss": 0.011895321309566498,
          "train_metric": 0.0014333591097965837,
          "validation_metric": 0.0011980532435700297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8337110480006231
        },
        {
          "epoch": 232,
          "updates": 232,
          "samples_seen": 237568,
          "online_loss": 0.014231645502150059,
          "train_loss": 0.014101490378379822,
          "validation_loss": 0.011783742345869541,
          "train_metric": 0.0014202503953129053,
          "validation_metric": 0.0011868153233081102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8393843220001145
        },
        {
          "epoch": 233,
          "updates": 233,
          "samples_seen": 238592,
          "online_loss": 0.014101488515734673,
          "train_loss": 0.01397313829511404,
          "validation_loss": 0.011673802509903908,
          "train_metric": 0.0014073234051465988,
          "validation_metric": 0.0011757425963878632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8452292920010223
        },
        {
          "epoch": 234,
          "updates": 234,
          "samples_seen": 239616,
          "online_loss": 0.01397313829511404,
          "train_loss": 0.013846595771610737,
          "validation_loss": 0.011565502732992172,
          "train_metric": 0.0013945784885436296,
          "validation_metric": 0.0011648350628092885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8509199250001984
        },
        {
          "epoch": 235,
          "updates": 235,
          "samples_seen": 240640,
          "online_loss": 0.013846594840288162,
          "train_loss": 0.013721850700676441,
          "validation_loss": 0.01145881973206997,
          "train_metric": 0.0013820144813507795,
          "validation_metric": 0.001154090161435306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8565330500005075
        },
        {
          "epoch": 236,
          "updates": 236,
          "samples_seen": 241664,
          "online_loss": 0.013721850700676441,
          "train_loss": 0.013598894700407982,
          "validation_loss": 0.01135371346026659,
          "train_metric": 0.0013696306850761175,
          "validation_metric": 0.0011435045162215829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8623020940012793
        },
        {
          "epoch": 237,
          "updates": 237,
          "samples_seen": 242688,
          "online_loss": 0.013598893769085407,
          "train_loss": 0.013477708213031292,
          "validation_loss": 0.011250190436840057,
          "train_metric": 0.001357425469905138,
          "validation_metric": 0.0011330778943374753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8682994580012746
        },
        {
          "epoch": 238,
          "updates": 238,
          "samples_seen": 243712,
          "online_loss": 0.013477709144353867,
          "train_loss": 0.013358290307223797,
          "validation_loss": 0.011148265562951565,
          "train_metric": 0.0013453979045152664,
          "validation_metric": 0.0011228123912587762,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.873954828000933
        },
        {
          "epoch": 239,
          "updates": 239,
          "samples_seen": 244736,
          "online_loss": 0.013358289375901222,
          "train_loss": 0.013240613043308258,
          "validation_loss": 0.011047950945794582,
          "train_metric": 0.0013335458934307098,
          "validation_metric": 0.0011127092875540257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8798768930009828
        },
        {
          "epoch": 240,
          "updates": 240,
          "samples_seen": 245760,
          "online_loss": 0.013240613043308258,
          "train_loss": 0.013124671764671803,
          "validation_loss": 0.010949228890240192,
          "train_metric": 0.0013218688545748591,
          "validation_metric": 0.0011027661385014653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8856096930012427
        },
        {
          "epoch": 241,
          "updates": 241,
          "samples_seen": 246784,
          "online_loss": 0.013124670833349228,
          "train_loss": 0.013010448776185513,
          "validation_loss": 0.010852059349417686,
          "train_metric": 0.0013103648088872433,
          "validation_metric": 0.001092979684472084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8911796320007852
        },
        {
          "epoch": 242,
          "updates": 242,
          "samples_seen": 247808,
          "online_loss": 0.013010447844862938,
          "train_loss": 0.012897924520075321,
          "validation_loss": 0.010756444185972214,
          "train_metric": 0.0012990316608920693,
          "validation_metric": 0.0010833496926352382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8969829360012227
        },
        {
          "epoch": 243,
          "updates": 243,
          "samples_seen": 248832,
          "online_loss": 0.012897923588752747,
          "train_loss": 0.01278708130121231,
          "validation_loss": 0.010662388056516647,
          "train_metric": 0.0012878677807748318,
          "validation_metric": 0.0010738767450675368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9027537350011698
        },
        {
          "epoch": 244,
          "updates": 244,
          "samples_seen": 249856,
          "online_loss": 0.012787080369889736,
          "train_loss": 0.012677901424467564,
          "validation_loss": 0.010569905862212181,
          "train_metric": 0.0012768716551363468,
          "validation_metric": 0.001064562238752842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.908567138001672
        },
        {
          "epoch": 245,
          "updates": 245,
          "samples_seen": 250880,
          "online_loss": 0.012677900493144989,
          "train_loss": 0.012570356950163841,
          "validation_loss": 0.010478966869413853,
          "train_metric": 0.001266040257178247,
          "validation_metric": 0.0010554032633081079,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9900891620000039
        },
        {
          "epoch": 246,
          "updates": 246,
          "samples_seen": 251904,
          "online_loss": 0.012570356018841267,
          "train_loss": 0.012464442290365696,
          "validation_loss": 0.010389539413154125,
          "train_metric": 0.0012553728884086013,
          "validation_metric": 0.0010463964426890016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9961395840000478
        },
        {
          "epoch": 247,
          "updates": 247,
          "samples_seen": 252928,
          "online_loss": 0.012464441359043121,
          "train_loss": 0.012360122054815292,
          "validation_loss": 0.010301602073013783,
          "train_metric": 0.0012448662891983986,
          "validation_metric": 0.0010375396814197302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.001741783000398
        },
        {
          "epoch": 248,
          "updates": 248,
          "samples_seen": 253952,
          "online_loss": 0.012360121123492718,
          "train_loss": 0.012257379479706287,
          "validation_loss": 0.010215160436928272,
          "train_metric": 0.0012345184804871678,
          "validation_metric": 0.001028833445161581,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0073881650005205
        },
        {
          "epoch": 249,
          "updates": 249,
          "samples_seen": 254976,
          "online_loss": 0.012257379479706287,
          "train_loss": 0.012156190350651741,
          "validation_loss": 0.010130196809768677,
          "train_metric": 0.0012243270175531507,
          "validation_metric": 0.0010202763369306922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.013388497000051
        },
        {
          "epoch": 250,
          "updates": 250,
          "samples_seen": 256000,
          "online_loss": 0.012156190350651741,
          "train_loss": 0.012056531384587288,
          "validation_loss": 0.010046692565083504,
          "train_metric": 0.0012142896885052323,
          "validation_metric": 0.0010118660284206271,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.019471716001135
        },
        {
          "epoch": 251,
          "updates": 251,
          "samples_seen": 257024,
          "online_loss": 0.012056531384587288,
          "train_loss": 0.01195838488638401,
          "validation_loss": 0.00996460672467947,
          "train_metric": 0.0012044047471135855,
          "validation_metric": 0.0010035987943410873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0252819000015734
        },
        {
          "epoch": 252,
          "updates": 252,
          "samples_seen": 258048,
          "online_loss": 0.01195838488638401,
          "train_loss": 0.011861705221235752,
          "validation_loss": 0.009883916936814785,
          "train_metric": 0.001194667536765337,
          "validation_metric": 0.000995471840724349,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0311017550011456
        },
        {
          "epoch": 253,
          "updates": 253,
          "samples_seen": 259072,
          "online_loss": 0.011861705221235752,
          "train_loss": 0.011766490526497364,
          "validation_loss": 0.009804601781070232,
          "train_metric": 0.001185077941045165,
          "validation_metric": 0.0009874835377559066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.036845660000836
        },
        {
          "epoch": 254,
          "updates": 254,
          "samples_seen": 260096,
          "online_loss": 0.011766492389142513,
          "train_loss": 0.01167270727455616,
          "validation_loss": 0.009726659394800663,
          "train_metric": 0.0011756321182474494,
          "validation_metric": 0.0009796335361897945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0424135940011183
        },
        {
          "epoch": 255,
          "updates": 255,
          "samples_seen": 261120,
          "online_loss": 0.011672706343233585,
          "train_loss": 0.011580324731767178,
          "validation_loss": 0.009650052525103092,
          "train_metric": 0.0011663277400657535,
          "validation_metric": 0.0009719179361127317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0482185630007734
        },
        {
          "epoch": 256,
          "updates": 256,
          "samples_seen": 262144,
          "online_loss": 0.011580323800444603,
          "train_loss": 0.011489317752420902,
          "validation_loss": 0.00957474298775196,
          "train_metric": 0.0011571618961170316,
          "validation_metric": 0.0009643329540267587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0539969920009753
        },
        {
          "epoch": 257,
          "updates": 257,
          "samples_seen": 263168,
          "online_loss": 0.011489317752420902,
          "train_loss": 0.011399668641388416,
          "validation_loss": 0.009500703774392605,
          "train_metric": 0.0011481328401714563,
          "validation_metric": 0.0009568760287947953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0596077520003746
        },
        {
          "epoch": 258,
          "updates": 258,
          "samples_seen": 264192,
          "online_loss": 0.011399669572710991,
          "train_loss": 0.011311333626508713,
          "validation_loss": 0.009427906945347786,
          "train_metric": 0.0011392361484467983,
          "validation_metric": 0.0009495442500337958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0653182820005895
        },
        {
          "epoch": 259,
          "updates": 259,
          "samples_seen": 265216,
          "online_loss": 0.011311334557831287,
          "train_loss": 0.011224311776459217,
          "validation_loss": 0.009356336668133736,
          "train_metric": 0.0011304715881124139,
          "validation_metric": 0.0009423360461369157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0711297180005204
        },
        {
          "epoch": 260,
          "updates": 260,
          "samples_seen": 266240,
          "online_loss": 0.011224311776459217,
          "train_loss": 0.011138547211885452,
          "validation_loss": 0.00928595568984747,
          "train_metric": 0.0011218336876481771,
          "validation_metric": 0.0009352475171908736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0769152880002366
        },
        {
          "epoch": 261,
          "updates": 261,
          "samples_seen": 267264,
          "online_loss": 0.011138548143208027,
          "train_loss": 0.011054027825593948,
          "validation_loss": 0.009216733276844025,
          "train_metric": 0.001113321166485548,
          "validation_metric": 0.0009282756946049631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0828366540008574
        },
        {
          "epoch": 262,
          "updates": 262,
          "samples_seen": 268288,
          "online_loss": 0.011054027825593948,
          "train_loss": 0.010970733128488064,
          "validation_loss": 0.009148631244897842,
          "train_metric": 0.0011049319291487336,
          "validation_metric": 0.0009214167366735637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0886092630007624
        },
        {
          "epoch": 263,
          "updates": 263,
          "samples_seen": 269312,
          "online_loss": 0.01097073219716549,
          "train_loss": 0.010888618417084217,
          "validation_loss": 0.009081624448299408,
          "train_metric": 0.0010966615518555045,
          "validation_metric": 0.0009146678494289517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0945047380009783
        },
        {
          "epoch": 264,
          "updates": 264,
          "samples_seen": 270336,
          "online_loss": 0.010888617485761642,
          "train_loss": 0.010807666927576065,
          "validation_loss": 0.009015685878694057,
          "train_metric": 0.0010885086376219988,
          "validation_metric": 0.0009080269373953342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1002400610013865
        },
        {
          "epoch": 265,
          "updates": 265,
          "samples_seen": 271360,
          "online_loss": 0.010807666927576065,
          "train_loss": 0.010727851651608944,
          "validation_loss": 0.00895079318434,
          "train_metric": 0.0010804698104038835,
          "validation_metric": 0.0009014911483973265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.105875721001212
        },
        {
          "epoch": 266,
          "updates": 266,
          "samples_seen": 272384,
          "online_loss": 0.010727851651608944,
          "train_loss": 0.010649146512150764,
          "validation_loss": 0.008886907249689102,
          "train_metric": 0.0010725429747253656,
          "validation_metric": 0.0008950568153522909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1117366930011485
        },
        {
          "epoch": 267,
          "updates": 267,
          "samples_seen": 273408,
          "online_loss": 0.010649146512150764,
          "train_loss": 0.010571524500846863,
          "validation_loss": 0.0088240010663867,
          "train_metric": 0.0010647251037880778,
          "validation_metric": 0.0008887210860848427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1174427480000304
        },
        {
          "epoch": 268,
          "updates": 268,
          "samples_seen": 274432,
          "online_loss": 0.010571522638201714,
          "train_loss": 0.010494955815374851,
          "validation_loss": 0.00876204390078783,
          "train_metric": 0.0010570134036242962,
          "validation_metric": 0.0008824811666272581,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1232215630006976
        },
        {
          "epoch": 269,
          "updates": 269,
          "samples_seen": 275456,
          "online_loss": 0.010494955815374851,
          "train_loss": 0.010419415310025215,
          "validation_loss": 0.008701012469828129,
          "train_metric": 0.0010494054295122623,
          "validation_metric": 0.0008763340883888304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.12903082000048
        },
        {
          "epoch": 270,
          "updates": 270,
          "samples_seen": 276480,
          "online_loss": 0.010419417172670364,
          "train_loss": 0.010344886220991611,
          "validation_loss": 0.008640888147056103,
          "train_metric": 0.0010418988531455398,
          "validation_metric": 0.0008702785708010197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1346762150005816
        },
        {
          "epoch": 271,
          "updates": 271,
          "samples_seen": 277504,
          "online_loss": 0.010344884358346462,
          "train_loss": 0.010271330364048481,
          "validation_loss": 0.008581623435020447,
          "train_metric": 0.0010344907641410828,
          "validation_metric": 0.0008643097244203091,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.140502900001593
        },
        {
          "epoch": 272,
          "updates": 272,
          "samples_seen": 278528,
          "online_loss": 0.010271328501403332,
          "train_loss": 0.010198733769357204,
          "validation_loss": 0.008523202501237392,
          "train_metric": 0.0010271790670230985,
          "validation_metric": 0.0008584257448092103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1465014970017364
        },
        {
          "epoch": 273,
          "updates": 273,
          "samples_seen": 279552,
          "online_loss": 0.01019873283803463,
          "train_loss": 0.010127069428563118,
          "validation_loss": 0.008465602993965149,
          "train_metric": 0.0010199613170698285,
          "validation_metric": 0.0008526245364919305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1521451940006955
        },
        {
          "epoch": 274,
          "updates": 274,
          "samples_seen": 280576,
          "online_loss": 0.010127069428563118,
          "train_loss": 0.010056310333311558,
          "validation_loss": 0.00840880535542965,
          "train_metric": 0.001012834720313549,
          "validation_metric": 0.0008469041204079986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.157763649000117
        },
        {
          "epoch": 275,
          "updates": 275,
          "samples_seen": 281600,
          "online_loss": 0.010056310333311558,
          "train_loss": 0.00998643133789301,
          "validation_loss": 0.008352772332727909,
          "train_metric": 0.0010057968320325017,
          "validation_metric": 0.0008412606548517942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1635842840005353
        },
        {
          "epoch": 276,
          "updates": 276,
          "samples_seen": 282624,
          "online_loss": 0.00998643133789301,
          "train_loss": 0.009917419403791428,
          "validation_loss": 0.00829748809337616,
          "train_metric": 0.0009988462552428246,
          "validation_metric": 0.0008356925682164729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.169215912001164
        },
        {
          "epoch": 277,
          "updates": 277,
          "samples_seen": 283648,
          "online_loss": 0.009917420335114002,
          "train_loss": 0.009849244728684425,
          "validation_loss": 0.00824293028563261,
          "train_metric": 0.0009919798467308283,
          "validation_metric": 0.0008301977650262415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1748765490010555
        },
        {
          "epoch": 278,
          "updates": 278,
          "samples_seen": 284672,
          "online_loss": 0.009849245660007,
          "train_loss": 0.009781887754797935,
          "validation_loss": 0.008189069107174873,
          "train_metric": 0.0009851959766820073,
          "validation_metric": 0.0008247729856520891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.180655480000496
        },
        {
          "epoch": 279,
          "updates": 279,
          "samples_seen": 285696,
          "online_loss": 0.009781887754797935,
          "train_loss": 0.00971533078700304,
          "validation_loss": 0.008135894313454628,
          "train_metric": 0.000978492433205247,
          "validation_metric": 0.0008194174151867628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1867514370005665
        },
        {
          "epoch": 280,
          "updates": 280,
          "samples_seen": 286720,
          "online_loss": 0.00971533078700304,
          "train_loss": 0.009649539366364479,
          "validation_loss": 0.008083371445536613,
          "train_metric": 0.0009718663059175014,
          "validation_metric": 0.0008141276193782687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1929363390008803
        },
        {
          "epoch": 281,
          "updates": 281,
          "samples_seen": 287744,
          "online_loss": 0.009649539366364479,
          "train_loss": 0.009584516286849976,
          "validation_loss": 0.008031489327549934,
          "train_metric": 0.0009653174201957881,
          "validation_metric": 0.0008089023176580667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.201075002001744
        },
        {
          "epoch": 282,
          "updates": 282,
          "samples_seen": 288768,
          "online_loss": 0.009584516286849976,
          "train_loss": 0.00952022522687912,
          "validation_loss": 0.007980226539075375,
          "train_metric": 0.0009588421671651304,
          "validation_metric": 0.0008037392399273813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.206662076001521
        },
        {
          "epoch": 283,
          "updates": 283,
          "samples_seen": 289792,
          "online_loss": 0.00952022336423397,
          "train_loss": 0.009456655010581017,
          "validation_loss": 0.007929569110274315,
          "train_metric": 0.0009524397319182754,
          "validation_metric": 0.0007986371638253331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3062252850013465
        },
        {
          "epoch": 284,
          "updates": 284,
          "samples_seen": 290816,
          "online_loss": 0.009456655941903591,
          "train_loss": 0.009393783286213875,
          "validation_loss": 0.007879496552050114,
          "train_metric": 0.000946107495110482,
          "validation_metric": 0.0007935939938761294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3135738400014816
        },
        {
          "epoch": 285,
          "updates": 285,
          "samples_seen": 291840,
          "online_loss": 0.00939378421753645,
          "train_loss": 0.00933159701526165,
          "validation_loss": 0.007829985581338406,
          "train_metric": 0.0009398443507961929,
          "validation_metric": 0.0007886074599809945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3195458380014315
        },
        {
          "epoch": 286,
          "updates": 286,
          "samples_seen": 292864,
          "online_loss": 0.009331596083939075,
          "train_loss": 0.009270076639950275,
          "validation_loss": 0.007781021762639284,
          "train_metric": 0.0009336482034996152,
          "validation_metric": 0.0007836759905330837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3270776430017577
        },
        {
          "epoch": 287,
          "updates": 287,
          "samples_seen": 293888,
          "online_loss": 0.009270076639950275,
          "train_loss": 0.00920921191573143,
          "validation_loss": 0.007732592988759279,
          "train_metric": 0.0009275181218981743,
          "validation_metric": 0.0007787983049638569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.333460207000826
        },
        {
          "epoch": 288,
          "updates": 288,
          "samples_seen": 294912,
          "online_loss": 0.00920921191573143,
          "train_loss": 0.009148967452347279,
          "validation_loss": 0.007684687152504921,
          "train_metric": 0.0009214504971168935,
          "validation_metric": 0.0007739735301584005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3395826340001804
        },
        {
          "epoch": 289,
          "updates": 289,
          "samples_seen": 295936,
          "online_loss": 0.009148968383669853,
          "train_loss": 0.009089349769055843,
          "validation_loss": 0.007637280970811844,
          "train_metric": 0.0009154460858553648,
          "validation_metric": 0.0007691989303566515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3502507350003725
        },
        {
          "epoch": 290,
          "updates": 290,
          "samples_seen": 296960,
          "online_loss": 0.009089350700378418,
          "train_loss": 0.009030335582792759,
          "validation_loss": 0.007590358145534992,
          "train_metric": 0.000909502268768847,
          "validation_metric": 0.0007644729921594262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.356187579000107
        },
        {
          "epoch": 291,
          "updates": 291,
          "samples_seen": 297984,
          "online_loss": 0.009030334651470184,
          "train_loss": 0.008971910923719406,
          "validation_loss": 0.007543907500803471,
          "train_metric": 0.0009036179981194437,
          "validation_metric": 0.0007597946678288281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3624579740007903
        },
        {
          "epoch": 292,
          "updates": 292,
          "samples_seen": 299008,
          "online_loss": 0.00897191185504198,
          "train_loss": 0.00891406275331974,
          "validation_loss": 0.007497924845665693,
          "train_metric": 0.0008977917023003101,
          "validation_metric": 0.0007551635499112308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.368750212001032
        },
        {
          "epoch": 293,
          "updates": 293,
          "samples_seen": 300032,
          "online_loss": 0.00891406275331974,
          "train_loss": 0.00885678268969059,
          "validation_loss": 0.007452378980815411,
          "train_metric": 0.0008920226828195155,
          "validation_metric": 0.0007505762623623013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.374393071000668
        },
        {
          "epoch": 294,
          "updates": 294,
          "samples_seen": 301056,
          "online_loss": 0.00885678268969059,
          "train_loss": 0.008800046518445015,
          "validation_loss": 0.007407279219478369,
          "train_metric": 0.0008863083785399795,
          "validation_metric": 0.000746033969335258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3806159510004363
        },
        {
          "epoch": 295,
          "updates": 295,
          "samples_seen": 302080,
          "online_loss": 0.00880004558712244,
          "train_loss": 0.008743850514292717,
          "validation_loss": 0.007362594362348318,
          "train_metric": 0.0008806486148387194,
          "validation_metric": 0.0007415335858240724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3865450320008676
        },
        {
          "epoch": 296,
          "updates": 296,
          "samples_seen": 303104,
          "online_loss": 0.008743851445615292,
          "train_loss": 0.008688176050782204,
          "validation_loss": 0.0073183318600058556,
          "train_metric": 0.0008750414126552641,
          "validation_metric": 0.0007370755774900317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3921156770011294
        },
        {
          "epoch": 297,
          "updates": 297,
          "samples_seen": 304128,
          "online_loss": 0.008688176050782204,
          "train_loss": 0.008633018471300602,
          "validation_loss": 0.0072744679637253284,
          "train_metric": 0.000869485957082361,
          "validation_metric": 0.0007326577324420214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3984826580017398
        },
        {
          "epoch": 298,
          "updates": 298,
          "samples_seen": 305152,
          "online_loss": 0.008633018471300602,
          "train_loss": 0.008578365668654442,
          "validation_loss": 0.007230993360280991,
          "train_metric": 0.0008639813750050962,
          "validation_metric": 0.0007282791193574667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.404422722000163
        },
        {
          "epoch": 299,
          "updates": 299,
          "samples_seen": 306176,
          "online_loss": 0.008578364737331867,
          "train_loss": 0.008524205535650253,
          "validation_loss": 0.007187894079834223,
          "train_metric": 0.0008585267933085561,
          "validation_metric": 0.0007239383994601667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.410046802000579
        },
        {
          "epoch": 300,
          "updates": 300,
          "samples_seen": 307200,
          "online_loss": 0.008524205535650253,
          "train_loss": 0.008470526896417141,
          "validation_loss": 0.007145164534449577,
          "train_metric": 0.0008531204657629132,
          "validation_metric": 0.0007196348742581904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4164002910001727
        },
        {
          "epoch": 301,
          "updates": 301,
          "samples_seen": 308224,
          "online_loss": 0.008470526896417141,
          "train_loss": 0.008417325094342232,
          "validation_loss": 0.007102798204869032,
          "train_metric": 0.0008477622177451849,
          "validation_metric": 0.000715367728844285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.422324896000646
        },
        {
          "epoch": 302,
          "updates": 302,
          "samples_seen": 309248,
          "online_loss": 0.008417324163019657,
          "train_loss": 0.008364587090909481,
          "validation_loss": 0.00706078065559268,
          "train_metric": 0.0008424505940638483,
          "validation_metric": 0.000711135973688215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4284321510003792
        },
        {
          "epoch": 303,
          "updates": 303,
          "samples_seen": 310272,
          "online_loss": 0.008364588022232056,
          "train_loss": 0.008312288671731949,
          "validation_loss": 0.007019096054136753,
          "train_metric": 0.0008371832664124668,
          "validation_metric": 0.0007069376297295094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4348586320011236
        },
        {
          "epoch": 304,
          "updates": 304,
          "samples_seen": 311296,
          "online_loss": 0.008312288671731949,
          "train_loss": 0.008260449394583702,
          "validation_loss": 0.006977748591452837,
          "train_metric": 0.0008319622720591724,
          "validation_metric": 0.0007027732208371162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.440696116000254
        },
        {
          "epoch": 305,
          "updates": 305,
          "samples_seen": 312320,
          "online_loss": 0.008260449394583702,
          "train_loss": 0.00820904690772295,
          "validation_loss": 0.006936721503734589,
          "train_metric": 0.0008267851662822068,
          "validation_metric": 0.0006986412336118519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.446662706001007
        },
        {
          "epoch": 306,
          "updates": 306,
          "samples_seen": 313344,
          "online_loss": 0.00820904690772295,
          "train_loss": 0.008158066309988499,
          "validation_loss": 0.006896011997014284,
          "train_metric": 0.0008216504938900471,
          "validation_metric": 0.0006945410277694464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4528279050009587
        },
        {
          "epoch": 307,
          "updates": 307,
          "samples_seen": 314368,
          "online_loss": 0.008158065378665924,
          "train_loss": 0.008107513189315796,
          "validation_loss": 0.006855614949017763,
          "train_metric": 0.0008165590697899461,
          "validation_metric": 0.0006904724286869168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.458674946001338
        },
        {
          "epoch": 308,
          "updates": 308,
          "samples_seen": 315392,
          "online_loss": 0.008107513189315796,
          "train_loss": 0.008057371713221073,
          "validation_loss": 0.00681550707668066,
          "train_metric": 0.0008115090313367546,
          "validation_metric": 0.000686432933434844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4651834260002943
        },
        {
          "epoch": 309,
          "updates": 309,
          "samples_seen": 316416,
          "online_loss": 0.008057371713221073,
          "train_loss": 0.008007632568478584,
          "validation_loss": 0.0067756990902125835,
          "train_metric": 0.0008064995636232197,
          "validation_metric": 0.0006824235897511244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4708150740007113
        },
        {
          "epoch": 310,
          "updates": 310,
          "samples_seen": 317440,
          "online_loss": 0.008007633499801159,
          "train_loss": 0.007958292961120605,
          "validation_loss": 0.006736179813742638,
          "train_metric": 0.0008015300845727324,
          "validation_metric": 0.0006784433498978615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4763949920015875
        },
        {
          "epoch": 311,
          "updates": 311,
          "samples_seen": 318464,
          "online_loss": 0.007958292961120605,
          "train_loss": 0.00790934357792139,
          "validation_loss": 0.006696933414787054,
          "train_metric": 0.0007966002449393272,
          "validation_metric": 0.0006744905840605497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.482114314001592
        },
        {
          "epoch": 312,
          "updates": 312,
          "samples_seen": 319488,
          "online_loss": 0.007909344509243965,
          "train_loss": 0.007860783487558365,
          "validation_loss": 0.006657956633716822,
          "train_metric": 0.0007917093462310731,
          "validation_metric": 0.0006705649429932237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4877873820005334
        },
        {
          "epoch": 313,
          "updates": 313,
          "samples_seen": 320512,
          "online_loss": 0.007860783487558365,
          "train_loss": 0.007812597788870335,
          "validation_loss": 0.006619246676564217,
          "train_metric": 0.0007868562825024128,
          "validation_metric": 0.0006666663102805614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4932713040016097
        },
        {
          "epoch": 314,
          "updates": 314,
          "samples_seen": 321536,
          "online_loss": 0.00781259872019291,
          "train_loss": 0.007764787878841162,
          "validation_loss": 0.006580804008990526,
          "train_metric": 0.0007820409955456853,
          "validation_metric": 0.0006627943948842585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4990450680015783
        },
        {
          "epoch": 315,
          "updates": 315,
          "samples_seen": 322560,
          "online_loss": 0.007764787413179874,
          "train_loss": 0.0077173421159386635,
          "validation_loss": 0.006542614195495844,
          "train_metric": 0.0007772624958306551,
          "validation_metric": 0.0006589481490664184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.504616086000169
        },
        {
          "epoch": 316,
          "updates": 316,
          "samples_seen": 323584,
          "online_loss": 0.007717342581599951,
          "train_loss": 0.0076702553778886795,
          "validation_loss": 0.006504668388515711,
          "train_metric": 0.0007725199102424085,
          "validation_metric": 0.000655126292258501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5101427660010813
        },
        {
          "epoch": 317,
          "updates": 317,
          "samples_seen": 324608,
          "online_loss": 0.0076702553778886795,
          "train_loss": 0.007623523473739624,
          "validation_loss": 0.006466970779001713,
          "train_metric": 0.0007678132969886065,
          "validation_metric": 0.0006513296975754201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.516079102000731
        },
        {
          "epoch": 318,
          "updates": 318,
          "samples_seen": 325632,
          "online_loss": 0.007623524405062199,
          "train_loss": 0.0075771389529109,
          "validation_loss": 0.006429516244679689,
          "train_metric": 0.0007631417247466743,
          "validation_metric": 0.0006475573172792792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5218580610016943
        },
        {
          "epoch": 319,
          "updates": 319,
          "samples_seen": 326656,
          "online_loss": 0.0075771380215883255,
          "train_loss": 0.007531104143708944,
          "validation_loss": 0.006392296869307756,
          "train_metric": 0.000758505251724273,
          "validation_metric": 0.0006438087439164519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5276118730016606
        },
        {
          "epoch": 320,
          "updates": 320,
          "samples_seen": 327680,
          "online_loss": 0.007531104143708944,
          "train_loss": 0.0074854036793112755,
          "validation_loss": 0.006355307996273041,
          "train_metric": 0.0007539024809375405,
          "validation_metric": 0.0006400833372026682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6111062690015387
        },
        {
          "epoch": 321,
          "updates": 321,
          "samples_seen": 328704,
          "online_loss": 0.0074854036793112755,
          "train_loss": 0.007440037094056606,
          "validation_loss": 0.006318538449704647,
          "train_metric": 0.0007493332377634943,
          "validation_metric": 0.0006363801076076925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6169070890009607
        },
        {
          "epoch": 322,
          "updates": 322,
          "samples_seen": 329728,
          "online_loss": 0.007440037094056606,
          "train_loss": 0.007394999265670776,
          "validation_loss": 0.006281993817538023,
          "train_metric": 0.0007447972893714905,
          "validation_metric": 0.0006326994625851512,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6226370280000992
        },
        {
          "epoch": 323,
          "updates": 323,
          "samples_seen": 330752,
          "online_loss": 0.007394999265670776,
          "train_loss": 0.007350290659815073,
          "validation_loss": 0.00624568248167634,
          "train_metric": 0.0007402942865155637,
          "validation_metric": 0.0006290421588346362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6284507360014686
        },
        {
          "epoch": 324,
          "updates": 324,
          "samples_seen": 331776,
          "online_loss": 0.007350289728492498,
          "train_loss": 0.007305893115699291,
          "validation_loss": 0.006209571845829487,
          "train_metric": 0.0007358227157965302,
          "validation_metric": 0.0006254054023884237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.634186634000798
        },
        {
          "epoch": 325,
          "updates": 325,
          "samples_seen": 332800,
          "online_loss": 0.007305892650038004,
          "train_loss": 0.007261819671839476,
          "validation_loss": 0.006173677276819944,
          "train_metric": 0.000731383915990591,
          "validation_metric": 0.0006217901245690882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.639895153000907
        },
        {
          "epoch": 326,
          "updates": 326,
          "samples_seen": 333824,
          "online_loss": 0.007261819206178188,
          "train_loss": 0.007218053098767996,
          "validation_loss": 0.006137995980679989,
          "train_metric": 0.000726975966244936,
          "validation_metric": 0.0006181964999996126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6458153010007663
        },
        {
          "epoch": 327,
          "updates": 327,
          "samples_seen": 334848,
          "online_loss": 0.007218053098767996,
          "train_loss": 0.007174591533839703,
          "validation_loss": 0.006102514453232288,
          "train_metric": 0.0007225985173135996,
          "validation_metric": 0.0006146228988654912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6516331900002115
        },
        {
          "epoch": 328,
          "updates": 328,
          "samples_seen": 335872,
          "online_loss": 0.007174591068178415,
          "train_loss": 0.007131432183086872,
          "validation_loss": 0.006067241076380014,
          "train_metric": 0.0007182517438195646,
          "validation_metric": 0.0006110703689046204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.657357324000259
        },
        {
          "epoch": 329,
          "updates": 329,
          "samples_seen": 336896,
          "online_loss": 0.007131432183086872,
          "train_loss": 0.007088575977832079,
          "validation_loss": 0.006032171193510294,
          "train_metric": 0.0007139353547245264,
          "validation_metric": 0.0006075382698327303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6632121990005544
        },
        {
          "epoch": 330,
          "updates": 330,
          "samples_seen": 337920,
          "online_loss": 0.007088575512170792,
          "train_loss": 0.007046013604849577,
          "validation_loss": 0.005997294560074806,
          "train_metric": 0.0007096485933288932,
          "validation_metric": 0.0006040255539119244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.668915421001657
        },
        {
          "epoch": 331,
          "updates": 331,
          "samples_seen": 338944,
          "online_loss": 0.007046013604849577,
          "train_loss": 0.007003741804510355,
          "validation_loss": 0.005962613504379988,
          "train_metric": 0.0007053912268020213,
          "validation_metric": 0.0006005325121805072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6745324000003166
        },
        {
          "epoch": 332,
          "updates": 332,
          "samples_seen": 339968,
          "online_loss": 0.007003741338849068,
          "train_loss": 0.006961757782846689,
          "validation_loss": 0.005928117781877518,
          "train_metric": 0.0007011627312749624,
          "validation_metric": 0.0005970583879388869,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.680435591000787
        },
        {
          "epoch": 333,
          "updates": 333,
          "samples_seen": 340992,
          "online_loss": 0.006961757782846689,
          "train_loss": 0.006920062005519867,
          "validation_loss": 0.005893822759389877,
          "train_metric": 0.0006969633395783603,
          "validation_metric": 0.000593604170717299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.686518413000158
        },
        {
          "epoch": 334,
          "updates": 334,
          "samples_seen": 342016,
          "online_loss": 0.006920062005519867,
          "train_loss": 0.006878640502691269,
          "validation_loss": 0.005859720055013895,
          "train_metric": 0.0006927914801053703,
          "validation_metric": 0.0005901695694774389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6941845930014097
        },
        {
          "epoch": 335,
          "updates": 335,
          "samples_seen": 343040,
          "online_loss": 0.006878640502691269,
          "train_loss": 0.006837498862296343,
          "validation_loss": 0.0058257970958948135,
          "train_metric": 0.0006886478513479233,
          "validation_metric": 0.000586753012612462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7004776840003615
        },
        {
          "epoch": 336,
          "updates": 336,
          "samples_seen": 344064,
          "online_loss": 0.00683749932795763,
          "train_loss": 0.006796634290367365,
          "validation_loss": 0.005792055744677782,
          "train_metric": 0.0006845321622677147,
          "validation_metric": 0.0005833546747453511,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.706268058000205
        },
        {
          "epoch": 337,
          "updates": 337,
          "samples_seen": 345088,
          "online_loss": 0.006796634756028652,
          "train_loss": 0.0067560407333076,
          "validation_loss": 0.005758501589298248,
          "train_metric": 0.0006804436561651528,
          "validation_metric": 0.0005799751961603761,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7129196260011668
        },
        {
          "epoch": 338,
          "updates": 338,
          "samples_seen": 346112,
          "online_loss": 0.006756040267646313,
          "train_loss": 0.006715713068842888,
          "validation_loss": 0.005725125782191753,
          "train_metric": 0.000676381983794272,
          "validation_metric": 0.0005766138201579452,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.718812610000896
        },
        {
          "epoch": 339,
          "updates": 339,
          "samples_seen": 347136,
          "online_loss": 0.006715712603181601,
          "train_loss": 0.006675653159618378,
          "validation_loss": 0.005691926460713148,
          "train_metric": 0.000672347319778055,
          "validation_metric": 0.0005732699646614492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7243730680002045
        },
        {
          "epoch": 340,
          "updates": 340,
          "samples_seen": 348160,
          "online_loss": 0.006675653625279665,
          "train_loss": 0.006635856814682484,
          "validation_loss": 0.00565890409052372,
          "train_metric": 0.0006683391984552145,
          "validation_metric": 0.0005699440953321755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.730286165000507
        },
        {
          "epoch": 341,
          "updates": 341,
          "samples_seen": 349184,
          "online_loss": 0.006635856814682484,
          "train_loss": 0.0065963175147771835,
          "validation_loss": 0.005626062396913767,
          "train_metric": 0.0006643570377491415,
          "validation_metric": 0.0005666363867931068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7359966920012084
        },
        {
          "epoch": 342,
          "updates": 342,
          "samples_seen": 350208,
          "online_loss": 0.0065963175147771835,
          "train_loss": 0.006557036191225052,
          "validation_loss": 0.005593384616076946,
          "train_metric": 0.0006604006048291922,
          "validation_metric": 0.0005633452092297375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7417278150005586
        },
        {
          "epoch": 343,
          "updates": 343,
          "samples_seen": 351232,
          "online_loss": 0.006557035725563765,
          "train_loss": 0.006518014241009951,
          "validation_loss": 0.0055608805269002914,
          "train_metric": 0.0006564705399796367,
          "validation_metric": 0.0005600714939646423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.747580581000875
        },
        {
          "epoch": 344,
          "updates": 344,
          "samples_seen": 352256,
          "online_loss": 0.006518014706671238,
          "train_loss": 0.0064792390912771225,
          "validation_loss": 0.00552855571731925,
          "train_metric": 0.0006525652715936303,
          "validation_metric": 0.0005568158230744302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7533332090006297
        },
        {
          "epoch": 345,
          "updates": 345,
          "samples_seen": 353280,
          "online_loss": 0.00647923955693841,
          "train_loss": 0.006440713070333004,
          "validation_loss": 0.005496395286172628,
          "train_metric": 0.0006486850325018167,
          "validation_metric": 0.0005535767995752394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7596261510007025
        },
        {
          "epoch": 346,
          "updates": 346,
          "samples_seen": 354304,
          "online_loss": 0.006440713070333004,
          "train_loss": 0.0064024352468550205,
          "validation_loss": 0.005464404821395874,
          "train_metric": 0.000644829822704196,
          "validation_metric": 0.0005503548891283572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7653528210012155
        },
        {
          "epoch": 347,
          "updates": 347,
          "samples_seen": 355328,
          "online_loss": 0.0064024352468550205,
          "train_loss": 0.006364408414810896,
          "validation_loss": 0.00543257175013423,
          "train_metric": 0.0006409999332390726,
          "validation_metric": 0.0005471486365422606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7710396210004546
        },
        {
          "epoch": 348,
          "updates": 348,
          "samples_seen": 356352,
          "online_loss": 0.006364408414810896,
          "train_loss": 0.006326612085103989,
          "validation_loss": 0.005400918889790773,
          "train_metric": 0.0006371933268383145,
          "validation_metric": 0.0005439607775770128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7771372700008214
        },
        {
          "epoch": 349,
          "updates": 349,
          "samples_seen": 357376,
          "online_loss": 0.006326613016426563,
          "train_loss": 0.006289068143814802,
          "validation_loss": 0.005369429010897875,
          "train_metric": 0.0006334118079394102,
          "validation_metric": 0.0005407892749644816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7830384760000015
        },
        {
          "epoch": 350,
          "updates": 350,
          "samples_seen": 358400,
          "online_loss": 0.006289067678153515,
          "train_loss": 0.006251761224120855,
          "validation_loss": 0.005338091403245926,
          "train_metric": 0.000629654445219785,
          "validation_metric": 0.0005376330227591097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7887558820002596
        },
        {
          "epoch": 351,
          "updates": 351,
          "samples_seen": 359424,
          "online_loss": 0.006251760758459568,
          "train_loss": 0.0062146843411028385,
          "validation_loss": 0.005306922364979982,
          "train_metric": 0.0006259201909415424,
          "validation_metric": 0.0005344938253983855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.794803259001128
        },
        {
          "epoch": 352,
          "updates": 352,
          "samples_seen": 360448,
          "online_loss": 0.0062146843411028385,
          "train_loss": 0.006177845411002636,
          "validation_loss": 0.005275918170809746,
          "train_metric": 0.000622210034634918,
          "validation_metric": 0.0005313711590133607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8004907240010652
        },
        {
          "epoch": 353,
          "updates": 353,
          "samples_seen": 361472,
          "online_loss": 0.006177845411002636,
          "train_loss": 0.006141239777207375,
          "validation_loss": 0.005245075095444918,
          "train_metric": 0.0006185232778079808,
          "validation_metric": 0.0005282646743580699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.80610327300019
        },
        {
          "epoch": 354,
          "updates": 354,
          "samples_seen": 362496,
          "online_loss": 0.006141240242868662,
          "train_loss": 0.006104866974055767,
          "validation_loss": 0.005214391276240349,
          "train_metric": 0.000614859804045409,
          "validation_metric": 0.0005251744296401739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8122366210009204
        },
        {
          "epoch": 355,
          "updates": 355,
          "samples_seen": 363520,
          "online_loss": 0.006104866974055767,
          "train_loss": 0.006068718153983355,
          "validation_loss": 0.0051838611252605915,
          "train_metric": 0.0006112190894782543,
          "validation_metric": 0.0005220995517447591,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.818028013000003
        },
        {
          "epoch": 356,
          "updates": 356,
          "samples_seen": 364544,
          "online_loss": 0.006068718153983355,
          "train_loss": 0.006032796576619148,
          "validation_loss": 0.005153493024408817,
          "train_metric": 0.0006076011923141778,
          "validation_metric": 0.0005190409719944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.823716504000913
        },
        {
          "epoch": 357,
          "updates": 357,
          "samples_seen": 365568,
          "online_loss": 0.006032796576619148,
          "train_loss": 0.005997104104608297,
          "validation_loss": 0.005123281851410866,
          "train_metric": 0.0006040063453838229,
          "validation_metric": 0.0005159982247278094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.829770764001296
        },
        {
          "epoch": 358,
          "updates": 358,
          "samples_seen": 366592,
          "online_loss": 0.005997104570269585,
          "train_loss": 0.005961638875305653,
          "validation_loss": 0.005093229003250599,
          "train_metric": 0.000600434432271868,
          "validation_metric": 0.0005129713681526482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8355052360002446
        },
        {
          "epoch": 359,
          "updates": 359,
          "samples_seen": 367616,
          "online_loss": 0.005961639806628227,
          "train_loss": 0.005926386918872595,
          "validation_loss": 0.005063326098024845,
          "train_metric": 0.000596884055994451,
          "validation_metric": 0.0005099596455693245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8411233850001736
        },
        {
          "epoch": 360,
          "updates": 360,
          "samples_seen": 368640,
          "online_loss": 0.005926387384533882,
          "train_loss": 0.0058913580141961575,
          "validation_loss": 0.005033584777265787,
          "train_metric": 0.0005933560896664858,
          "validation_metric": 0.0005069642211310565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9263449770005536
        },
        {
          "epoch": 361,
          "updates": 361,
          "samples_seen": 369664,
          "online_loss": 0.005891358479857445,
          "train_loss": 0.005856554955244064,
          "validation_loss": 0.005004005506634712,
          "train_metric": 0.0005898507661186159,
          "validation_metric": 0.0005039851530455053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.932519576001141
        },
        {
          "epoch": 362,
          "updates": 362,
          "samples_seen": 370688,
          "online_loss": 0.005856554955244064,
          "train_loss": 0.005821968428790569,
          "validation_loss": 0.0049745733849704266,
          "train_metric": 0.0005863673868589103,
          "validation_metric": 0.000501020869705826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.938323129001219
        },
        {
          "epoch": 363,
          "updates": 363,
          "samples_seen": 371712,
          "online_loss": 0.005821968894451857,
          "train_loss": 0.005787594243884087,
          "validation_loss": 0.004945292603224516,
          "train_metric": 0.0005829053116030991,
          "validation_metric": 0.0004980718367733061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9443482940005197
        },
        {
          "epoch": 364,
          "updates": 364,
          "samples_seen": 372736,
          "online_loss": 0.005787594709545374,
          "train_loss": 0.005753437522798777,
          "validation_loss": 0.004916168749332428,
          "train_metric": 0.0005794651224277914,
          "validation_metric": 0.0004951385199092329,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9501405360006174
        },
        {
          "epoch": 365,
          "updates": 365,
          "samples_seen": 373760,
          "online_loss": 0.005753437057137489,
          "train_loss": 0.005719493143260479,
          "validation_loss": 0.004887202754616737,
          "train_metric": 0.0005760464118793607,
          "validation_metric": 0.000492221093736589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.955437079001058
        },
        {
          "epoch": 366,
          "updates": 366,
          "samples_seen": 374784,
          "online_loss": 0.005719493143260479,
          "train_loss": 0.005685767158865929,
          "validation_loss": 0.004858385771512985,
          "train_metric": 0.0005726495874114335,
          "validation_metric": 0.0004893188597634435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9612096090004343
        },
        {
          "epoch": 367,
          "updates": 367,
          "samples_seen": 375808,
          "online_loss": 0.005685767158865929,
          "train_loss": 0.0056522502563893795,
          "validation_loss": 0.00482972152531147,
          "train_metric": 0.0005692739505320787,
          "validation_metric": 0.0004864318179897964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9670837100002245
        },
        {
          "epoch": 368,
          "updates": 368,
          "samples_seen": 376832,
          "online_loss": 0.005652249790728092,
          "train_loss": 0.005618936847895384,
          "validation_loss": 0.004801205825060606,
          "train_metric": 0.0005659187445417047,
          "validation_metric": 0.00048355991020798683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9728085470014776
        },
        {
          "epoch": 369,
          "updates": 369,
          "samples_seen": 377856,
          "online_loss": 0.005618937313556671,
          "train_loss": 0.0055858418345451355,
          "validation_loss": 0.004772844724357128,
          "train_metric": 0.0005625855992548168,
          "validation_metric": 0.00048070342745631933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.978781414001787
        },
        {
          "epoch": 370,
          "updates": 370,
          "samples_seen": 378880,
          "online_loss": 0.005585842300206423,
          "train_loss": 0.005552951246500015,
          "validation_loss": 0.004744637291878462,
          "train_metric": 0.0005592729430645704,
          "validation_metric": 0.00047786245704628527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9846293480004533
        },
        {
          "epoch": 371,
          "updates": 371,
          "samples_seen": 379904,
          "online_loss": 0.005552951246500015,
          "train_loss": 0.005520271137356758,
          "validation_loss": 0.004716580733656883,
          "train_metric": 0.0005559815326705575,
          "validation_metric": 0.00047503673704341054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9906212100013363
        },
        {
          "epoch": 372,
          "updates": 372,
          "samples_seen": 380928,
          "online_loss": 0.005520271137356758,
          "train_loss": 0.005487794987857342,
          "validation_loss": 0.004688672721385956,
          "train_metric": 0.0005527105531655252,
          "validation_metric": 0.00047222591820172966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9965690880017064
        },
        {
          "epoch": 373,
          "updates": 373,
          "samples_seen": 381952,
          "online_loss": 0.005487794987857342,
          "train_loss": 0.005455524194985628,
          "validation_loss": 0.004660910461097956,
          "train_metric": 0.0005494604702107608,
          "validation_metric": 0.0004694298841059208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.002644123000209
        },
        {
          "epoch": 374,
          "updates": 374,
          "samples_seen": 382976,
          "online_loss": 0.005455524194985628,
          "train_loss": 0.005423455033451319,
          "validation_loss": 0.0046333023346960545,
          "train_metric": 0.0005462305271066725,
          "validation_metric": 0.0004666491877287626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.008222646001741
        },
        {
          "epoch": 375,
          "updates": 375,
          "samples_seen": 384000,
          "online_loss": 0.005423454567790031,
          "train_loss": 0.005391598213464022,
          "validation_loss": 0.004605852533131838,
          "train_metric": 0.0005430220626294613,
          "validation_metric": 0.00046388470218516886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0138267170004838
        },
        {
          "epoch": 376,
          "updates": 376,
          "samples_seen": 385024,
          "online_loss": 0.005391599144786596,
          "train_loss": 0.005359936039894819,
          "validation_loss": 0.0045785363763570786,
          "train_metric": 0.0005398330977186561,
          "validation_metric": 0.000461133400676772,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0223492570003145
        },
        {
          "epoch": 377,
          "updates": 377,
          "samples_seen": 386048,
          "online_loss": 0.005359936505556107,
          "train_loss": 0.005328480154275894,
          "validation_loss": 0.00455138087272644,
          "train_metric": 0.0005366650293581188,
          "validation_metric": 0.00045839836820960045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.039632865000385
        },
        {
          "epoch": 378,
          "updates": 378,
          "samples_seen": 387072,
          "online_loss": 0.005328480154275894,
          "train_loss": 0.005297222640365362,
          "validation_loss": 0.004524379502981901,
          "train_metric": 0.000533516809809953,
          "validation_metric": 0.00045567896449938416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.066923123000379
        },
        {
          "epoch": 379,
          "updates": 379,
          "samples_seen": 388096,
          "online_loss": 0.005297222640365362,
          "train_loss": 0.005266167223453522,
          "validation_loss": 0.00449750991538167,
          "train_metric": 0.0005303890793584287,
          "validation_metric": 0.00045297268661670387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.086992956001268
        },
        {
          "epoch": 380,
          "updates": 380,
          "samples_seen": 389120,
          "online_loss": 0.005266167223453522,
          "train_loss": 0.005235312506556511,
          "validation_loss": 0.004470801912248135,
          "train_metric": 0.0005272815469652414,
          "validation_metric": 0.0004502827941905707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1047436210010346
        },
        {
          "epoch": 381,
          "updates": 381,
          "samples_seen": 390144,
          "online_loss": 0.005235312040895224,
          "train_loss": 0.005204654298722744,
          "validation_loss": 0.004444245249032974,
          "train_metric": 0.0005241937469691038,
          "validation_metric": 0.0004476080648601055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1223790210005973
        },
        {
          "epoch": 382,
          "updates": 382,
          "samples_seen": 391168,
          "online_loss": 0.005204654298722744,
          "train_loss": 0.005174197256565094,
          "validation_loss": 0.004417827818542719,
          "train_metric": 0.0005211262032389641,
          "validation_metric": 0.0004449474217835814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.139769156001421
        },
        {
          "epoch": 383,
          "updates": 383,
          "samples_seen": 392192,
          "online_loss": 0.005174197256565094,
          "train_loss": 0.005143935792148113,
          "validation_loss": 0.004391561262309551,
          "train_metric": 0.0005180783336982131,
          "validation_metric": 0.00044230197090655565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1546959770003014
        },
        {
          "epoch": 384,
          "updates": 384,
          "samples_seen": 393216,
          "online_loss": 0.005143935326486826,
          "train_loss": 0.005113875959068537,
          "validation_loss": 0.004365453030914068,
          "train_metric": 0.0005150508368387818,
          "validation_metric": 0.00043967243982478976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1698805300002277
        },
        {
          "epoch": 385,
          "updates": 385,
          "samples_seen": 394240,
          "online_loss": 0.005113875959068537,
          "train_loss": 0.005084005184471607,
          "validation_loss": 0.004339486360549927,
          "train_metric": 0.0005120424320921302,
          "validation_metric": 0.00043705716961994767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1826144970000314
        },
        {
          "epoch": 386,
          "updates": 386,
          "samples_seen": 395264,
          "online_loss": 0.0050840056501328945,
          "train_loss": 0.005054331850260496,
          "validation_loss": 0.004313659388571978,
          "train_metric": 0.0005090538179501891,
          "validation_metric": 0.00043445598566904664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1885479370012035
        },
        {
          "epoch": 387,
          "updates": 387,
          "samples_seen": 396288,
          "online_loss": 0.005054331384599209,
          "train_loss": 0.005024855025112629,
          "validation_loss": 0.004287994001060724,
          "train_metric": 0.0005060849944129586,
          "validation_metric": 0.0004318710707593709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1952761390002706
        },
        {
          "epoch": 388,
          "updates": 388,
          "samples_seen": 397312,
          "online_loss": 0.005024854559451342,
          "train_loss": 0.004995578434318304,
          "validation_loss": 0.004262478090822697,
          "train_metric": 0.0005031363107264042,
          "validation_metric": 0.0004293012316338718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2012221940003656
        },
        {
          "epoch": 389,
          "updates": 389,
          "samples_seen": 398336,
          "online_loss": 0.004995578434318304,
          "train_loss": 0.00496648857370019,
          "validation_loss": 0.004237092100083828,
          "train_metric": 0.0005002065445296466,
          "validation_metric": 0.00042674446012824774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2070431600004667
        },
        {
          "epoch": 390,
          "updates": 390,
          "samples_seen": 399360,
          "online_loss": 0.004966489039361477,
          "train_loss": 0.004937594756484032,
          "validation_loss": 0.004211872816085815,
          "train_metric": 0.0004972963943146169,
          "validation_metric": 0.00042420433601364493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.220998218001114
        },
        {
          "epoch": 391,
          "updates": 391,
          "samples_seen": 400384,
          "online_loss": 0.004937594756484032,
          "train_loss": 0.004908895585685968,
          "validation_loss": 0.004186804406344891,
          "train_metric": 0.000494405918288976,
          "validation_metric": 0.0004216795787215233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.238852771000893
        },
        {
          "epoch": 392,
          "updates": 392,
          "samples_seen": 401408,
          "online_loss": 0.004908895120024681,
          "train_loss": 0.0048803891986608505,
          "validation_loss": 0.004161865916103125,
          "train_metric": 0.000491534941829741,
          "validation_metric": 0.00041916791815310717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.247962492001534
        },
        {
          "epoch": 393,
          "updates": 393,
          "samples_seen": 402432,
          "online_loss": 0.0048803891986608505,
          "train_loss": 0.004852072335779667,
          "validation_loss": 0.004137079231441021,
          "train_metric": 0.0004886830574832857,
          "validation_metric": 0.00041667153709568083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.256101391001721
        },
        {
          "epoch": 394,
          "updates": 394,
          "samples_seen": 403456,
          "online_loss": 0.004852072801440954,
          "train_loss": 0.004823945462703705,
          "validation_loss": 0.004112457390874624,
          "train_metric": 0.0004858501488342881,
          "validation_metric": 0.000414191628806293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2652273070016236
        },
        {
          "epoch": 395,
          "updates": 395,
          "samples_seen": 404480,
          "online_loss": 0.004823945462703705,
          "train_loss": 0.004796016030013561,
          "validation_loss": 0.004087962210178375,
          "train_metric": 0.000483037147205323,
          "validation_metric": 0.00041172458440996706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2734174730012455
        },
        {
          "epoch": 396,
          "updates": 396,
          "samples_seen": 405504,
          "online_loss": 0.004796016030013561,
          "train_loss": 0.00476826773956418,
          "validation_loss": 0.0040636202320456505,
          "train_metric": 0.0004802424518857151,
          "validation_metric": 0.00040927293593995273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.290550928000812
        },
        {
          "epoch": 397,
          "updates": 397,
          "samples_seen": 406528,
          "online_loss": 0.004768267273902893,
          "train_loss": 0.004740716889500618,
          "validation_loss": 0.004039436113089323,
          "train_metric": 0.0004774676344823092,
          "validation_metric": 0.0004068371490575373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5087917530017876
        },
        {
          "epoch": 398,
          "updates": 398,
          "samples_seen": 407552,
          "online_loss": 0.004740716889500618,
          "train_loss": 0.004713353235274553,
          "validation_loss": 0.004015389829874039,
          "train_metric": 0.00047471170546486974,
          "validation_metric": 0.0004044154193252325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5179865610007255
        },
        {
          "epoch": 399,
          "updates": 399,
          "samples_seen": 408576,
          "online_loss": 0.004713353235274553,
          "train_loss": 0.004686175845563412,
          "validation_loss": 0.003991493955254555,
          "train_metric": 0.00047197440289892256,
          "validation_metric": 0.0004020086780656129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5271823940001923
        },
        {
          "epoch": 400,
          "updates": 400,
          "samples_seen": 409600,
          "online_loss": 0.004686175379902124,
          "train_loss": 0.004659190773963928,
          "validation_loss": 0.00396775035187602,
          "train_metric": 0.00046925662900321186,
          "validation_metric": 0.0003996173036284745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.535719619001611
        },
        {
          "epoch": 401,
          "updates": 401,
          "samples_seen": 410624,
          "online_loss": 0.004659190773963928,
          "train_loss": 0.004632391035556793,
          "validation_loss": 0.003944151569157839,
          "train_metric": 0.00046655748155899346,
          "validation_metric": 0.0003972405393142253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.544823141000961
        },
        {
          "epoch": 402,
          "updates": 402,
          "samples_seen": 411648,
          "online_loss": 0.0046323915012180805,
          "train_loss": 0.004605777096003294,
          "validation_loss": 0.003920692950487137,
          "train_metric": 0.00046387698967009783,
          "validation_metric": 0.0003948778030462563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5534599750008056
        },
        {
          "epoch": 403,
          "updates": 403,
          "samples_seen": 412672,
          "online_loss": 0.004605777096003294,
          "train_loss": 0.004579352214932442,
          "validation_loss": 0.0038973838090896606,
          "train_metric": 0.00046121561899781227,
          "validation_metric": 0.0003925302007701248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5617740900015633
        },
        {
          "epoch": 404,
          "updates": 404,
          "samples_seen": 413696,
          "online_loss": 0.004579352680593729,
          "train_loss": 0.004553108476102352,
          "validation_loss": 0.003874229034408927,
          "train_metric": 0.0004585724091157317,
          "validation_metric": 0.0003901981399394572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5700459140007297
        },
        {
          "epoch": 405,
          "updates": 405,
          "samples_seen": 414720,
          "online_loss": 0.0045531089417636395,
          "train_loss": 0.004527058452367783,
          "validation_loss": 0.003851222340017557,
          "train_metric": 0.00045594872790388763,
          "validation_metric": 0.00038788100937381387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.578401931001281
        },
        {
          "epoch": 406,
          "updates": 406,
          "samples_seen": 415744,
          "online_loss": 0.004527058452367783,
          "train_loss": 0.0045011877082288265,
          "validation_loss": 0.003828349057585001,
          "train_metric": 0.0004533431783784181,
          "validation_metric": 0.00038557720836251974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5868663340006606
        },
        {
          "epoch": 407,
          "updates": 407,
          "samples_seen": 416768,
          "online_loss": 0.004501188173890114,
          "train_loss": 0.004475498106330633,
          "validation_loss": 0.0038056313060224056,
          "train_metric": 0.0004507557605393231,
          "validation_metric": 0.0003832892398349941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.595631680000224
        },
        {
          "epoch": 408,
          "updates": 408,
          "samples_seen": 417792,
          "online_loss": 0.004475497640669346,
          "train_loss": 0.004450000822544098,
          "validation_loss": 0.0037830714136362076,
          "train_metric": 0.00044818781316280365,
          "validation_metric": 0.0003810171037912369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.604213343000083
        },
        {
          "epoch": 409,
          "updates": 409,
          "samples_seen": 418816,
          "online_loss": 0.004450000822544098,
          "train_loss": 0.004424684680998325,
          "validation_loss": 0.0037606442347168922,
          "train_metric": 0.00044563799747265875,
          "validation_metric": 0.0003787583264056593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.612483864000751
        },
        {
          "epoch": 410,
          "updates": 410,
          "samples_seen": 419840,
          "online_loss": 0.004424684215337038,
          "train_loss": 0.004399546887725592,
          "validation_loss": 0.0037383688613772392,
          "train_metric": 0.00044310628436505795,
          "validation_metric": 0.000376514857634902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6208614880015375
        },
        {
          "epoch": 411,
          "updates": 411,
          "samples_seen": 420864,
          "online_loss": 0.004399547353386879,
          "train_loss": 0.004374592565000057,
          "validation_loss": 0.0037162385415285826,
          "train_metric": 0.00044059293577447534,
          "validation_metric": 0.00037428588257171214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.629099630001292
        },
        {
          "epoch": 412,
          "updates": 412,
          "samples_seen": 421888,
          "online_loss": 0.0043745930306613445,
          "train_loss": 0.004349817056208849,
          "validation_loss": 0.0036942502483725548,
          "train_metric": 0.0004380976315587759,
          "validation_metric": 0.0003720713430084288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6376214620013343
        },
        {
          "epoch": 413,
          "updates": 413,
          "samples_seen": 422912,
          "online_loss": 0.004349817521870136,
          "train_loss": 0.004325219430029392,
          "validation_loss": 0.00367241189815104,
          "train_metric": 0.00043562031351029873,
          "validation_metric": 0.0003698718792293221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6461357490006776
        },
        {
          "epoch": 414,
          "updates": 414,
          "samples_seen": 423936,
          "online_loss": 0.004325220361351967,
          "train_loss": 0.004300805274397135,
          "validation_loss": 0.003650721162557602,
          "train_metric": 0.00043316135997883976,
          "validation_metric": 0.0003676872292999178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.65491536600166
        },
        {
          "epoch": 415,
          "updates": 415,
          "samples_seen": 424960,
          "online_loss": 0.004300805740058422,
          "train_loss": 0.004276570864021778,
          "validation_loss": 0.003629175014793873,
          "train_metric": 0.0004307205672375858,
          "validation_metric": 0.00036551718949340284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.663301232001686
        },
        {
          "epoch": 416,
          "updates": 416,
          "samples_seen": 425984,
          "online_loss": 0.004276570864021778,
          "train_loss": 0.004252515733242035,
          "validation_loss": 0.003607773222029209,
          "train_metric": 0.00042829784797504544,
          "validation_metric": 0.0003633617307059467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6719294160011486
        },
        {
          "epoch": 417,
          "updates": 417,
          "samples_seen": 427008,
          "online_loss": 0.004252515733242035,
          "train_loss": 0.004228633362799883,
          "validation_loss": 0.0035865209065377712,
          "train_metric": 0.00042589253280311823,
          "validation_metric": 0.0003612212312873453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.687429970001176
        },
        {
          "epoch": 418,
          "updates": 418,
          "samples_seen": 428032,
          "online_loss": 0.00422863382846117,
          "train_loss": 0.004204926081001759,
          "validation_loss": 0.0035654082894325256,
          "train_metric": 0.0004235047090332955,
          "validation_metric": 0.0003590949054341763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.709324187000675
        },
        {
          "epoch": 419,
          "updates": 419,
          "samples_seen": 429056,
          "online_loss": 0.004204926081001759,
          "train_loss": 0.004181400872766972,
          "validation_loss": 0.0035444435197860003,
          "train_metric": 0.0004211353079881519,
          "validation_metric": 0.00035698339343070984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.728820024000015
        },
        {
          "epoch": 420,
          "updates": 420,
          "samples_seen": 430080,
          "online_loss": 0.004181400407105684,
          "train_loss": 0.004158043302595615,
          "validation_loss": 0.0035236200783401728,
          "train_metric": 0.0004187828744761646,
          "validation_metric": 0.00035488608409650624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7506669180002064
        },
        {
          "epoch": 421,
          "updates": 421,
          "samples_seen": 431104,
          "online_loss": 0.004158042371273041,
          "train_loss": 0.004134862218052149,
          "validation_loss": 0.003502938896417618,
          "train_metric": 0.0004164481651969254,
          "validation_metric": 0.00035280315205454826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7697068940015015
        },
        {
          "epoch": 422,
          "updates": 422,
          "samples_seen": 432128,
          "online_loss": 0.004134862218052149,
          "train_loss": 0.004111858084797859,
          "validation_loss": 0.003482404863461852,
          "train_metric": 0.00041413132566958666,
          "validation_metric": 0.00035073503386229277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.788424885000495
        },
        {
          "epoch": 423,
          "updates": 423,
          "samples_seen": 433152,
          "online_loss": 0.004111858084797859,
          "train_loss": 0.004089021589607,
          "validation_loss": 0.003462023800238967,
          "train_metric": 0.00041183127905242145,
          "validation_metric": 0.00034868234070017934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8072801070011337
        },
        {
          "epoch": 424,
          "updates": 424,
          "samples_seen": 434176,
          "online_loss": 0.004089022055268288,
          "train_loss": 0.004066355060786009,
          "validation_loss": 0.003441769164055586,
          "train_metric": 0.0004095484036952257,
          "validation_metric": 0.00034664239501580596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8227006060005806
        },
        {
          "epoch": 425,
          "updates": 425,
          "samples_seen": 435200,
          "online_loss": 0.004066355060786009,
          "train_loss": 0.00404386268928647,
          "validation_loss": 0.0034216681960970163,
          "train_metric": 0.0004072830197401345,
          "validation_metric": 0.0003446178452577442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8317997600006493
        },
        {
          "epoch": 426,
          "updates": 426,
          "samples_seen": 436224,
          "online_loss": 0.004043862223625183,
          "train_loss": 0.004021535627543926,
          "validation_loss": 0.003401704365387559,
          "train_metric": 0.0004050343995913863,
          "validation_metric": 0.00034260720713064075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.840462076001131
        },
        {
          "epoch": 427,
          "updates": 427,
          "samples_seen": 437248,
          "online_loss": 0.004021535627543926,
          "train_loss": 0.003999382257461548,
          "validation_loss": 0.0033818846568465233,
          "train_metric": 0.00040280315442942083,
          "validation_metric": 0.0003406110336072743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.849033224001687
        },
        {
          "epoch": 428,
          "updates": 428,
          "samples_seen": 438272,
          "online_loss": 0.003999382257461548,
          "train_loss": 0.003977390006184578,
          "validation_loss": 0.003362201852723956,
          "train_metric": 0.00040058812010101974,
          "validation_metric": 0.00033862865529954433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8572607020014402
        },
        {
          "epoch": 429,
          "updates": 429,
          "samples_seen": 439296,
          "online_loss": 0.003977390006184578,
          "train_loss": 0.003955568186938763,
          "validation_loss": 0.003342672251164913,
          "train_metric": 0.00039839037344790995,
          "validation_metric": 0.000336661672918126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8657407210012025
        },
        {
          "epoch": 430,
          "updates": 430,
          "samples_seen": 440320,
          "online_loss": 0.00395556865260005,
          "train_loss": 0.003933907952159643,
          "validation_loss": 0.0033232741989195347,
          "train_metric": 0.00039620883762836456,
          "validation_metric": 0.0003347080491948873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.874196901000687
        },
        {
          "epoch": 431,
          "updates": 431,
          "samples_seen": 441344,
          "online_loss": 0.0039339084178209305,
          "train_loss": 0.003912412095814943,
          "validation_loss": 0.0033040139824151993,
          "train_metric": 0.00039404380368068814,
          "validation_metric": 0.00033276822068728507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8828404510004475
        },
        {
          "epoch": 432,
          "updates": 432,
          "samples_seen": 442368,
          "online_loss": 0.003912411630153656,
          "train_loss": 0.0038910778239369392,
          "validation_loss": 0.0032848971895873547,
          "train_metric": 0.0003918951260857284,
          "validation_metric": 0.0003308428276795894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.89153371500106
        },
        {
          "epoch": 433,
          "updates": 433,
          "samples_seen": 443392,
          "online_loss": 0.0038910782895982265,
          "train_loss": 0.0038699086289852858,
          "validation_loss": 0.003265920327976346,
          "train_metric": 0.00038976300857029855,
          "validation_metric": 0.0003289315209258348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9001800720016035
        },
        {
          "epoch": 434,
          "updates": 434,
          "samples_seen": 444416,
          "online_loss": 0.003869908396154642,
          "train_loss": 0.0038488993886858225,
          "validation_loss": 0.0032470799051225185,
          "train_metric": 0.0003876470436807722,
          "validation_metric": 0.00032703400938771665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9085793230005947
        },
        {
          "epoch": 435,
          "updates": 435,
          "samples_seen": 445440,
          "online_loss": 0.0038488993886858225,
          "train_loss": 0.0038280487060546875,
          "validation_loss": 0.003228376852348447,
          "train_metric": 0.0003855470276903361,
          "validation_metric": 0.0003251503221690655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.015925685000184
        },
        {
          "epoch": 436,
          "updates": 436,
          "samples_seen": 446464,
          "online_loss": 0.003828048938885331,
          "train_loss": 0.00380735844373703,
          "validation_loss": 0.0032098223455250263,
          "train_metric": 0.0003834632516372949,
          "validation_metric": 0.0003232815070077777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0248534660004225
        },
        {
          "epoch": 437,
          "updates": 437,
          "samples_seen": 447488,
          "online_loss": 0.0038073586765676737,
          "train_loss": 0.003786821383982897,
          "validation_loss": 0.003191385418176651,
          "train_metric": 0.00038139475509524345,
          "validation_metric": 0.00032142462441697717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.034028795000268
        },
        {
          "epoch": 438,
          "updates": 438,
          "samples_seen": 448512,
          "online_loss": 0.003786821151152253,
          "train_loss": 0.00376644148491323,
          "validation_loss": 0.003173090983182192,
          "train_metric": 0.00037934223655611277,
          "validation_metric": 0.0003195820900145918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.042610824000803
        },
        {
          "epoch": 439,
          "updates": 439,
          "samples_seen": 449536,
          "online_loss": 0.0037664417177438736,
          "train_loss": 0.003746220376342535,
          "validation_loss": 0.0031549381092190742,
          "train_metric": 0.000377305579604581,
          "validation_metric": 0.00031775381648913026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.051338951001526
        },
        {
          "epoch": 440,
          "updates": 440,
          "samples_seen": 450560,
          "online_loss": 0.0037462201435118914,
          "train_loss": 0.0037261468823999166,
          "validation_loss": 0.0031369184143841267,
          "train_metric": 0.00037528385291807353,
          "validation_metric": 0.0003159389307256788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.060096114000771
        },
        {
          "epoch": 441,
          "updates": 441,
          "samples_seen": 451584,
          "online_loss": 0.003726146649569273,
          "train_loss": 0.0037062300834804773,
          "validation_loss": 0.0031190321315079927,
          "train_metric": 0.00037327795871533453,
          "validation_metric": 0.00031413749093189836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0695855370013305
        },
        {
          "epoch": 442,
          "updates": 442,
          "samples_seen": 452608,
          "online_loss": 0.003706230316311121,
          "train_loss": 0.0036864611320197582,
          "validation_loss": 0.00310128228738904,
          "train_metric": 0.000371286878362298,
          "validation_metric": 0.00031234981724992394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.078750460001174
        },
        {
          "epoch": 443,
          "updates": 443,
          "samples_seen": 453632,
          "online_loss": 0.003686461364850402,
          "train_loss": 0.003666844917461276,
          "validation_loss": 0.0030836528167128563,
          "train_metric": 0.00036931116483174264,
          "validation_metric": 0.0003105741925537586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.087190780001038
        },
        {
          "epoch": 444,
          "updates": 444,
          "samples_seen": 454656,
          "online_loss": 0.0036668446846306324,
          "train_loss": 0.003647375851869583,
          "validation_loss": 0.0030661700293421745,
          "train_metric": 0.00036735032335855067,
          "validation_metric": 0.00030881341081112623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.095826736000163
        },
        {
          "epoch": 445,
          "updates": 445,
          "samples_seen": 455680,
          "online_loss": 0.0036473756190389395,
          "train_loss": 0.0036280553322285414,
          "validation_loss": 0.0030488118063658476,
          "train_metric": 0.00036540444125421345,
          "validation_metric": 0.00030706514371559024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.104240941000171
        },
        {
          "epoch": 446,
          "updates": 446,
          "samples_seen": 456704,
          "online_loss": 0.003628055565059185,
          "train_loss": 0.003608875907957554,
          "validation_loss": 0.0030315774492919445,
          "train_metric": 0.0003634727909229696,
          "validation_metric": 0.0003053294203709811,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.112232598001356
        },
        {
          "epoch": 447,
          "updates": 447,
          "samples_seen": 457728,
          "online_loss": 0.003608875907957554,
          "train_loss": 0.00358985154889524,
          "validation_loss": 0.003014488611370325,
          "train_metric": 0.00036155665293335915,
          "validation_metric": 0.00030360821983776987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.120679364001262
        },
        {
          "epoch": 448,
          "updates": 448,
          "samples_seen": 458752,
          "online_loss": 0.003589851316064596,
          "train_loss": 0.003570959670469165,
          "validation_loss": 0.002997511299327016,
          "train_metric": 0.0003596540482249111,
          "validation_metric": 0.00030189836979843676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.129491724001127
        },
        {
          "epoch": 449,
          "updates": 449,
          "samples_seen": 459776,
          "online_loss": 0.003570960136130452,
          "train_loss": 0.0035522179678082466,
          "validation_loss": 0.0029806778766214848,
          "train_metric": 0.00035776643198914826,
          "validation_metric": 0.0003002029552590102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.138443122001263
        },
        {
          "epoch": 450,
          "updates": 450,
          "samples_seen": 460800,
          "online_loss": 0.0035522182006388903,
          "train_loss": 0.003533612471073866,
          "validation_loss": 0.0029639597050845623,
          "train_metric": 0.00035589246544986963,
          "validation_metric": 0.0002985191240441054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.147410429000956
        },
        {
          "epoch": 451,
          "updates": 451,
          "samples_seen": 461824,
          "online_loss": 0.003533612471073866,
          "train_loss": 0.0035151473712176085,
          "validation_loss": 0.0029473777394741774,
          "train_metric": 0.0003540327597875148,
          "validation_metric": 0.0002968490880448371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.156026892000227
        },
        {
          "epoch": 452,
          "updates": 452,
          "samples_seen": 462848,
          "online_loss": 0.003515147138386965,
          "train_loss": 0.003496821504086256,
          "validation_loss": 0.002930917777121067,
          "train_metric": 0.0003521870821714401,
          "validation_metric": 0.00029519127565436065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.165697935000935
        },
        {
          "epoch": 453,
          "updates": 453,
          "samples_seen": 463872,
          "online_loss": 0.0034968217369168997,
          "train_loss": 0.0034786330070346594,
          "validation_loss": 0.0029145819135010242,
          "train_metric": 0.0003503551706671715,
          "validation_metric": 0.0002935460361186415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.17464015600126
        },
        {
          "epoch": 454,
          "updates": 454,
          "samples_seen": 464896,
          "online_loss": 0.0034786330070346594,
          "train_loss": 0.0034605753608047962,
          "validation_loss": 0.0028983678203076124,
          "train_metric": 0.0003485365305095911,
          "validation_metric": 0.0002919129910878837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.1841043300009915
        },
        {
          "epoch": 455,
          "updates": 455,
          "samples_seen": 465920,
          "online_loss": 0.0034605758264660835,
          "train_loss": 0.0034426581114530563,
          "validation_loss": 0.0028822741005569696,
          "train_metric": 0.00034673194750212133,
          "validation_metric": 0.0002902920823544264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.192848516000595
        },
        {
          "epoch": 456,
          "updates": 456,
          "samples_seen": 466944,
          "online_loss": 0.0034426585771143436,
          "train_loss": 0.0034248733427375555,
          "validation_loss": 0.002866310765966773,
          "train_metric": 0.00034494069404900074,
          "validation_metric": 0.00028868435765616596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.201612423001279
        },
        {
          "epoch": 457,
          "updates": 457,
          "samples_seen": 467968,
          "online_loss": 0.003424873575568199,
          "train_loss": 0.0034072156995534897,
          "validation_loss": 0.0028504584915935993,
          "train_metric": 0.000343162304488942,
          "validation_metric": 0.00028708775062114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2099787760016625
        },
        {
          "epoch": 458,
          "updates": 458,
          "samples_seen": 468992,
          "online_loss": 0.0034072156995534897,
          "train_loss": 0.003389692632481456,
          "validation_loss": 0.0028347375337034464,
          "train_metric": 0.0003413973608985543,
          "validation_metric": 0.00028550438582897186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.21920719100126
        },
        {
          "epoch": 459,
          "updates": 459,
          "samples_seen": 470016,
          "online_loss": 0.0033896921668201685,
          "train_loss": 0.003372299950569868,
          "validation_loss": 0.0028191301971673965,
          "train_metric": 0.0003396456886548549,
          "validation_metric": 0.0002839324588421732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.229460135000409
        },
        {
          "epoch": 460,
          "updates": 460,
          "samples_seen": 471040,
          "online_loss": 0.003372299950569868,
          "train_loss": 0.0033550336956977844,
          "validation_loss": 0.002803641138598323,
          "train_metric": 0.00033790667657740414,
          "validation_metric": 0.0002823724935296923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2380798640006105
        },
        {
          "epoch": 461,
          "updates": 461,
          "samples_seen": 472064,
          "online_loss": 0.0033550341613590717,
          "train_loss": 0.0033378899097442627,
          "validation_loss": 0.002788279438391328,
          "train_metric": 0.00033618006273172796,
          "validation_metric": 0.0002808252756949514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2469857260002755
        },
        {
          "epoch": 462,
          "updates": 462,
          "samples_seen": 473088,
          "online_loss": 0.0033378901425749063,
          "train_loss": 0.0033208788372576237,
          "validation_loss": 0.002773015759885311,
          "train_metric": 0.00033446672023274004,
          "validation_metric": 0.00027928801137022674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.255492257001606
        },
        {
          "epoch": 463,
          "updates": 463,
          "samples_seen": 474112,
          "online_loss": 0.00332087860442698,
          "train_loss": 0.00330398790538311,
          "validation_loss": 0.0027578857261687517,
          "train_metric": 0.00033276560134254396,
          "validation_metric": 0.00027776413480751216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.26396642400141
        },
        {
          "epoch": 464,
          "updates": 464,
          "samples_seen": 475136,
          "online_loss": 0.00330398790538311,
          "train_loss": 0.003287220373749733,
          "validation_loss": 0.002742853481322527,
          "train_metric": 0.00033107682247646153,
          "validation_metric": 0.0002762501535471529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.272739959000319
        },
        {
          "epoch": 465,
          "updates": 465,
          "samples_seen": 476160,
          "online_loss": 0.0032872206065803766,
          "train_loss": 0.003270575311034918,
          "validation_loss": 0.0027279492933303118,
          "train_metric": 0.0003294002963230014,
          "validation_metric": 0.000274749007076025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.281418234000739
        },
        {
          "epoch": 466,
          "updates": 466,
          "samples_seen": 477184,
          "online_loss": 0.0032705748453736305,
          "train_loss": 0.0032540466636419296,
          "validation_loss": 0.0027131508104503155,
          "train_metric": 0.00032773567363619804,
          "validation_metric": 0.00027325862902216613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2900743930003955
        },
        {
          "epoch": 467,
          "updates": 467,
          "samples_seen": 478208,
          "online_loss": 0.0032540466636419296,
          "train_loss": 0.0032376390881836414,
          "validation_loss": 0.00269846199080348,
          "train_metric": 0.00032608318724669516,
          "validation_metric": 0.000271779194008559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.298846114001208
        },
        {
          "epoch": 468,
          "updates": 468,
          "samples_seen": 479232,
          "online_loss": 0.003237639321014285,
          "train_loss": 0.00322134792804718,
          "validation_loss": 0.00268388818949461,
          "train_metric": 0.0003244424005970359,
          "validation_metric": 0.00027031140052713454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.307281843000965
        },
        {
          "epoch": 469,
          "updates": 469,
          "samples_seen": 480256,
          "online_loss": 0.0032213483937084675,
          "train_loss": 0.003205169690772891,
          "validation_loss": 0.002669419627636671,
          "train_metric": 0.00032281296444125473,
          "validation_metric": 0.0002688541717361659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.316270104000068
        },
        {
          "epoch": 470,
          "updates": 470,
          "samples_seen": 481280,
          "online_loss": 0.0032051701564341784,
          "train_loss": 0.0031891160178929567,
          "validation_loss": 0.00265506561845541,
          "train_metric": 0.0003211961011402309,
          "validation_metric": 0.00026740849716588855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.324741781001649
        },
        {
          "epoch": 471,
          "updates": 471,
          "samples_seen": 482304,
          "online_loss": 0.0031891160178929567,
          "train_loss": 0.003173172939568758,
          "validation_loss": 0.002640817780047655,
          "train_metric": 0.00031959032639861107,
          "validation_metric": 0.00026597350370138884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.333500870001444
        },
        {
          "epoch": 472,
          "updates": 472,
          "samples_seen": 483328,
          "online_loss": 0.003173172939568758,
          "train_loss": 0.00315733696334064,
          "validation_loss": 0.002626671688631177,
          "train_metric": 0.00031799543648958206,
          "validation_metric": 0.00026454872568137944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.44128553200062
        },
        {
          "epoch": 473,
          "updates": 473,
          "samples_seen": 484352,
          "online_loss": 0.0031573371961712837,
          "train_loss": 0.003141620196402073,
          "validation_loss": 0.0026126361917704344,
          "train_metric": 0.0003164124791510403,
          "validation_metric": 0.0002631351235322654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.450404650000564
        },
        {
          "epoch": 474,
          "updates": 474,
          "samples_seen": 485376,
          "online_loss": 0.003141620196402073,
          "train_loss": 0.0031260063406080008,
          "validation_loss": 0.002598700113594532,
          "train_metric": 0.0003148399118799716,
          "validation_metric": 0.00026173150399699807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.458777388001181
        },
        {
          "epoch": 475,
          "updates": 475,
          "samples_seen": 486400,
          "online_loss": 0.0031260063406080008,
          "train_loss": 0.003110511228442192,
          "validation_loss": 0.00258487556129694,
          "train_metric": 0.00031327930628322065,
          "validation_metric": 0.00026033917674794793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.467229733001659
        },
        {
          "epoch": 476,
          "updates": 476,
          "samples_seen": 487424,
          "online_loss": 0.0031105109956115484,
          "train_loss": 0.0030951194930821657,
          "validation_loss": 0.002571149030700326,
          "train_metric": 0.00031172906165011227,
          "validation_metric": 0.0002589567156974226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.475631098001031
        },
        {
          "epoch": 477,
          "updates": 477,
          "samples_seen": 488448,
          "online_loss": 0.0030951190274208784,
          "train_loss": 0.003079830203205347,
          "validation_loss": 0.002557522151619196,
          "train_metric": 0.0003101892361883074,
          "validation_metric": 0.00025758426636457443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.48410708800111
        },
        {
          "epoch": 478,
          "updates": 478,
          "samples_seen": 489472,
          "online_loss": 0.0030798304360359907,
          "train_loss": 0.003064648946747184,
          "validation_loss": 0.002544001443311572,
          "train_metric": 0.00030866017914377153,
          "validation_metric": 0.0002562224690336734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.492885592000675
        },
        {
          "epoch": 479,
          "updates": 479,
          "samples_seen": 490496,
          "online_loss": 0.003064648946747184,
          "train_loss": 0.003049573628231883,
          "validation_loss": 0.0025305773597210646,
          "train_metric": 0.0003071418614126742,
          "validation_metric": 0.0002548704796936363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.501436450000256
        },
        {
          "epoch": 480,
          "updates": 480,
          "samples_seen": 491520,
          "online_loss": 0.0030495733954012394,
          "train_loss": 0.003034604247659445,
          "validation_loss": 0.0025172512978315353,
          "train_metric": 0.00030563422478735447,
          "validation_metric": 0.0002535282983444631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.510043910000604
        },
        {
          "epoch": 481,
          "updates": 481,
          "samples_seen": 492544,
          "online_loss": 0.003034604247659445,
          "train_loss": 0.0030197324231266975,
          "validation_loss": 0.0025040293112397194,
          "train_metric": 0.0003041363670490682,
          "validation_metric": 0.0002521966234780848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.518742757001746
        },
        {
          "epoch": 482,
          "updates": 482,
          "samples_seen": 493568,
          "online_loss": 0.0030197324231266975,
          "train_loss": 0.003004959085956216,
          "validation_loss": 0.002490884391590953,
          "train_metric": 0.0003026484337169677,
          "validation_metric": 0.0002508727484382689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.5275490840012935
        },
        {
          "epoch": 483,
          "updates": 483,
          "samples_seen": 494592,
          "online_loss": 0.003004958853125572,
          "train_loss": 0.002990291453897953,
          "validation_loss": 0.002477861475199461,
          "train_metric": 0.00030117121059447527,
          "validation_metric": 0.00024956115521490574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.5363802120009495
        },
        {
          "epoch": 484,
          "updates": 484,
          "samples_seen": 495616,
          "online_loss": 0.002990291453897953,
          "train_loss": 0.0029757204465568066,
          "validation_loss": 0.002464905846863985,
          "train_metric": 0.0002997037081513554,
          "validation_metric": 0.0002482562558725476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.544980918000874
        },
        {
          "epoch": 485,
          "updates": 485,
          "samples_seen": 496640,
          "online_loss": 0.002975720912218094,
          "train_loss": 0.0029612474609166384,
          "validation_loss": 0.0024520724546164274,
          "train_metric": 0.0002982460136990994,
          "validation_metric": 0.0002469637547619641,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.553451870000572
        },
        {
          "epoch": 486,
          "updates": 486,
          "samples_seen": 497664,
          "online_loss": 0.0029612474609166384,
          "train_loss": 0.0029468690045177937,
          "validation_loss": 0.0024392956402152777,
          "train_metric": 0.00029679786530323327,
          "validation_metric": 0.0002456769288983196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.562808086000587
        },
        {
          "epoch": 487,
          "updates": 487,
          "samples_seen": 498688,
          "online_loss": 0.0029468685388565063,
          "train_loss": 0.0029325929936021566,
          "validation_loss": 0.002426659455522895,
          "train_metric": 0.0002953600196633488,
          "validation_metric": 0.0002444042474962771,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.571400625000024
        },
        {
          "epoch": 488,
          "updates": 488,
          "samples_seen": 499712,
          "online_loss": 0.002932592760771513,
          "train_loss": 0.0029184080194681883,
          "validation_loss": 0.002414063084870577,
          "train_metric": 0.0002939313999377191,
          "validation_metric": 0.0002431355678709224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.5804068420002295
        },
        {
          "epoch": 489,
          "updates": 489,
          "samples_seen": 500736,
          "online_loss": 0.0029184080194681883,
          "train_loss": 0.002904314547777176,
          "validation_loss": 0.0024016029201447964,
          "train_metric": 0.0002925119479186833,
          "validation_metric": 0.00024188066890928894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.589266897000925
        },
        {
          "epoch": 490,
          "updates": 490,
          "samples_seen": 501760,
          "online_loss": 0.002904315013438463,
          "train_loss": 0.0028903139755129814,
          "validation_loss": 0.002389180473983288,
          "train_metric": 0.00029110180912539363,
          "validation_metric": 0.00024062949523795396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.598182908001036
        },
        {
          "epoch": 491,
          "updates": 491,
          "samples_seen": 502784,
          "online_loss": 0.002890313509851694,
          "train_loss": 0.002876406069844961,
          "validation_loss": 0.0023769112303853035,
          "train_metric": 0.00028970109997317195,
          "validation_metric": 0.00023939376114867628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.606567407001421
        },
        {
          "epoch": 492,
          "updates": 492,
          "samples_seen": 503808,
          "online_loss": 0.002876406302675605,
          "train_loss": 0.0028625933919101954,
          "validation_loss": 0.0023646445479243994,
          "train_metric": 0.0002883099368773401,
          "validation_metric": 0.00023815836175344884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.6154413590011245
        },
        {
          "epoch": 493,
          "updates": 493,
          "samples_seen": 504832,
          "online_loss": 0.002862593624740839,
          "train_loss": 0.002848860342055559,
          "validation_loss": 0.0023525631986558437,
          "train_metric": 0.00028692680643871427,
          "validation_metric": 0.0002369415306020528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.623907502000293
        },
        {
          "epoch": 494,
          "updates": 494,
          "samples_seen": 505856,
          "online_loss": 0.0028488601092249155,
          "train_loss": 0.002835220890119672,
          "validation_loss": 0.002340457635000348,
          "train_metric": 0.00028555310564115644,
          "validation_metric": 0.00023572231293655932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.6328964730000735
        },
        {
          "epoch": 495,
          "updates": 495,
          "samples_seen": 506880,
          "online_loss": 0.002835220890119672,
          "train_loss": 0.002821672707796097,
          "validation_loss": 0.00232856092043221,
          "train_metric": 0.0002841885725501925,
          "validation_metric": 0.00023452413734048605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.641450895000162
        },
        {
          "epoch": 496,
          "updates": 496,
          "samples_seen": 507904,
          "online_loss": 0.002821672707796097,
          "train_loss": 0.002808202989399433,
          "validation_loss": 0.0023165903985500336,
          "train_metric": 0.0002828319265972823,
          "validation_metric": 0.00023331846750807017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.650740103001226
        },
        {
          "epoch": 497,
          "updates": 497,
          "samples_seen": 508928,
          "online_loss": 0.002808202989399433,
          "train_loss": 0.0027948261704295874,
          "validation_loss": 0.002304901834577322,
          "train_metric": 0.0002814846229739487,
          "validation_metric": 0.00023214129032567143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.6596657630016125
        },
        {
          "epoch": 498,
          "updates": 498,
          "samples_seen": 509952,
          "online_loss": 0.0027948261704295874,
          "train_loss": 0.0027815322391688824,
          "validation_loss": 0.0022930277045816183,
          "train_metric": 0.00028014573035761714,
          "validation_metric": 0.00023094532662071288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.668497715001649
        },
        {
          "epoch": 499,
          "updates": 499,
          "samples_seen": 510976,
          "online_loss": 0.0027815322391688824,
          "train_loss": 0.0027683195658028126,
          "validation_loss": 0.002281603869050741,
          "train_metric": 0.0002788150159176439,
          "validation_metric": 0.00022979477944318205,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.679426048000096
        },
        {
          "epoch": 500,
          "updates": 500,
          "samples_seen": 512000,
          "online_loss": 0.0027683197986334562,
          "train_loss": 0.0027551904786378145,
          "validation_loss": 0.0022697472013533115,
          "train_metric": 0.00027749271248467267,
          "validation_metric": 0.00022860063472762704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.689476894000109
        },
        {
          "epoch": 501,
          "updates": 501,
          "samples_seen": 513024,
          "online_loss": 0.002755190711468458,
          "train_loss": 0.00274214637465775,
          "validation_loss": 0.0022586905397474766,
          "train_metric": 0.0002761789655778557,
          "validation_metric": 0.00022748703486286104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.70165086400084
        },
        {
          "epoch": 502,
          "updates": 502,
          "samples_seen": 514048,
          "online_loss": 0.0027421461418271065,
          "train_loss": 0.002729182131588459,
          "validation_loss": 0.002246682532131672,
          "train_metric": 0.00027487322222441435,
          "validation_metric": 0.0002262776397401467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.7132129940000596
        },
        {
          "epoch": 503,
          "updates": 503,
          "samples_seen": 515072,
          "online_loss": 0.002729182131588459,
          "train_loss": 0.0027163021732121706,
          "validation_loss": 0.0022362403105944395,
          "train_metric": 0.00027357597718946636,
          "validation_metric": 0.00022522592917084694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.723668604001432
        },
        {
          "epoch": 504,
          "updates": 504,
          "samples_seen": 516096,
          "online_loss": 0.0027163017075508833,
          "train_loss": 0.0027035081293433905,
          "validation_loss": 0.002223687246441841,
          "train_metric": 0.00027228740509599447,
          "validation_metric": 0.00022396162967197597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.73665265100135
        },
        {
          "epoch": 505,
          "updates": 505,
          "samples_seen": 517120,
          "online_loss": 0.002703507896512747,
          "train_loss": 0.0026907948777079582,
          "validation_loss": 0.002214499283581972,
          "train_metric": 0.0002710069820750505,
          "validation_metric": 0.0002230362588306889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.752258801001517
        },
        {
          "epoch": 506,
          "updates": 506,
          "samples_seen": 518144,
          "online_loss": 0.0026907948777079582,
          "train_loss": 0.002678187098354101,
          "validation_loss": 0.002200389513745904,
          "train_metric": 0.00026973721105605364,
          "validation_metric": 0.0002216151769971475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.762169846000688
        },
        {
          "epoch": 507,
          "updates": 507,
          "samples_seen": 519168,
          "online_loss": 0.002678187098354101,
          "train_loss": 0.0026657076086848974,
          "validation_loss": 0.0021941554732620716,
          "train_metric": 0.0002684803039301187,
          "validation_metric": 0.0002209873346146196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.774873391001165
        },
        {
          "epoch": 508,
          "updates": 508,
          "samples_seen": 520192,
          "online_loss": 0.002665707841515541,
          "train_loss": 0.0026534576900303364,
          "validation_loss": 0.0021758684888482094,
          "train_metric": 0.0002672465343493968,
          "validation_metric": 0.00021914546960033476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.785647507000249
        },
        {
          "epoch": 509,
          "updates": 509,
          "samples_seen": 521216,
          "online_loss": 0.0026534576900303364,
          "train_loss": 0.002641678787767887,
          "validation_loss": 0.002177418675273657,
          "train_metric": 0.0002660602331161499,
          "validation_metric": 0.00021930165530648082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.7934456410002895
        },
        {
          "epoch": 510,
          "updates": 510,
          "samples_seen": 522240,
          "online_loss": 0.002641678787767887,
          "train_loss": 0.00263110245577991,
          "validation_loss": 0.002148180967196822,
          "train_metric": 0.00026499497471377254,
          "validation_metric": 0.00021635690063703805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.8016700690004654
        },
        {
          "epoch": 511,
          "updates": 511,
          "samples_seen": 523264,
          "online_loss": 0.00263110245577991,
          "train_loss": 0.0026236793491989374,
          "validation_loss": 0.0021727930288761854,
          "train_metric": 0.0002642473846208304,
          "validation_metric": 0.0002188357466366142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.809445403001519
        },
        {
          "epoch": 512,
          "updates": 512,
          "samples_seen": 524288,
          "online_loss": 0.0026236793491989374,
          "train_loss": 0.002625631634145975,
          "validation_loss": 0.0021174901630729437,
          "train_metric": 0.00026444398099556565,
          "validation_metric": 0.0002132658555638045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.92223309200017
        },
        {
          "epoch": 513,
          "updates": 513,
          "samples_seen": 525312,
          "online_loss": 0.002625631634145975,
          "train_loss": 0.002651493065059185,
          "validation_loss": 0.002221157308667898,
          "train_metric": 0.0002670486574061215,
          "validation_metric": 0.0002237068401882425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.930139793001217
        },
        {
          "epoch": 514,
          "updates": 514,
          "samples_seen": 526336,
          "online_loss": 0.0026514928322285414,
          "train_loss": 0.002750449813902378,
          "validation_loss": 0.002132639056071639,
          "train_metric": 0.00027701520593836904,
          "validation_metric": 0.00021479159477166831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.93774811700132
        },
        {
          "epoch": 515,
          "updates": 515,
          "samples_seen": 527360,
          "online_loss": 0.0027504493482410908,
          "train_loss": 0.002949522342532873,
          "validation_loss": 0.0024948955979198217,
          "train_metric": 0.00029706506757065654,
          "validation_metric": 0.0002512767678126693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.945444724000481
        },
        {
          "epoch": 516,
          "updates": 516,
          "samples_seen": 528384,
          "online_loss": 0.002949522342532873,
          "train_loss": 0.0033362347166985273,
          "validation_loss": 0.002401884412392974,
          "train_metric": 0.0003360133559908718,
          "validation_metric": 0.00024190900148823857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.953046127000562
        },
        {
          "epoch": 517,
          "updates": 517,
          "samples_seen": 529408,
          "online_loss": 0.003336234949529171,
          "train_loss": 0.0031529313419014215,
          "validation_loss": 0.002668536501005292,
          "train_metric": 0.0003175517194904387,
          "validation_metric": 0.0002687652304302901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.961105884000062
        },
        {
          "epoch": 518,
          "updates": 518,
          "samples_seen": 530432,
          "online_loss": 0.0031529313419014215,
          "train_loss": 0.002744146389886737,
          "validation_loss": 0.0021086623892188072,
          "train_metric": 0.00027638039318844676,
          "validation_metric": 0.00021237677719909698,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.969588669000586
        },
        {
          "epoch": 519,
          "updates": 519,
          "samples_seen": 531456,
          "online_loss": 0.0027441466227173805,
          "train_loss": 0.0025401473976671696,
          "validation_loss": 0.0020500142127275467,
          "train_metric": 0.00025583431124687195,
          "validation_metric": 0.00020646993652917445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.978894336001758
        },
        {
          "epoch": 520,
          "updates": 520,
          "samples_seen": 532480,
          "online_loss": 0.002540147164836526,
          "train_loss": 0.0028348038904368877,
          "validation_loss": 0.0023872745223343372,
          "train_metric": 0.00028551105060614645,
          "validation_metric": 0.00024043757002800703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.986568731001171
        },
        {
          "epoch": 521,
          "updates": 521,
          "samples_seen": 533504,
          "online_loss": 0.0028348041232675314,
          "train_loss": 0.002977511379867792,
          "validation_loss": 0.0021864953450858593,
          "train_metric": 0.0002998840354848653,
          "validation_metric": 0.00022021580662112683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.99440585100092
        },
        {
          "epoch": 522,
          "updates": 522,
          "samples_seen": 534528,
          "online_loss": 0.002977510914206505,
          "train_loss": 0.002569155767560005,
          "validation_loss": 0.0021407657768577337,
          "train_metric": 0.0002587559283711016,
          "validation_metric": 0.00021561005269177258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.001773487001628
        },
        {
          "epoch": 523,
          "updates": 523,
          "samples_seen": 535552,
          "online_loss": 0.002569155301898718,
          "train_loss": 0.0025789067149162292,
          "validation_loss": 0.0021591216791421175,
          "train_metric": 0.0002597380371298641,
          "validation_metric": 0.00021745884441770613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.009274327001549
        },
        {
          "epoch": 524,
          "updates": 524,
          "samples_seen": 536576,
          "online_loss": 0.002578906249254942,
          "train_loss": 0.002838701708242297,
          "validation_loss": 0.0021286034025251865,
          "train_metric": 0.0002859036612790078,
          "validation_metric": 0.00021438515977934003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.016731221001464
        },
        {
          "epoch": 525,
          "updates": 525,
          "samples_seen": 537600,
          "online_loss": 0.002838701941072941,
          "train_loss": 0.0026048170402646065,
          "validation_loss": 0.002196491928771138,
          "train_metric": 0.0002623476320877671,
          "validation_metric": 0.00022122260998003185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.024359931001527
        },
        {
          "epoch": 526,
          "updates": 526,
          "samples_seen": 538624,
          "online_loss": 0.00260481727309525,
          "train_loss": 0.0024743378162384033,
          "validation_loss": 0.002050721785053611,
          "train_metric": 0.00024920626310631633,
          "validation_metric": 0.00020654118270613253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.031898103999993
        },
        {
          "epoch": 527,
          "updates": 527,
          "samples_seen": 539648,
          "online_loss": 0.0024743378162384033,
          "train_loss": 0.0026526376605033875,
          "validation_loss": 0.002036049962043762,
          "train_metric": 0.0002671639376785606,
          "validation_metric": 0.00020506347937043756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.040112850001606
        },
        {
          "epoch": 528,
          "updates": 528,
          "samples_seen": 540672,
          "online_loss": 0.0026526376605033875,
          "train_loss": 0.0025822410825639963,
          "validation_loss": 0.002169980900362134,
          "train_metric": 0.00026007392443716526,
          "validation_metric": 0.0002185525227105245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.0478248670006
        },
        {
          "epoch": 529,
          "updates": 529,
          "samples_seen": 541696,
          "online_loss": 0.00258224131539464,
          "train_loss": 0.0024367093574255705,
          "validation_loss": 0.001981276087462902,
          "train_metric": 0.0002454164787195623,
          "validation_metric": 0.00019954687741119415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.0568990820011095
        },
        {
          "epoch": 530,
          "updates": 530,
          "samples_seen": 542720,
          "online_loss": 0.002436709590256214,
          "train_loss": 0.0025141339283436537,
          "validation_loss": 0.0019694429356604815,
          "train_metric": 0.0002532143844291568,
          "validation_metric": 0.0001983550755539909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.065936634000536
        },
        {
          "epoch": 531,
          "updates": 531,
          "samples_seen": 543744,
          "online_loss": 0.0025141339283436537,
          "train_loss": 0.0025354214012622833,
          "validation_loss": 0.0021350174210965633,
          "train_metric": 0.0002553583472035825,
          "validation_metric": 0.00021503113384824246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.073825398001645
        },
        {
          "epoch": 532,
          "updates": 532,
          "samples_seen": 544768,
          "online_loss": 0.0025354214012622833,
          "train_loss": 0.0024264706298708916,
          "validation_loss": 0.0019514979794621468,
          "train_metric": 0.0002443852135911584,
          "validation_metric": 0.00019654774223454297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.082295447000433
        },
        {
          "epoch": 533,
          "updates": 533,
          "samples_seen": 545792,
          "online_loss": 0.002426470397040248,
          "train_loss": 0.0024283002130687237,
          "validation_loss": 0.0019419798627495766,
          "train_metric": 0.0002445694408379495,
          "validation_metric": 0.00019558909116312861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.090721617001691
        },
        {
          "epoch": 534,
          "updates": 534,
          "samples_seen": 546816,
          "online_loss": 0.0024282997474074364,
          "train_loss": 0.002480710856616497,
          "validation_loss": 0.002080702455714345,
          "train_metric": 0.00024984811898320913,
          "validation_metric": 0.00020956071966793388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.099589154000569
        },
        {
          "epoch": 535,
          "updates": 535,
          "samples_seen": 547840,
          "online_loss": 0.002480710856616497,
          "train_loss": 0.0024210389237850904,
          "validation_loss": 0.0019232995109632611,
          "train_metric": 0.00024383817799389362,
          "validation_metric": 0.00019370770314708352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.108225306001259
        },
        {
          "epoch": 536,
          "updates": 536,
          "samples_seen": 548864,
          "online_loss": 0.0024210389237850904,
          "train_loss": 0.00237838551402092,
          "validation_loss": 0.0019242150010541081,
          "train_metric": 0.00023954226344358176,
          "validation_metric": 0.00019379988953005522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.116868645000068
        },
        {
          "epoch": 537,
          "updates": 537,
          "samples_seen": 549888,
          "online_loss": 0.002378385281190276,
          "train_loss": 0.0024249448906630278,
          "validation_loss": 0.0020221476443111897,
          "train_metric": 0.0002442315744701773,
          "validation_metric": 0.00020366327953524888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.1246829540014005
        },
        {
          "epoch": 538,
          "updates": 538,
          "samples_seen": 550912,
          "online_loss": 0.0024249448906630278,
          "train_loss": 0.0024088118225336075,
          "validation_loss": 0.0019068175461143255,
          "train_metric": 0.00024260670761577785,
          "validation_metric": 0.00019204769341740757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.133283653000035
        },
        {
          "epoch": 539,
          "updates": 539,
          "samples_seen": 551936,
          "online_loss": 0.0024088118225336075,
          "train_loss": 0.002350125228986144,
          "validation_loss": 0.001925799297168851,
          "train_metric": 0.00023669603979215026,
          "validation_metric": 0.00019395945128053427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.140906252001514
        },
        {
          "epoch": 540,
          "updates": 540,
          "samples_seen": 552960,
          "online_loss": 0.0023501254618167877,
          "train_loss": 0.0023723579943180084,
          "validation_loss": 0.001972814090549946,
          "train_metric": 0.00023893518664408475,
          "validation_metric": 0.00019869461539201438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.148670899001445
        },
        {
          "epoch": 541,
          "updates": 541,
          "samples_seen": 553984,
          "online_loss": 0.0023723579943180084,
          "train_loss": 0.0023838281631469727,
          "validation_loss": 0.0018874796805903316,
          "train_metric": 0.0002400904195383191,
          "validation_metric": 0.00019010007963515818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.157205548001002
        },
        {
          "epoch": 542,
          "updates": 542,
          "samples_seen": 555008,
          "online_loss": 0.0023838281631469727,
          "train_loss": 0.0023318459279835224,
          "validation_loss": 0.0019195883069187403,
          "train_metric": 0.00023485497513320297,
          "validation_metric": 0.00019333389354869723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.165526829001465
        },
        {
          "epoch": 543,
          "updates": 543,
          "samples_seen": 556032,
          "online_loss": 0.002331846160814166,
          "train_loss": 0.00232705264352262,
          "validation_loss": 0.0019173931796103716,
          "train_metric": 0.0002343722153455019,
          "validation_metric": 0.0001931128354044631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.173831993000931
        },
        {
          "epoch": 544,
          "updates": 544,
          "samples_seen": 557056,
          "online_loss": 0.002327052876353264,
          "train_loss": 0.0023474995978176594,
          "validation_loss": 0.0018600660841912031,
          "train_metric": 0.0002364315587328747,
          "validation_metric": 0.00018733904289547354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.18252104900057
        },
        {
          "epoch": 545,
          "updates": 545,
          "samples_seen": 558080,
          "online_loss": 0.0023474995978176594,
          "train_loss": 0.0023153789807111025,
          "validation_loss": 0.001912523526698351,
          "train_metric": 0.00023319647880271077,
          "validation_metric": 0.00019262237765360624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.18995186200118
        },
        {
          "epoch": 546,
          "updates": 546,
          "samples_seen": 559104,
          "online_loss": 0.002315378747880459,
          "train_loss": 0.0022921303752809763,
          "validation_loss": 0.0018754542106762528,
          "train_metric": 0.0002308549446752295,
          "validation_metric": 0.00018888889462687075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.201130911000291
        },
        {
          "epoch": 547,
          "updates": 547,
          "samples_seen": 560128,
          "online_loss": 0.0022921301424503326,
          "train_loss": 0.0023057376965880394,
          "validation_loss": 0.0018405112205073237,
          "train_metric": 0.00023222548770718277,
          "validation_metric": 0.00018536954303272069,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.220217676000175
        },
        {
          "epoch": 548,
          "updates": 548,
          "samples_seen": 561152,
          "online_loss": 0.0023057376965880394,
          "train_loss": 0.002295062178745866,
          "validation_loss": 0.0018996961880475283,
          "train_metric": 0.00023115026124287397,
          "validation_metric": 0.00019133042951580137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.228798766000182
        },
        {
          "epoch": 549,
          "updates": 549,
          "samples_seen": 562176,
          "online_loss": 0.002295062178745866,
          "train_loss": 0.002267661737278104,
          "validation_loss": 0.0018382949056103826,
          "train_metric": 0.0002283905923832208,
          "validation_metric": 0.000185146345756948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.336968178000461
        },
        {
          "epoch": 550,
          "updates": 550,
          "samples_seen": 563200,
          "online_loss": 0.0022676619701087475,
          "train_loss": 0.002266320399940014,
          "validation_loss": 0.0018222657963633537,
          "train_metric": 0.00022825549240224063,
          "validation_metric": 0.00018353194172959775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.345807400000922
        },
        {
          "epoch": 551,
          "updates": 551,
          "samples_seen": 564224,
          "online_loss": 0.002266320399940014,
          "train_loss": 0.0022684400901198387,
          "validation_loss": 0.0018730245064944029,
          "train_metric": 0.00022846896899864078,
          "validation_metric": 0.0001886441750684753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.353655444001561
        },
        {
          "epoch": 552,
          "updates": 552,
          "samples_seen": 565248,
          "online_loss": 0.002268439857289195,
          "train_loss": 0.0022491756826639175,
          "validation_loss": 0.0018097651191055775,
          "train_metric": 0.0002265287475893274,
          "validation_metric": 0.00018227291002403945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.3618866030010395
        },
        {
          "epoch": 553,
          "updates": 553,
          "samples_seen": 566272,
          "online_loss": 0.002249175449833274,
          "train_loss": 0.002234745305031538,
          "validation_loss": 0.0018132879631593823,
          "train_metric": 0.00022507537505589426,
          "validation_metric": 0.00018262770026922226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.36991661400134
        },
        {
          "epoch": 554,
          "updates": 554,
          "samples_seen": 567296,
          "online_loss": 0.0022347455378621817,
          "train_loss": 0.002236779546365142,
          "validation_loss": 0.001841682125814259,
          "train_metric": 0.0002252802369184792,
          "validation_metric": 0.00018548750085756183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.377482758000042
        },
        {
          "epoch": 555,
          "updates": 555,
          "samples_seen": 568320,
          "online_loss": 0.002236779313534498,
          "train_loss": 0.002229425823315978,
          "validation_loss": 0.0017872776370495558,
          "train_metric": 0.00022453961719293147,
          "validation_metric": 0.00018000804993789643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.385653919000106
        },
        {
          "epoch": 556,
          "updates": 556,
          "samples_seen": 569344,
          "online_loss": 0.002229425823315978,
          "train_loss": 0.002211475744843483,
          "validation_loss": 0.0018044484313577414,
          "train_metric": 0.00022273173090070486,
          "validation_metric": 0.000181737428647466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.393470957000318
        },
        {
          "epoch": 557,
          "updates": 557,
          "samples_seen": 570368,
          "online_loss": 0.002211475744843483,
          "train_loss": 0.002205221215263009,
          "validation_loss": 0.001801775535568595,
          "train_metric": 0.00022210183669812977,
          "validation_metric": 0.00018146824731957167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.401254036001774
        },
        {
          "epoch": 558,
          "updates": 558,
          "samples_seen": 571392,
          "online_loss": 0.0022052209824323654,
          "train_loss": 0.002203937154263258,
          "validation_loss": 0.0017650050576776266,
          "train_metric": 0.00022197251382749528,
          "validation_metric": 0.00017776484310161322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4097380740004155
        },
        {
          "epoch": 559,
          "updates": 559,
          "samples_seen": 572416,
          "online_loss": 0.0022039373870939016,
          "train_loss": 0.002191675826907158,
          "validation_loss": 0.0017932670889422297,
          "train_metric": 0.00022073753643780947,
          "validation_metric": 0.00018061127047985792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4177161830011755
        },
        {
          "epoch": 560,
          "updates": 560,
          "samples_seen": 573440,
          "online_loss": 0.0021916755940765142,
          "train_loss": 0.002178789582103491,
          "validation_loss": 0.0017675135750323534,
          "train_metric": 0.00021943972387816757,
          "validation_metric": 0.00017801749345380813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.425378087000354
        },
        {
          "epoch": 561,
          "updates": 561,
          "samples_seen": 574464,
          "online_loss": 0.002178789582103491,
          "train_loss": 0.0021747639402747154,
          "validation_loss": 0.0017503969138488173,
          "train_metric": 0.00021903426386415958,
          "validation_metric": 0.0001762935717124492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.433907697000905
        },
        {
          "epoch": 562,
          "updates": 562,
          "samples_seen": 575488,
          "online_loss": 0.0021747639402747154,
          "train_loss": 0.002169450046494603,
          "validation_loss": 0.00177699641790241,
          "train_metric": 0.00021849905897397548,
          "validation_metric": 0.00017897257930599153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.441548929000419
        },
        {
          "epoch": 563,
          "updates": 563,
          "samples_seen": 576512,
          "online_loss": 0.002169450046494603,
          "train_loss": 0.00215759570710361,
          "validation_loss": 0.00173934455960989,
          "train_metric": 0.0002173051325371489,
          "validation_metric": 0.00017518039385322481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.449931033001121
        },
        {
          "epoch": 564,
          "updates": 564,
          "samples_seen": 577536,
          "online_loss": 0.00215759570710361,
          "train_loss": 0.0021478631533682346,
          "validation_loss": 0.0017389185959473252,
          "train_metric": 0.00021632490097545087,
          "validation_metric": 0.00017513750935904682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4585539150011755
        },
        {
          "epoch": 565,
          "updates": 565,
          "samples_seen": 578560,
          "online_loss": 0.0021478626877069473,
          "train_loss": 0.002143339952453971,
          "validation_loss": 0.001749986782670021,
          "train_metric": 0.00021586933871731162,
          "validation_metric": 0.0001762522733770311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.466443733001142
        },
        {
          "epoch": 566,
          "updates": 566,
          "samples_seen": 579584,
          "online_loss": 0.002143339952453971,
          "train_loss": 0.0021365375723689795,
          "validation_loss": 0.001717257546260953,
          "train_metric": 0.00021518424910027534,
          "validation_metric": 0.0001729559007799253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.474845560000176
        },
        {
          "epoch": 567,
          "updates": 567,
          "samples_seen": 580608,
          "online_loss": 0.002136537339538336,
          "train_loss": 0.0021257437765598297,
          "validation_loss": 0.0017307328525930643,
          "train_metric": 0.0002140971046173945,
          "validation_metric": 0.0001743130706017837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.48269168600018
        },
        {
          "epoch": 568,
          "updates": 568,
          "samples_seen": 581632,
          "online_loss": 0.0021257437765598297,
          "train_loss": 0.00211736629717052,
          "validation_loss": 0.0017204565228894353,
          "train_metric": 0.0002132533845724538,
          "validation_metric": 0.00017327808018308133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.490408162000676
        },
        {
          "epoch": 569,
          "updates": 569,
          "samples_seen": 582656,
          "online_loss": 0.00211736629717052,
          "train_loss": 0.00211203470826149,
          "validation_loss": 0.0017005548579618335,
          "train_metric": 0.00021271640434861183,
          "validation_metric": 0.00017127368482761085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.498966369001209
        },
        {
          "epoch": 570,
          "updates": 570,
          "samples_seen": 583680,
          "online_loss": 0.002112035406753421,
          "train_loss": 0.002104678424075246,
          "validation_loss": 0.0017164365854114294,
          "train_metric": 0.00021197550813667476,
          "validation_metric": 0.0001728732167975977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.506836315000328
        },
        {
          "epoch": 571,
          "updates": 571,
          "samples_seen": 584704,
          "online_loss": 0.002104678424075246,
          "train_loss": 0.002095080679282546,
          "validation_loss": 0.0016909092664718628,
          "train_metric": 0.00021100880985613912,
          "validation_metric": 0.00017030218441504985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.515585753000778
        },
        {
          "epoch": 572,
          "updates": 572,
          "samples_seen": 585728,
          "online_loss": 0.0020950804464519024,
          "train_loss": 0.0020871064625680447,
          "validation_loss": 0.0016866405494511127,
          "train_metric": 0.00021020571875851601,
          "validation_metric": 0.00016987223352771252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.524085502000162
        },
        {
          "epoch": 573,
          "updates": 573,
          "samples_seen": 586752,
          "online_loss": 0.0020871064625680447,
          "train_loss": 0.0020810936111956835,
          "validation_loss": 0.0016929981065914035,
          "train_metric": 0.00020960011170245707,
          "validation_metric": 0.0001705125905573368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.531694647001132
        },
        {
          "epoch": 574,
          "updates": 574,
          "samples_seen": 587776,
          "online_loss": 0.0020810936111956835,
          "train_loss": 0.002073944779112935,
          "validation_loss": 0.0016684376168996096,
          "train_metric": 0.00020888011204078794,
          "validation_metric": 0.00016803893959149718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.539924913000505
        },
        {
          "epoch": 575,
          "updates": 575,
          "samples_seen": 588800,
          "online_loss": 0.002073944779112935,
          "train_loss": 0.002065167995169759,
          "validation_loss": 0.0016763132298365235,
          "train_metric": 0.0002079961559502408,
          "validation_metric": 0.00016883212083484977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.547571984001479
        },
        {
          "epoch": 576,
          "updates": 576,
          "samples_seen": 589824,
          "online_loss": 0.002065167995169759,
          "train_loss": 0.0020572508219629526,
          "validation_loss": 0.0016658012755215168,
          "train_metric": 0.00020719875465147197,
          "validation_metric": 0.00016777341079432517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.556062567000481
        },
        {
          "epoch": 577,
          "updates": 577,
          "samples_seen": 590848,
          "online_loss": 0.0020572508219629526,
          "train_loss": 0.002050776034593582,
          "validation_loss": 0.0016516918549314141,
          "train_metric": 0.0002065466105705127,
          "validation_metric": 0.000166352343512699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.564643516001524
        },
        {
          "epoch": 578,
          "updates": 578,
          "samples_seen": 591872,
          "online_loss": 0.0020507758017629385,
          "train_loss": 0.0020438632927834988,
          "validation_loss": 0.0016602688701823354,
          "train_metric": 0.00020585040329024196,
          "validation_metric": 0.00016721618885640055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.57247017500049
        },
        {
          "epoch": 579,
          "updates": 579,
          "samples_seen": 592896,
          "online_loss": 0.002043863059952855,
          "train_loss": 0.002035802463069558,
          "validation_loss": 0.001639966038055718,
          "train_metric": 0.00020503855193965137,
          "validation_metric": 0.00016517135372851044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.5805841500005045
        },
        {
          "epoch": 580,
          "updates": 580,
          "samples_seen": 593920,
          "online_loss": 0.0020358022302389145,
          "train_loss": 0.0020279057789593935,
          "validation_loss": 0.0016393412370234728,
          "train_metric": 0.0002042432315647602,
          "validation_metric": 0.00016510843124706298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.588914471001772
        },
        {
          "epoch": 581,
          "updates": 581,
          "samples_seen": 594944,
          "online_loss": 0.00202790554612875,
          "train_loss": 0.00202102935872972,
          "validation_loss": 0.0016374809201806784,
          "train_metric": 0.000203550691367127,
          "validation_metric": 0.00016492108989041299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.597483651001312
        },
        {
          "epoch": 582,
          "updates": 582,
          "samples_seen": 595968,
          "online_loss": 0.0020210295915603638,
          "train_loss": 0.00201431754976511,
          "validation_loss": 0.0016208499437198043,
          "train_metric": 0.00020287469669710845,
          "validation_metric": 0.00016324606258422136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.605995528001586
        },
        {
          "epoch": 583,
          "updates": 583,
          "samples_seen": 596992,
          "online_loss": 0.0020143173169344664,
          "train_loss": 0.0020068699959665537,
          "validation_loss": 0.0016266735037788749,
          "train_metric": 0.00020212460367474705,
          "validation_metric": 0.0001638325775275007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.613784607001435
        },
        {
          "epoch": 584,
          "updates": 584,
          "samples_seen": 598016,
          "online_loss": 0.0020068702287971973,
          "train_loss": 0.0019991332665085793,
          "validation_loss": 0.0016120312502607703,
          "train_metric": 0.00020134537771809846,
          "validation_metric": 0.000162357886438258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.622230246000981
        },
        {
          "epoch": 585,
          "updates": 585,
          "samples_seen": 599040,
          "online_loss": 0.001999133499339223,
          "train_loss": 0.0019919327460229397,
          "validation_loss": 0.0016062880167737603,
          "train_metric": 0.00020062021212652326,
          "validation_metric": 0.0001617794478079304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.6308003880003525
        },
        {
          "epoch": 586,
          "updates": 586,
          "samples_seen": 600064,
          "online_loss": 0.0019919329788535833,
          "train_loss": 0.0019851853139698505,
          "validation_loss": 0.0016067482065409422,
          "train_metric": 0.0001999405794776976,
          "validation_metric": 0.0001618257665541023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.638691513000595
        },
        {
          "epoch": 587,
          "updates": 587,
          "samples_seen": 601088,
          "online_loss": 0.001985185081139207,
          "train_loss": 0.0019782320596277714,
          "validation_loss": 0.0015910009387880564,
          "train_metric": 0.0001992402976611629,
          "validation_metric": 0.00016023978241719306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.647211875000721
        },
        {
          "epoch": 588,
          "updates": 588,
          "samples_seen": 602112,
          "online_loss": 0.0019782320596277714,
          "train_loss": 0.0019708597101271152,
          "validation_loss": 0.00159403623547405,
          "train_metric": 0.00019849777163472027,
          "validation_metric": 0.00016054545994848013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.755996489000609
        },
        {
          "epoch": 589,
          "updates": 589,
          "samples_seen": 603136,
          "online_loss": 0.0019708597101271152,
          "train_loss": 0.001963529270142317,
          "validation_loss": 0.0015833494253456593,
          "train_metric": 0.00019775946566369385,
          "validation_metric": 0.0001594691420905292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.764760544001547
        },
        {
          "epoch": 590,
          "updates": 590,
          "samples_seen": 604160,
          "online_loss": 0.001963529270142317,
          "train_loss": 0.0019565788097679615,
          "validation_loss": 0.001576212584041059,
          "train_metric": 0.00019705944578163326,
          "validation_metric": 0.00015875036478973925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.7734012470009475
        },
        {
          "epoch": 591,
          "updates": 591,
          "samples_seen": 605184,
          "online_loss": 0.001956579042598605,
          "train_loss": 0.0019498133333399892,
          "validation_loss": 0.0015763118863105774,
          "train_metric": 0.0001963780669029802,
          "validation_metric": 0.00015876036195550114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.781333438000729
        },
        {
          "epoch": 592,
          "updates": 592,
          "samples_seen": 606208,
          "online_loss": 0.001949813449755311,
          "train_loss": 0.001942880218848586,
          "validation_loss": 0.0015622737118974328,
          "train_metric": 0.0001956797786988318,
          "validation_metric": 0.0001573464833199978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.789410049001162
        },
        {
          "epoch": 593,
          "updates": 593,
          "samples_seen": 607232,
          "online_loss": 0.001942880218848586,
          "train_loss": 0.0019357397686690092,
          "validation_loss": 0.00156325101852417,
          "train_metric": 0.0001949606230482459,
          "validation_metric": 0.0001574449270265177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.796990221000669
        },
        {
          "epoch": 594,
          "updates": 594,
          "samples_seen": 608256,
          "online_loss": 0.0019357397686690092,
          "train_loss": 0.0019286517053842545,
          "validation_loss": 0.0015538671286776662,
          "train_metric": 0.00019424672063905746,
          "validation_metric": 0.0001564997946843505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.8055067300010705
        },
        {
          "epoch": 595,
          "updates": 595,
          "samples_seen": 609280,
          "online_loss": 0.0019286517053842545,
          "train_loss": 0.001921795541420579,
          "validation_loss": 0.0015472606755793095,
          "train_metric": 0.000193556203157641,
          "validation_metric": 0.00015583443746436387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.81426059200021
        },
        {
          "epoch": 596,
          "updates": 596,
          "samples_seen": 610304,
          "online_loss": 0.0019217953085899353,
          "train_loss": 0.001915063476189971,
          "validation_loss": 0.0015460500726476312,
          "train_metric": 0.00019287820032332093,
          "validation_metric": 0.00015571249241475016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.822683829001107
        },
        {
          "epoch": 597,
          "updates": 597,
          "samples_seen": 611328,
          "online_loss": 0.0019150637090206146,
          "train_loss": 0.0019082622602581978,
          "validation_loss": 0.0015338815283030272,
          "train_metric": 0.00019219321256969124,
          "validation_metric": 0.00015448694466613233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.831431257000077
        },
        {
          "epoch": 598,
          "updates": 598,
          "samples_seen": 612352,
          "online_loss": 0.0019082623766735196,
          "train_loss": 0.001901349751278758,
          "validation_loss": 0.0015338738448917866,
          "train_metric": 0.00019149697618559003,
          "validation_metric": 0.00015448615886271,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.84015558000101
        },
        {
          "epoch": 599,
          "updates": 599,
          "samples_seen": 613376,
          "online_loss": 0.001901349751278758,
          "train_loss": 0.001894434099085629,
          "validation_loss": 0.0015242925146594644,
          "train_metric": 0.00019080044876318425,
          "validation_metric": 0.0001535211777081713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.84883172300033
        },
        {
          "epoch": 600,
          "updates": 600,
          "samples_seen": 614400,
          "online_loss": 0.0018944339826703072,
          "train_loss": 0.0018876395188272,
          "validation_loss": 0.0015191688435152173,
          "train_metric": 0.0001901161449495703,
          "validation_metric": 0.00015300509403459728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.857635376001781
        },
        {
          "epoch": 601,
          "updates": 601,
          "samples_seen": 615424,
          "online_loss": 0.0018876394024118781,
          "train_loss": 0.001880962634459138,
          "validation_loss": 0.0015160111943259835,
          "train_metric": 0.00018944368639495224,
          "validation_metric": 0.00015268707647919655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.866496831000404
        },
        {
          "epoch": 602,
          "updates": 602,
          "samples_seen": 616448,
          "online_loss": 0.0018809627508744597,
          "train_loss": 0.0018743057735264301,
          "validation_loss": 0.0015059909783303738,
          "train_metric": 0.00018877320690080523,
          "validation_metric": 0.00015167791571002454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.874622523000653
        },
        {
          "epoch": 603,
          "updates": 603,
          "samples_seen": 617472,
          "online_loss": 0.001874305889941752,
          "train_loss": 0.001867598737590015,
          "validation_loss": 0.0015053891111165285,
          "train_metric": 0.00018809770699590445,
          "validation_metric": 0.00015161724877543747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.8832036730000254
        },
        {
          "epoch": 604,
          "updates": 604,
          "samples_seen": 618496,
          "online_loss": 0.001867598621174693,
          "train_loss": 0.00186085794121027,
          "validation_loss": 0.0014953865902498364,
          "train_metric": 0.00018741880194284022,
          "validation_metric": 0.0001506098487880081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.891496878000908
        },
        {
          "epoch": 605,
          "updates": 605,
          "samples_seen": 619520,
          "online_loss": 0.00186085794121027,
          "train_loss": 0.0018541492754593492,
          "validation_loss": 0.0014921302208676934,
          "train_metric": 0.00018674314196687192,
          "validation_metric": 0.00015028187772259116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.900209423001797
        },
        {
          "epoch": 606,
          "updates": 606,
          "samples_seen": 620544,
          "online_loss": 0.001854149391874671,
          "train_loss": 0.0018475211691111326,
          "validation_loss": 0.0014862706884741783,
          "train_metric": 0.00018607558740768582,
          "validation_metric": 0.00014969173935241997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.908680151000226
        },
        {
          "epoch": 607,
          "updates": 607,
          "samples_seen": 621568,
          "online_loss": 0.0018475212855264544,
          "train_loss": 0.0018409562762826681,
          "validation_loss": 0.0014788773842155933,
          "train_metric": 0.00018541440658736974,
          "validation_metric": 0.0001489471032982692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.917532874000244
        },
        {
          "epoch": 608,
          "updates": 608,
          "samples_seen": 622592,
          "online_loss": 0.0018409565091133118,
          "train_loss": 0.0018344164127483964,
          "validation_loss": 0.0014766689855605364,
          "train_metric": 0.00018475569959264249,
          "validation_metric": 0.00014872469182591885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.926147371001207
        },
        {
          "epoch": 609,
          "updates": 609,
          "samples_seen": 623616,
          "online_loss": 0.0018344164127483964,
          "train_loss": 0.001827870262786746,
          "validation_loss": 0.0014672488905489445,
          "train_metric": 0.0001840964105213061,
          "validation_metric": 0.0001477759360568598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.934827438000866
        },
        {
          "epoch": 610,
          "updates": 610,
          "samples_seen": 624640,
          "online_loss": 0.0018278701463714242,
          "train_loss": 0.0018213107250630856,
          "validation_loss": 0.0014652927638962865,
          "train_metric": 0.00018343570991419256,
          "validation_metric": 0.0001475789467804134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.943491397001708
        },
        {
          "epoch": 611,
          "updates": 611,
          "samples_seen": 625664,
          "online_loss": 0.001821310492232442,
          "train_loss": 0.001814774819649756,
          "validation_loss": 0.0014571112114936113,
          "train_metric": 0.00018277748313266784,
          "validation_metric": 0.00014675491547677666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.952184880001369
        },
        {
          "epoch": 612,
          "updates": 612,
          "samples_seen": 626688,
          "online_loss": 0.001814774819649756,
          "train_loss": 0.0018082725582644343,
          "validation_loss": 0.0014528431929647923,
          "train_metric": 0.00018212258873973042,
          "validation_metric": 0.00014632503734901547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.961099157000717
        },
        {
          "epoch": 613,
          "updates": 613,
          "samples_seen": 627712,
          "online_loss": 0.0018082725582644343,
          "train_loss": 0.0018018176779150963,
          "validation_loss": 0.001447667134925723,
          "train_metric": 0.00018147248192690313,
          "validation_metric": 0.0001458037440897897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.969628727001691
        },
        {
          "epoch": 614,
          "updates": 614,
          "samples_seen": 628736,
          "online_loss": 0.0018018183764070272,
          "train_loss": 0.0017954037757590413,
          "validation_loss": 0.0014407843118533492,
          "train_metric": 0.0001808265078580007,
          "validation_metric": 0.00014511051995214075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.97825149900018
        },
        {
          "epoch": 615,
          "updates": 615,
          "samples_seen": 629760,
          "online_loss": 0.0017954038921743631,
          "train_loss": 0.001789010944776237,
          "validation_loss": 0.0014378979103639722,
          "train_metric": 0.00018018262926489115,
          "validation_metric": 0.0001448198308935389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.986998328000482
        },
        {
          "epoch": 616,
          "updates": 616,
          "samples_seen": 630784,
          "online_loss": 0.001789010944776237,
          "train_loss": 0.0017826305702328682,
          "validation_loss": 0.0014296466251835227,
          "train_metric": 0.0001795400312403217,
          "validation_metric": 0.0001439887855667621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.996018397001535
        },
        {
          "epoch": 617,
          "updates": 617,
          "samples_seen": 631808,
          "online_loss": 0.0017826309194788337,
          "train_loss": 0.0017762613715603948,
          "validation_loss": 0.0014271704712882638,
          "train_metric": 0.00017889855371322483,
          "validation_metric": 0.00014373939484357834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.004939559001286
        },
        {
          "epoch": 618,
          "updates": 618,
          "samples_seen": 632832,
          "online_loss": 0.0017762613715603948,
          "train_loss": 0.0017699062591418624,
          "validation_loss": 0.0014193058013916016,
          "train_metric": 0.00017825845861807466,
          "validation_metric": 0.00014294729044195265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.0137860620016
        },
        {
          "epoch": 619,
          "updates": 619,
          "samples_seen": 633856,
          "online_loss": 0.0017699062591418624,
          "train_loss": 0.0017635734984651208,
          "validation_loss": 0.0014158138073980808,
          "train_metric": 0.0001776206772774458,
          "validation_metric": 0.00014259557065088302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.022495823001009
        },
        {
          "epoch": 620,
          "updates": 620,
          "samples_seen": 634880,
          "online_loss": 0.0017635737312957644,
          "train_loss": 0.0017572708893567324,
          "validation_loss": 0.0014095058431848884,
          "train_metric": 0.00017698589363135397,
          "validation_metric": 0.00014196026313584298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.031044177001604
        },
        {
          "epoch": 621,
          "updates": 621,
          "samples_seen": 635904,
          "online_loss": 0.0017572710057720542,
          "train_loss": 0.0017510021571069956,
          "validation_loss": 0.0014044343261048198,
          "train_metric": 0.00017635452968534082,
          "validation_metric": 0.00014144947635941207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.039424071001122
        },
        {
          "epoch": 622,
          "updates": 622,
          "samples_seen": 636928,
          "online_loss": 0.0017510021571069956,
          "train_loss": 0.001744758104905486,
          "validation_loss": 0.0013997892383486032,
          "train_metric": 0.00017572561046108603,
          "validation_metric": 0.0001409816468367353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.048131993000425
        },
        {
          "epoch": 623,
          "updates": 623,
          "samples_seen": 637952,
          "online_loss": 0.0017447578720748425,
          "train_loss": 0.0017385409446433187,
          "validation_loss": 0.0013932945439592004,
          "train_metric": 0.00017509949975647032,
          "validation_metric": 0.00014032753824722022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.056813600000169
        },
        {
          "epoch": 624,
          "updates": 624,
          "samples_seen": 638976,
          "online_loss": 0.0017385410610586405,
          "train_loss": 0.001732347416691482,
          "validation_loss": 0.001389821758493781,
          "train_metric": 0.00017447570280637592,
          "validation_metric": 0.00013997775386087596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.065836171001138
        },
        {
          "epoch": 625,
          "updates": 625,
          "samples_seen": 640000,
          "online_loss": 0.001732347416691482,
          "train_loss": 0.0017261796165257692,
          "validation_loss": 0.0013825424248352647,
          "train_metric": 0.00017385449609719217,
          "validation_metric": 0.00013924461381975561,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.074247941000067
        },
        {
          "epoch": 626,
          "updates": 626,
          "samples_seen": 641024,
          "online_loss": 0.0017261796165257692,
          "train_loss": 0.0017200232250615954,
          "validation_loss": 0.0013796521816402674,
          "train_metric": 0.00017323443898931146,
          "validation_metric": 0.00013895351730752736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.1805808340013755
        },
        {
          "epoch": 627,
          "updates": 627,
          "samples_seen": 642048,
          "online_loss": 0.0017200233414769173,
          "train_loss": 0.0017138909315690398,
          "validation_loss": 0.0013722021831199527,
          "train_metric": 0.00017261681205127388,
          "validation_metric": 0.00013820317690260708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.189012940001703
        },
        {
          "epoch": 628,
          "updates": 628,
          "samples_seen": 643072,
          "online_loss": 0.001713890815153718,
          "train_loss": 0.0017077784286811948,
          "validation_loss": 0.0013693857472389936,
          "train_metric": 0.00017200116417370737,
          "validation_metric": 0.00013791953097097576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.197733883000183
        },
        {
          "epoch": 629,
          "updates": 629,
          "samples_seen": 644096,
          "online_loss": 0.0017077786615118384,
          "train_loss": 0.0017016875790432096,
          "validation_loss": 0.0013620840618386865,
          "train_metric": 0.0001713877427391708,
          "validation_metric": 0.00013718412083107978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.206241776000752
        },
        {
          "epoch": 630,
          "updates": 630,
          "samples_seen": 645120,
          "online_loss": 0.0017016876954585314,
          "train_loss": 0.0016956189647316933,
          "validation_loss": 0.0013590636663138866,
          "train_metric": 0.0001707765186438337,
          "validation_metric": 0.00013687991304323077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.218763511000361
        },
        {
          "epoch": 631,
          "updates": 631,
          "samples_seen": 646144,
          "online_loss": 0.0016956187319010496,
          "train_loss": 0.0016895754961296916,
          "validation_loss": 0.0013520455686375499,
          "train_metric": 0.00017016787023749202,
          "validation_metric": 0.00013617305376101285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.236841499001457
        },
        {
          "epoch": 632,
          "updates": 632,
          "samples_seen": 647168,
          "online_loss": 0.0016895756125450134,
          "train_loss": 0.001683550886809826,
          "validation_loss": 0.0013488336699083447,
          "train_metric": 0.00016956106992438436,
          "validation_metric": 0.00013584959378931671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.248872125001071
        },
        {
          "epoch": 633,
          "updates": 633,
          "samples_seen": 648192,
          "online_loss": 0.001683550886809826,
          "train_loss": 0.001677550608292222,
          "validation_loss": 0.0013420616742223501,
          "train_metric": 0.00016895672888495028,
          "validation_metric": 0.00013516753097064793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.262566930001412
        },
        {
          "epoch": 634,
          "updates": 634,
          "samples_seen": 649216,
          "online_loss": 0.0016775504918769002,
          "train_loss": 0.00167157466057688,
          "validation_loss": 0.0013387957587838173,
          "train_metric": 0.0001683549053268507,
          "validation_metric": 0.00013483861403074116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.275134995001281
        },
        {
          "epoch": 635,
          "updates": 635,
          "samples_seen": 650240,
          "online_loss": 0.00167157466057688,
          "train_loss": 0.0016656193183735013,
          "validation_loss": 0.0013320627622306347,
          "train_metric": 0.00016775507538113743,
          "validation_metric": 0.00013416048022918403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.288895855001101
        },
        {
          "epoch": 636,
          "updates": 636,
          "samples_seen": 651264,
          "online_loss": 0.0016656194347888231,
          "train_loss": 0.0016596965724602342,
          "validation_loss": 0.0013289391063153744,
          "train_metric": 0.000167158548720181,
          "validation_metric": 0.00013384586782194674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.295356774000538
        },
        {
          "epoch": 637,
          "updates": 637,
          "samples_seen": 652288,
          "online_loss": 0.0016596964560449123,
          "train_loss": 0.0016537872143089771,
          "validation_loss": 0.0013219768879935145,
          "train_metric": 0.000166563389939256,
          "validation_metric": 0.00013314468378666788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.30113971100036
        },
        {
          "epoch": 638,
          "updates": 638,
          "samples_seen": 653312,
          "online_loss": 0.0016537869814783335,
          "train_loss": 0.0016479069599881768,
          "validation_loss": 0.001319355913437903,
          "train_metric": 0.00016597115609329194,
          "validation_metric": 0.00013288068294059485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.3070289050010615
        },
        {
          "epoch": 639,
          "updates": 639,
          "samples_seen": 654336,
          "online_loss": 0.0016479071928188205,
          "train_loss": 0.0016420583706349134,
          "validation_loss": 0.0013117752969264984,
          "train_metric": 0.00016538208001293242,
          "validation_metric": 0.00013211718760430813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.312944449000497
        },
        {
          "epoch": 640,
          "updates": 640,
          "samples_seen": 655360,
          "online_loss": 0.0016420581378042698,
          "train_loss": 0.0016362385358661413,
          "validation_loss": 0.0013101993827149272,
          "train_metric": 0.00016479595797136426,
          "validation_metric": 0.00013195846986491233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.318794162001723
        },
        {
          "epoch": 641,
          "updates": 641,
          "samples_seen": 656384,
          "online_loss": 0.001636238768696785,
          "train_loss": 0.0016304575838148594,
          "validation_loss": 0.0013012768467888236,
          "train_metric": 0.00016421370673924685,
          "validation_metric": 0.00013105983089189976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.32466177400056
        },
        {
          "epoch": 642,
          "updates": 642,
          "samples_seen": 657408,
          "online_loss": 0.0016304573509842157,
          "train_loss": 0.0016247362364083529,
          "validation_loss": 0.0013017266755923629,
          "train_metric": 0.00016363746544811875,
          "validation_metric": 0.00013110516010783613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.330349590001788
        },
        {
          "epoch": 643,
          "updates": 643,
          "samples_seen": 658432,
          "online_loss": 0.0016247362364083529,
          "train_loss": 0.0016191044123843312,
          "validation_loss": 0.001290152664296329,
          "train_metric": 0.00016307026089634746,
          "validation_metric": 0.00012993944983463734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.335928279000655
        },
        {
          "epoch": 644,
          "updates": 644,
          "samples_seen": 659456,
          "online_loss": 0.0016191044123843312,
          "train_loss": 0.0016136205522343516,
          "validation_loss": 0.0012946263886988163,
          "train_metric": 0.0001625179429538548,
          "validation_metric": 0.00013039002078585327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.3412078020010085
        },
        {
          "epoch": 645,
          "updates": 645,
          "samples_seen": 660480,
          "online_loss": 0.0016136204358190298,
          "train_loss": 0.0016084205126389861,
          "validation_loss": 0.0012778927339240909,
          "train_metric": 0.00016199421952478588,
          "validation_metric": 0.0001287046616198495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.347094969001773
        },
        {
          "epoch": 646,
          "updates": 646,
          "samples_seen": 661504,
          "online_loss": 0.0016084205126389861,
          "train_loss": 0.0016037380555644631,
          "validation_loss": 0.0012907043565064669,
          "train_metric": 0.00016152262105606496,
          "validation_metric": 0.00012999500904697925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.352198600001429
        },
        {
          "epoch": 647,
          "updates": 647,
          "samples_seen": 662528,
          "online_loss": 0.0016037379391491413,
          "train_loss": 0.0016001967014744878,
          "validation_loss": 0.0012638012412935495,
          "train_metric": 0.00016116596816573292,
          "validation_metric": 0.00012728544243145734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.357775598000444
        },
        {
          "epoch": 648,
          "updates": 648,
          "samples_seen": 663552,
          "online_loss": 0.0016001968178898096,
          "train_loss": 0.001598820905201137,
          "validation_loss": 0.0012950869277119637,
          "train_metric": 0.00016102739027701318,
          "validation_metric": 0.0001304364122916013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.363418041000841
        },
        {
          "epoch": 649,
          "updates": 649,
          "samples_seen": 664576,
          "online_loss": 0.0015988207887858152,
          "train_loss": 0.001602677395567298,
          "validation_loss": 0.0012490260414779186,
          "train_metric": 0.00016141576634254307,
          "validation_metric": 0.0001257973344763741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.369091603000925
        },
        {
          "epoch": 650,
          "updates": 650,
          "samples_seen": 665600,
          "online_loss": 0.0016026771627366543,
          "train_loss": 0.001615991466678679,
          "validation_loss": 0.001325068762525916,
          "train_metric": 0.00016275672533083707,
          "validation_metric": 0.00013345605111680925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.374391479001133
        },
        {
          "epoch": 651,
          "updates": 651,
          "samples_seen": 666624,
          "online_loss": 0.0016159913502633572,
          "train_loss": 0.0016549833817407489,
          "validation_loss": 0.0012497017160058022,
          "train_metric": 0.00016668385069351643,
          "validation_metric": 0.00012586537923198193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.380057663000116
        },
        {
          "epoch": 652,
          "updates": 652,
          "samples_seen": 667648,
          "online_loss": 0.001654983265325427,
          "train_loss": 0.0017300434410572052,
          "validation_loss": 0.0014414639445021749,
          "train_metric": 0.00017424362886231393,
          "validation_metric": 0.00014517895760945976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.385302599001079
        },
        {
          "epoch": 653,
          "updates": 653,
          "samples_seen": 668672,
          "online_loss": 0.0017300433246418834,
          "train_loss": 0.0019178674556314945,
          "validation_loss": 0.0013549174182116985,
          "train_metric": 0.0001931605947902426,
          "validation_metric": 0.00013646233128383756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.3905518440005835
        },
        {
          "epoch": 654,
          "updates": 654,
          "samples_seen": 669696,
          "online_loss": 0.0019178672228008509,
          "train_loss": 0.0021248594857752323,
          "validation_loss": 0.0017713962588459253,
          "train_metric": 0.00021400807600002736,
          "validation_metric": 0.00017840854707174003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.396031162001236
        },
        {
          "epoch": 655,
          "updates": 655,
          "samples_seen": 670720,
          "online_loss": 0.002124859718605876,
          "train_loss": 0.002526247641071677,
          "validation_loss": 0.0016685379669070244,
          "train_metric": 0.00025443441700190306,
          "validation_metric": 0.00016804903862066567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.401113582000107
        },
        {
          "epoch": 656,
          "updates": 656,
          "samples_seen": 671744,
          "online_loss": 0.002526247873902321,
          "train_loss": 0.002281953115016222,
          "validation_loss": 0.0018980421591550112,
          "train_metric": 0.00022982995142228901,
          "validation_metric": 0.0001911638246383518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.406273323000278
        },
        {
          "epoch": 657,
          "updates": 657,
          "samples_seen": 672768,
          "online_loss": 0.002281953115016222,
          "train_loss": 0.0019463439239189029,
          "validation_loss": 0.001352455816231668,
          "train_metric": 0.0001960286172106862,
          "validation_metric": 0.00013621438120026141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.411367782000525
        },
        {
          "epoch": 658,
          "updates": 658,
          "samples_seen": 673792,
          "online_loss": 0.0019463440403342247,
          "train_loss": 0.001561179175041616,
          "validation_loss": 0.0012610781705006957,
          "train_metric": 0.00015723623801022768,
          "validation_metric": 0.00012701116793323308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.416552792001312
        },
        {
          "epoch": 659,
          "updates": 659,
          "samples_seen": 674816,
          "online_loss": 0.0015611785929650068,
          "train_loss": 0.0016703384462743998,
          "validation_loss": 0.0013741971924901009,
          "train_metric": 0.00016823035548441112,
          "validation_metric": 0.00013840410974808037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.421740265001063
        },
        {
          "epoch": 660,
          "updates": 660,
          "samples_seen": 675840,
          "online_loss": 0.0016703384462743998,
          "train_loss": 0.002021032851189375,
          "validation_loss": 0.0013854913413524628,
          "train_metric": 0.00020355102606117725,
          "validation_metric": 0.00013954161840956658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.427092421999987
        },
        {
          "epoch": 661,
          "updates": 661,
          "samples_seen": 676864,
          "online_loss": 0.0020210326183587313,
          "train_loss": 0.0019191072788089514,
          "validation_loss": 0.0015970816602930427,
          "train_metric": 0.00019328547932673246,
          "validation_metric": 0.00016085219976957887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.4323746410009335
        },
        {
          "epoch": 662,
          "updates": 662,
          "samples_seen": 677888,
          "online_loss": 0.0019191073952242732,
          "train_loss": 0.0016470681875944138,
          "validation_loss": 0.0012292952742427588,
          "train_metric": 0.00016588666767347604,
          "validation_metric": 0.00012381011038087308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.4381902090008225
        },
        {
          "epoch": 663,
          "updates": 663,
          "samples_seen": 678912,
          "online_loss": 0.0016470683040097356,
          "train_loss": 0.0015232807490974665,
          "validation_loss": 0.0012070589000359178,
          "train_metric": 0.00015341927064582705,
          "validation_metric": 0.0001215705560753122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.444118167000852
        },
        {
          "epoch": 664,
          "updates": 664,
          "samples_seen": 679936,
          "online_loss": 0.0015232806326821446,
          "train_loss": 0.0016998077044263482,
          "validation_loss": 0.0014194643590599298,
          "train_metric": 0.00017119839321821928,
          "validation_metric": 0.00014296325389295816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.526123097000891
        },
        {
          "epoch": 665,
          "updates": 665,
          "samples_seen": 680960,
          "online_loss": 0.0016998075880110264,
          "train_loss": 0.001851845532655716,
          "validation_loss": 0.0013088597916066647,
          "train_metric": 0.00018651106802280992,
          "validation_metric": 0.00013182355905883014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.531882601000689
        },
        {
          "epoch": 666,
          "updates": 666,
          "samples_seen": 681984,
          "online_loss": 0.0018518454162403941,
          "train_loss": 0.00164842011872679,
          "validation_loss": 0.0013716514222323895,
          "train_metric": 0.0001660228444961831,
          "validation_metric": 0.000138147734105587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.536951089001377
        },
        {
          "epoch": 667,
          "updates": 667,
          "samples_seen": 683008,
          "online_loss": 0.00164842011872679,
          "train_loss": 0.0014989454066380858,
          "validation_loss": 0.0011892300099134445,
          "train_metric": 0.00015096830611582845,
          "validation_metric": 0.00011977489339187741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.542406174999996
        },
        {
          "epoch": 668,
          "updates": 668,
          "samples_seen": 684032,
          "online_loss": 0.0014989454066380858,
          "train_loss": 0.0015806193696334958,
          "validation_loss": 0.0011855255579575896,
          "train_metric": 0.0001591941836522892,
          "validation_metric": 0.00011940180411329493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.548309493000488
        },
        {
          "epoch": 669,
          "updates": 669,
          "samples_seen": 685056,
          "online_loss": 0.001580619253218174,
          "train_loss": 0.001671160920523107,
          "validation_loss": 0.0014025218551978469,
          "train_metric": 0.00016831321408972144,
          "validation_metric": 0.0001412568672094494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.553508992001298
        },
        {
          "epoch": 670,
          "updates": 670,
          "samples_seen": 686080,
          "online_loss": 0.0016711608041077852,
          "train_loss": 0.0016136766644194722,
          "validation_loss": 0.001188051188364625,
          "train_metric": 0.00016252360364887863,
          "validation_metric": 0.00011965615703957155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.558582336001564
        },
        {
          "epoch": 671,
          "updates": 671,
          "samples_seen": 687104,
          "online_loss": 0.0016136764315888286,
          "train_loss": 0.0014887446304783225,
          "validation_loss": 0.001200599828734994,
          "train_metric": 0.0001499409117968753,
          "validation_metric": 0.00012092001270502806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.564061414001117
        },
        {
          "epoch": 672,
          "updates": 672,
          "samples_seen": 688128,
          "online_loss": 0.0014887447468936443,
          "train_loss": 0.0015163341304287314,
          "validation_loss": 0.0012404646258801222,
          "train_metric": 0.00015271964366547763,
          "validation_metric": 0.00012493504618760198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.569168257001365
        },
        {
          "epoch": 673,
          "updates": 673,
          "samples_seen": 689152,
          "online_loss": 0.0015163342468440533,
          "train_loss": 0.0016050263075158,
          "validation_loss": 0.0011858268408104777,
          "train_metric": 0.00016165236593224108,
          "validation_metric": 0.00011943213758058846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.574273006001022
        },
        {
          "epoch": 674,
          "updates": 674,
          "samples_seen": 690176,
          "online_loss": 0.0016050260746851563,
          "train_loss": 0.0015592009294778109,
          "validation_loss": 0.0012942652683705091,
          "train_metric": 0.0001570370077388361,
          "validation_metric": 0.00013035365554969758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.58008626100127
        },
        {
          "epoch": 675,
          "updates": 675,
          "samples_seen": 691200,
          "online_loss": 0.0015592009294778109,
          "train_loss": 0.0014746538363397121,
          "validation_loss": 0.0011593354865908623,
          "train_metric": 0.00014852173626422882,
          "validation_metric": 0.00011676402937155217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.585956030001398
        },
        {
          "epoch": 676,
          "updates": 676,
          "samples_seen": 692224,
          "online_loss": 0.0014746538363397121,
          "train_loss": 0.0014745797961950302,
          "validation_loss": 0.0011539694387465715,
          "train_metric": 0.0001485142856836319,
          "validation_metric": 0.00011622360034380108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.5917912860004435
        },
        {
          "epoch": 677,
          "updates": 677,
          "samples_seen": 693248,
          "online_loss": 0.0014745797961950302,
          "train_loss": 0.0015274368925020099,
          "validation_loss": 0.0012625448871403933,
          "train_metric": 0.00015383787103928626,
          "validation_metric": 0.00012715888442471623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.597460850000061
        },
        {
          "epoch": 678,
          "updates": 678,
          "samples_seen": 694272,
          "online_loss": 0.0015274370089173317,
          "train_loss": 0.001527465065009892,
          "validation_loss": 0.001152304233983159,
          "train_metric": 0.00015384067955892533,
          "validation_metric": 0.00011605585314100608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.603234056001384
        },
        {
          "epoch": 679,
          "updates": 679,
          "samples_seen": 695296,
          "online_loss": 0.001527465065009892,
          "train_loss": 0.0014631535159423947,
          "validation_loss": 0.0011922346893697977,
          "train_metric": 0.00014736347657162696,
          "validation_metric": 0.00012007752957288176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.608438162000311
        },
        {
          "epoch": 680,
          "updates": 680,
          "samples_seen": 696320,
          "online_loss": 0.0014631536323577166,
          "train_loss": 0.001445298665203154,
          "validation_loss": 0.001167269074358046,
          "train_metric": 0.00014556519454345107,
          "validation_metric": 0.00011756306776078418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.613928805001706
        },
        {
          "epoch": 681,
          "updates": 681,
          "samples_seen": 697344,
          "online_loss": 0.001445298665203154,
          "train_loss": 0.0014801130164414644,
          "validation_loss": 0.001131653320044279,
          "train_metric": 0.00014907158038113266,
          "validation_metric": 0.00011397598427720368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.6197552000012365
        },
        {
          "epoch": 682,
          "updates": 682,
          "samples_seen": 698368,
          "online_loss": 0.001480113249272108,
          "train_loss": 0.0014846606645733118,
          "validation_loss": 0.0012264249380677938,
          "train_metric": 0.00014952958736103028,
          "validation_metric": 0.00012352102203294635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.625050649001423
        },
        {
          "epoch": 683,
          "updates": 683,
          "samples_seen": 699392,
          "online_loss": 0.0014846606645733118,
          "train_loss": 0.0014492771588265896,
          "validation_loss": 0.0011199760483577847,
          "train_metric": 0.00014596588152926415,
          "validation_metric": 0.00011279989848844707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.630932312000368
        },
        {
          "epoch": 684,
          "updates": 684,
          "samples_seen": 700416,
          "online_loss": 0.0014492771588265896,
          "train_loss": 0.0014243531040847301,
          "validation_loss": 0.0011325521627441049,
          "train_metric": 0.00014345563249662519,
          "validation_metric": 0.00011406650446588174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.636411589001
        },
        {
          "epoch": 685,
          "updates": 685,
          "samples_seen": 701440,
          "online_loss": 0.001424353220500052,
          "train_loss": 0.0014387578703463078,
          "validation_loss": 0.0011722748167812824,
          "train_metric": 0.00014490641478914768,
          "validation_metric": 0.0001180672479677014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.641425463001724
        },
        {
          "epoch": 686,
          "updates": 686,
          "samples_seen": 702464,
          "online_loss": 0.0014387579867616296,
          "train_loss": 0.0014537799870595336,
          "validation_loss": 0.0011129422346130013,
          "train_metric": 0.00014641940651927143,
          "validation_metric": 0.00011209145304746926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.647405007000998
        },
        {
          "epoch": 687,
          "updates": 687,
          "samples_seen": 703488,
          "online_loss": 0.0014537799870595336,
          "train_loss": 0.001433923956938088,
          "validation_loss": 0.0011725632939487696,
          "train_metric": 0.00014441956591326743,
          "validation_metric": 0.000118096279038582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.652789167001174
        },
        {
          "epoch": 688,
          "updates": 688,
          "samples_seen": 704512,
          "online_loss": 0.0014339238405227661,
          "train_loss": 0.0014091881457716227,
          "validation_loss": 0.001115376246161759,
          "train_metric": 0.0001419282634742558,
          "validation_metric": 0.00011233659461140633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.657919379000305
        },
        {
          "epoch": 689,
          "updates": 689,
          "samples_seen": 705536,
          "online_loss": 0.0014091881457716227,
          "train_loss": 0.0014083446003496647,
          "validation_loss": 0.0011066754814237356,
          "train_metric": 0.00014184329484123737,
          "validation_metric": 0.00011146031465614215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.663904422000996
        },
        {
          "epoch": 690,
          "updates": 690,
          "samples_seen": 706560,
          "online_loss": 0.0014083446003496647,
          "train_loss": 0.0014195566764101386,
          "validation_loss": 0.0011555933160707355,
          "train_metric": 0.0001429725525667891,
          "validation_metric": 0.00011638712021522224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.669136847000118
        },
        {
          "epoch": 691,
          "updates": 691,
          "samples_seen": 707584,
          "online_loss": 0.0014195565599948168,
          "train_loss": 0.0014165964676067233,
          "validation_loss": 0.001093667116947472,
          "train_metric": 0.00014267439837567508,
          "validation_metric": 0.00011015014752047136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.674812771001598
        },
        {
          "epoch": 692,
          "updates": 692,
          "samples_seen": 708608,
          "online_loss": 0.0014165963511914015,
          "train_loss": 0.001397411455400288,
          "validation_loss": 0.0011275260476395488,
          "train_metric": 0.00014074215141590685,
          "validation_metric": 0.00011356028699083254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.680467398000474
        },
        {
          "epoch": 693,
          "updates": 693,
          "samples_seen": 709632,
          "online_loss": 0.0013974113389849663,
          "train_loss": 0.0013864771462976933,
          "validation_loss": 0.0011050765169784427,
          "train_metric": 0.00013964090612716973,
          "validation_metric": 0.0001112992613343522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.685784104000049
        },
        {
          "epoch": 694,
          "updates": 694,
          "samples_seen": 710656,
          "online_loss": 0.0013864769134670496,
          "train_loss": 0.0013902627397328615,
          "validation_loss": 0.00108575657941401,
          "train_metric": 0.00014002218085806817,
          "validation_metric": 0.00010935343016171828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.691440146001696
        },
        {
          "epoch": 695,
          "updates": 695,
          "samples_seen": 711680,
          "online_loss": 0.0013902628561481833,
          "train_loss": 0.001393008278682828,
          "validation_loss": 0.0011295950971543789,
          "train_metric": 0.00014029868179932237,
          "validation_metric": 0.0001137686922447756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.697013592000076
        },
        {
          "epoch": 696,
          "updates": 696,
          "samples_seen": 712704,
          "online_loss": 0.001393008278682828,
          "train_loss": 0.0013847004156559706,
          "validation_loss": 0.0010766988852992654,
          "train_metric": 0.00013946194667369127,
          "validation_metric": 0.00010844117059605196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.70270892300141
        },
        {
          "epoch": 697,
          "updates": 697,
          "samples_seen": 713728,
          "online_loss": 0.001384700182825327,
          "train_loss": 0.0013715917011722922,
          "validation_loss": 0.001098404754884541,
          "train_metric": 0.00013814169506076723,
          "validation_metric": 0.00011062730482080951,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.707822127000327
        },
        {
          "epoch": 698,
          "updates": 698,
          "samples_seen": 714752,
          "online_loss": 0.0013715915847569704,
          "train_loss": 0.0013663828140124679,
          "validation_loss": 0.0010924582602456212,
          "train_metric": 0.00013761708396486938,
          "validation_metric": 0.00011002839892171323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.713036970000758
        },
        {
          "epoch": 699,
          "updates": 699,
          "samples_seen": 715776,
          "online_loss": 0.0013663830468431115,
          "train_loss": 0.0013682349817827344,
          "validation_loss": 0.0010712582152336836,
          "train_metric": 0.00013780361041426659,
          "validation_metric": 0.00010789321095217019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.718735587000992
        },
        {
          "epoch": 700,
          "updates": 700,
          "samples_seen": 716800,
          "online_loss": 0.0013682349817827344,
          "train_loss": 0.0013669769978150725,
          "validation_loss": 0.001106509123928845,
          "train_metric": 0.00013767692144028842,
          "validation_metric": 0.00011144355812575668,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.723951983000006
        },
        {
          "epoch": 701,
          "updates": 701,
          "samples_seen": 717824,
          "online_loss": 0.0013669768813997507,
          "train_loss": 0.0013592320028692484,
          "validation_loss": 0.001064647687599063,
          "train_metric": 0.00013689688057638705,
          "validation_metric": 0.00010722740989876911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.8063209990014
        },
        {
          "epoch": 702,
          "updates": 702,
          "samples_seen": 718848,
          "online_loss": 0.0013592320028692484,
          "train_loss": 0.001350049744360149,
          "validation_loss": 0.0010788331273943186,
          "train_metric": 0.00013597206270787865,
          "validation_metric": 0.00010865611693589017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.812044850001257
        },
        {
          "epoch": 703,
          "updates": 703,
          "samples_seen": 719872,
          "online_loss": 0.001350049627944827,
          "train_loss": 0.001345761469565332,
          "validation_loss": 0.0010743469465523958,
          "train_metric": 0.00013554016186390072,
          "validation_metric": 0.00010820427996804938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.817487856000298
        },
        {
          "epoch": 704,
          "updates": 704,
          "samples_seen": 720896,
          "online_loss": 0.001345761469565332,
          "train_loss": 0.0013452671701088548,
          "validation_loss": 0.0010552663588896394,
          "train_metric": 0.00013549037976190448,
          "validation_metric": 0.00010628255404299125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.823266599001727
        },
        {
          "epoch": 705,
          "updates": 705,
          "samples_seen": 721920,
          "online_loss": 0.0013452671701088548,
          "train_loss": 0.001343017560429871,
          "validation_loss": 0.0010818514274433255,
          "train_metric": 0.00013526382099371403,
          "validation_metric": 0.00010896010644501075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.828716888001509
        },
        {
          "epoch": 706,
          "updates": 706,
          "samples_seen": 722944,
          "online_loss": 0.001343017676845193,
          "train_loss": 0.001337007386609912,
          "validation_loss": 0.001048424979671836,
          "train_metric": 0.0001346584758721292,
          "validation_metric": 0.00010559351358097047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.83463506300177
        },
        {
          "epoch": 707,
          "updates": 707,
          "samples_seen": 723968,
          "online_loss": 0.00133700727019459,
          "train_loss": 0.0013298039557412267,
          "validation_loss": 0.0010610073804855347,
          "train_metric": 0.00013393296103458852,
          "validation_metric": 0.00010686078167054802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.839726623000388
        },
        {
          "epoch": 708,
          "updates": 708,
          "samples_seen": 724992,
          "online_loss": 0.0013298038393259048,
          "train_loss": 0.0013251224299892783,
          "validation_loss": 0.0010535988258197904,
          "train_metric": 0.00013346147898118943,
          "validation_metric": 0.00010611460311338305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.844970002000991
        },
        {
          "epoch": 709,
          "updates": 709,
          "samples_seen": 726016,
          "online_loss": 0.0013251223135739565,
          "train_loss": 0.0013229685137048364,
          "validation_loss": 0.0010403692722320557,
          "train_metric": 0.0001332445244770497,
          "validation_metric": 0.00010478219337528571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.850634280001032
        },
        {
          "epoch": 710,
          "updates": 710,
          "samples_seen": 727040,
          "online_loss": 0.0013229682808741927,
          "train_loss": 0.0013205645373091102,
          "validation_loss": 0.0010595620842650533,
          "train_metric": 0.0001330024388153106,
          "validation_metric": 0.00010671521886251867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.855852548000257
        },
        {
          "epoch": 711,
          "updates": 711,
          "samples_seen": 728064,
          "online_loss": 0.001320564653724432,
          "train_loss": 0.0013161352835595608,
          "validation_loss": 0.0010333084501326084,
          "train_metric": 0.00013255632075015455,
          "validation_metric": 0.00010407104127807543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.8616491760003555
        },
        {
          "epoch": 712,
          "updates": 712,
          "samples_seen": 729088,
          "online_loss": 0.0013161353999748826,
          "train_loss": 0.0013102798257023096,
          "validation_loss": 0.0010470213601365685,
          "train_metric": 0.00013196658983360976,
          "validation_metric": 0.00010545216355239972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.866876920001232
        },
        {
          "epoch": 713,
          "updates": 713,
          "samples_seen": 730112,
          "online_loss": 0.0013102798257023096,
          "train_loss": 0.001305137644521892,
          "validation_loss": 0.0010344288311898708,
          "train_metric": 0.00013144867261871696,
          "validation_metric": 0.0001041838841047138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.87192583200158
        },
        {
          "epoch": 714,
          "updates": 714,
          "samples_seen": 731136,
          "online_loss": 0.0013051375281065702,
          "train_loss": 0.00130161561537534,
          "validation_loss": 0.0010281694121658802,
          "train_metric": 0.00013109395513311028,
          "validation_metric": 0.00010355345148127526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.877997373001563
        },
        {
          "epoch": 715,
          "updates": 715,
          "samples_seen": 732160,
          "online_loss": 0.0013016154989600182,
          "train_loss": 0.0012988595990464091,
          "validation_loss": 0.001037972280755639,
          "train_metric": 0.0001308163773501292,
          "validation_metric": 0.00010454078437760472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.883287156000733
        },
        {
          "epoch": 716,
          "updates": 716,
          "samples_seen": 733184,
          "online_loss": 0.0012988594826310873,
          "train_loss": 0.0012955237179994583,
          "validation_loss": 0.001017785514704883,
          "train_metric": 0.00013048040273133665,
          "validation_metric": 0.00010250761988572776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.889035173000593
        },
        {
          "epoch": 717,
          "updates": 717,
          "samples_seen": 734208,
          "online_loss": 0.0012955237179994583,
          "train_loss": 0.0012910395162180066,
          "validation_loss": 0.0010318036656826735,
          "train_metric": 0.0001300287840422243,
          "validation_metric": 0.00010391948308097199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.894503610001266
        },
        {
          "epoch": 718,
          "updates": 718,
          "samples_seen": 735232,
          "online_loss": 0.0012910396326333284,
          "train_loss": 0.0012861109571531415,
          "validation_loss": 0.0010148235596716404,
          "train_metric": 0.000129532374558039,
          "validation_metric": 0.00010220930562354624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.900208446000761
        },
        {
          "epoch": 719,
          "updates": 719,
          "samples_seen": 736256,
          "online_loss": 0.0012861109571531415,
          "train_loss": 0.001281643402762711,
          "validation_loss": 0.0010167530272156,
          "train_metric": 0.00012908244389109313,
          "validation_metric": 0.00010240366100333631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.905282823001471
        },
        {
          "epoch": 720,
          "updates": 720,
          "samples_seen": 737280,
          "online_loss": 0.001281643402762711,
          "train_loss": 0.0012779986718669534,
          "validation_loss": 0.0010159628000110388,
          "train_metric": 0.00012871534272562712,
          "validation_metric": 0.00010232406202703714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.910809460001474
        },
        {
          "epoch": 721,
          "updates": 721,
          "samples_seen": 738304,
          "online_loss": 0.0012779987882822752,
          "train_loss": 0.0012747659347951412,
          "validation_loss": 0.001004502410069108,
          "train_metric": 0.00012838977272622287,
          "validation_metric": 0.00010116980411112309,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.916546920001565
        },
        {
          "epoch": 722,
          "updates": 722,
          "samples_seen": 739328,
          "online_loss": 0.0012747659347951412,
          "train_loss": 0.00127130257897079,
          "validation_loss": 0.0010148824658244848,
          "train_metric": 0.00012804093421436846,
          "validation_metric": 0.00010221524280495942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.921808469000098
        },
        {
          "epoch": 723,
          "updates": 723,
          "samples_seen": 740352,
          "online_loss": 0.00127130257897079,
          "train_loss": 0.001267330371774733,
          "validation_loss": 0.0009983094641938806,
          "train_metric": 0.0001276408729609102,
          "validation_metric": 0.0001005460653686896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.927543830001014
        },
        {
          "epoch": 724,
          "updates": 724,
          "samples_seen": 741376,
          "online_loss": 0.001267330488190055,
          "train_loss": 0.0012629885459318757,
          "validation_loss": 0.0010063610970973969,
          "train_metric": 0.00012720357335638255,
          "validation_metric": 0.00010135700722457841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.933053088001543
        },
        {
          "epoch": 725,
          "updates": 725,
          "samples_seen": 742400,
          "online_loss": 0.0012629885459318757,
          "train_loss": 0.0012587114470079541,
          "validation_loss": 0.0009957243455573916,
          "train_metric": 0.00012677280756179243,
          "validation_metric": 0.00010028571705333889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.9387160840014985
        },
        {
          "epoch": 726,
          "updates": 726,
          "samples_seen": 743424,
          "online_loss": 0.0012587113305926323,
          "train_loss": 0.0012547816149890423,
          "validation_loss": 0.0009943118784576654,
          "train_metric": 0.00012637699546758085,
          "validation_metric": 0.00010014345025410876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.944631391001167
        },
        {
          "epoch": 727,
          "updates": 727,
          "samples_seen": 744448,
          "online_loss": 0.0012547816149890423,
          "train_loss": 0.0012511962559074163,
          "validation_loss": 0.0009947052458301187,
          "train_metric": 0.0001260159187950194,
          "validation_metric": 0.0001001830751192756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.9498039270001755
        },
        {
          "epoch": 728,
          "updates": 728,
          "samples_seen": 745472,
          "online_loss": 0.00125119648873806,
          "train_loss": 0.0012477251002565026,
          "validation_loss": 0.0009845838649198413,
          "train_metric": 0.0001256663235835731,
          "validation_metric": 9.916368435369805e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.955696734001322
        },
        {
          "epoch": 729,
          "updates": 729,
          "samples_seen": 746496,
          "online_loss": 0.0012477251002565026,
          "train_loss": 0.0012441447470337152,
          "validation_loss": 0.0009914268739521503,
          "train_metric": 0.00012530569802038372,
          "validation_metric": 9.985288488678634e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.961651939000149
        },
        {
          "epoch": 730,
          "updates": 730,
          "samples_seen": 747520,
          "online_loss": 0.0012441447470337152,
          "train_loss": 0.0012403441360220313,
          "validation_loss": 0.0009779331739991903,
          "train_metric": 0.00012492292444221675,
          "validation_metric": 9.84938524197787e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.967515030000868
        },
        {
          "epoch": 731,
          "updates": 731,
          "samples_seen": 748544,
          "online_loss": 0.0012403441360220313,
          "train_loss": 0.0012363840360194445,
          "validation_loss": 0.0009838471887633204,
          "train_metric": 0.00012452407099772245,
          "validation_metric": 9.90894841379486e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.972673930000383
        },
        {
          "epoch": 732,
          "updates": 732,
          "samples_seen": 749568,
          "online_loss": 0.0012363840360194445,
          "train_loss": 0.001232403563335538,
          "validation_loss": 0.0009738494991324842,
          "train_metric": 0.00012412316573318094,
          "validation_metric": 9.808255708776414e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.978559059001782
        },
        {
          "epoch": 733,
          "updates": 733,
          "samples_seen": 750592,
          "online_loss": 0.0012324033305048943,
          "train_loss": 0.0012285251868888736,
          "validation_loss": 0.0009745362331159413,
          "train_metric": 0.0001237325486727059,
          "validation_metric": 9.815170778892934e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.983821096000611
        },
        {
          "epoch": 734,
          "updates": 734,
          "samples_seen": 751616,
          "online_loss": 0.0012285251868888736,
          "train_loss": 0.00122480234131217,
          "validation_loss": 0.0009712952305562794,
          "train_metric": 0.0001233576040249318,
          "validation_metric": 9.782530833035707e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.989506921001521
        },
        {
          "epoch": 735,
          "updates": 735,
          "samples_seen": 752640,
          "online_loss": 0.0012248021084815264,
          "train_loss": 0.0012211999855935574,
          "validation_loss": 0.0009657734190113842,
          "train_metric": 0.0001229947665706277,
          "validation_metric": 9.726917051011696e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.995818754001448
        },
        {
          "epoch": 736,
          "updates": 736,
          "samples_seen": 753664,
          "online_loss": 0.0012211999855935574,
          "train_loss": 0.001217647222802043,
          "validation_loss": 0.0009678869391791523,
          "train_metric": 0.0001226369640789926,
          "validation_metric": 9.74820286501199e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.001277439001569
        },
        {
          "epoch": 737,
          "updates": 737,
          "samples_seen": 754688,
          "online_loss": 0.001217647222802043,
          "train_loss": 0.0012140823528170586,
          "validation_loss": 0.0009582139318808913,
          "train_metric": 0.0001222779246745631,
          "validation_metric": 9.65078070294112e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.007019742000921
        },
        {
          "epoch": 738,
          "updates": 738,
          "samples_seen": 755712,
          "online_loss": 0.0012140822364017367,
          "train_loss": 0.0012104560155421495,
          "validation_loss": 0.0009627051185816526,
          "train_metric": 0.00012191270798211917,
          "validation_metric": 9.696013876236975e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.012469711000449
        },
        {
          "epoch": 739,
          "updates": 739,
          "samples_seen": 756736,
          "online_loss": 0.0012104561319574714,
          "train_loss": 0.0012067683273926377,
          "validation_loss": 0.0009523113258183002,
          "train_metric": 0.00012154129944974557,
          "validation_metric": 9.591331036062911e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.018260316001033
        },
        {
          "epoch": 740,
          "updates": 740,
          "samples_seen": 757760,
          "online_loss": 0.0012067683273926377,
          "train_loss": 0.0012030473444610834,
          "validation_loss": 0.0009561579208821058,
          "train_metric": 0.0001211665294249542,
          "validation_metric": 9.630072599975392e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.103077454001323
        },
        {
          "epoch": 741,
          "updates": 741,
          "samples_seen": 758784,
          "online_loss": 0.0012030473444610834,
          "train_loss": 0.0011993213556706905,
          "validation_loss": 0.0009474526741541922,
          "train_metric": 0.00012079125008312985,
          "validation_metric": 9.542397310724482e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.108840114000486
        },
        {
          "epoch": 742,
          "updates": 742,
          "samples_seen": 759808,
          "online_loss": 0.0011993214720860124,
          "train_loss": 0.0011956216767430305,
          "validation_loss": 0.0009486889466643333,
          "train_metric": 0.00012041863374179229,
          "validation_metric": 9.554847929393873e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.1143676080009755
        },
        {
          "epoch": 743,
          "updates": 743,
          "samples_seen": 760832,
          "online_loss": 0.0011956217931583524,
          "train_loss": 0.001191961346194148,
          "validation_loss": 0.0009430210920982063,
          "train_metric": 0.00012004996824543923,
          "validation_metric": 9.497762948740274e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.120134214001155
        },
        {
          "epoch": 744,
          "updates": 744,
          "samples_seen": 761856,
          "online_loss": 0.0011919612297788262,
          "train_loss": 0.0011883482802659273,
          "validation_loss": 0.0009413242805749178,
          "train_metric": 0.00011968606850132346,
          "validation_metric": 9.480673907091841e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.125711492000846
        },
        {
          "epoch": 745,
          "updates": 745,
          "samples_seen": 762880,
          "online_loss": 0.0011883482802659273,
          "train_loss": 0.0011847674613818526,
          "validation_loss": 0.0009388496982865036,
          "train_metric": 0.00011932542838621885,
          "validation_metric": 9.455750114284456e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.132032537001578
        },
        {
          "epoch": 746,
          "updates": 746,
          "samples_seen": 763904,
          "online_loss": 0.0011847673449665308,
          "train_loss": 0.0011812233133241534,
          "validation_loss": 0.0009344621794298291,
          "train_metric": 0.00011896847718162462,
          "validation_metric": 9.411561768501997e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.1379442930010555
        },
        {
          "epoch": 747,
          "updates": 747,
          "samples_seen": 764928,
          "online_loss": 0.0011812231969088316,
          "train_loss": 0.0011776987230405211,
          "validation_loss": 0.0009343561250716448,
          "train_metric": 0.00011861350503750145,
          "validation_metric": 9.41049147513695e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.1440474780010845
        },
        {
          "epoch": 748,
          "updates": 748,
          "samples_seen": 765952,
          "online_loss": 0.0011776987230405211,
          "train_loss": 0.0011741957860067487,
          "validation_loss": 0.0009278104407712817,
          "train_metric": 0.0001182606938527897,
          "validation_metric": 9.344566205982119e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.149990450001496
        },
        {
          "epoch": 749,
          "updates": 749,
          "samples_seen": 766976,
          "online_loss": 0.0011741957860067487,
          "train_loss": 0.0011707014637067914,
          "validation_loss": 0.0009294442133978009,
          "train_metric": 0.00011790876305894926,
          "validation_metric": 9.361021511722356e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.155252539001594
        },
        {
          "epoch": 750,
          "updates": 750,
          "samples_seen": 768000,
          "online_loss": 0.0011707014637067914,
          "train_loss": 0.0011672151740640402,
          "validation_loss": 0.0009214765159413218,
          "train_metric": 0.000117557639896404,
          "validation_metric": 9.280773520004004e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.161239317001673
        },
        {
          "epoch": 751,
          "updates": 751,
          "samples_seen": 769024,
          "online_loss": 0.0011672148248180747,
          "train_loss": 0.0011637471616268158,
          "validation_loss": 0.0009244628599844873,
          "train_metric": 0.00011720836482709274,
          "validation_metric": 9.310850145993754e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.16656241200144
        },
        {
          "epoch": 752,
          "updates": 752,
          "samples_seen": 770048,
          "online_loss": 0.0011637472780421376,
          "train_loss": 0.0011602990562096238,
          "validation_loss": 0.0009152793209068477,
          "train_metric": 0.00011686109064612538,
          "validation_metric": 9.218358900398016e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.172326925001471
        },
        {
          "epoch": 753,
          "updates": 753,
          "samples_seen": 771072,
          "online_loss": 0.0011602992890402675,
          "train_loss": 0.0011568652698770165,
          "validation_loss": 0.000919497397262603,
          "train_metric": 0.00011651523527689278,
          "validation_metric": 9.260841034119949e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.177797750000536
        },
        {
          "epoch": 754,
          "updates": 754,
          "samples_seen": 772096,
          "online_loss": 0.0011568652698770165,
          "train_loss": 0.0011534711811691523,
          "validation_loss": 0.0009089487721212208,
          "train_metric": 0.00011617339623626322,
          "validation_metric": 9.154600411420688e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.183371283001179
        },
        {
          "epoch": 755,
          "updates": 755,
          "samples_seen": 773120,
          "online_loss": 0.0011534711811691523,
          "train_loss": 0.0011501205153763294,
          "validation_loss": 0.0009149337420240045,
          "train_metric": 0.00011583593004615977,
          "validation_metric": 9.214877354679629e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.188680505001685
        },
        {
          "epoch": 756,
          "updates": 756,
          "samples_seen": 774144,
          "online_loss": 0.0011501205153763294,
          "train_loss": 0.00114684016443789,
          "validation_loss": 0.0009024436585605145,
          "train_metric": 0.0001155055288108997,
          "validation_metric": 9.08908259589225e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.194344554000054
        },
        {
          "epoch": 757,
          "updates": 757,
          "samples_seen": 775168,
          "online_loss": 0.0011468400480225682,
          "train_loss": 0.0011436508502811193,
          "validation_loss": 0.0009112732950598001,
          "train_metric": 0.00011518432438606396,
          "validation_metric": 9.178010805044323e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.199646623001172
        },
        {
          "epoch": 758,
          "updates": 758,
          "samples_seen": 776192,
          "online_loss": 0.001143651083111763,
          "train_loss": 0.001140636159107089,
          "validation_loss": 0.0008953846408985555,
          "train_metric": 0.00011488069139886647,
          "validation_metric": 9.017984848469496e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.205353476001619
        },
        {
          "epoch": 759,
          "updates": 759,
          "samples_seen": 777216,
          "online_loss": 0.0011406360426917672,
          "train_loss": 0.0011378565104678273,
          "validation_loss": 0.0009090955136343837,
          "train_metric": 0.00011460074165370315,
          "validation_metric": 9.156076703220606e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.211248505000185
        },
        {
          "epoch": 760,
          "updates": 760,
          "samples_seen": 778240,
          "online_loss": 0.0011378566268831491,
          "train_loss": 0.0011355523020029068,
          "validation_loss": 0.0008874381892383099,
          "train_metric": 0.00011436866770964116,
          "validation_metric": 8.937952952692285e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.217014238000047
        },
        {
          "epoch": 761,
          "updates": 761,
          "samples_seen": 779264,
          "online_loss": 0.0011355524184182286,
          "train_loss": 0.0011338979238644242,
          "validation_loss": 0.0009101931354962289,
          "train_metric": 0.00011420205555623397,
          "validation_metric": 9.167131065623835e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.222171979999985
        },
        {
          "epoch": 762,
          "updates": 762,
          "samples_seen": 780288,
          "online_loss": 0.0011338978074491024,
          "train_loss": 0.0011336765019223094,
          "validation_loss": 0.0008787760743871331,
          "train_metric": 0.00011417974747018889,
          "validation_metric": 8.850710582919419e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.2282012260002375
        },
        {
          "epoch": 763,
          "updates": 763,
          "samples_seen": 781312,
          "online_loss": 0.0011336765019223094,
          "train_loss": 0.0011354289017617702,
          "validation_loss": 0.0009187362156808376,
          "train_metric": 0.00011435625492595136,
          "validation_metric": 9.253175812773407e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.233538948001296
        },
        {
          "epoch": 764,
          "updates": 764,
          "samples_seen": 782336,
          "online_loss": 0.0011354289017617702,
          "train_loss": 0.0011419369839131832,
          "validation_loss": 0.000871324329636991,
          "train_metric": 0.0001150117241195403,
          "validation_metric": 8.775659080129117e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.239501292000568
        },
        {
          "epoch": 765,
          "updates": 765,
          "samples_seen": 783360,
          "online_loss": 0.0011419369839131832,
          "train_loss": 0.0011544449953362346,
          "validation_loss": 0.0009458434651605785,
          "train_metric": 0.0001162714688689448,
          "validation_metric": 9.526190115138888e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.245192638001754
        },
        {
          "epoch": 766,
          "updates": 766,
          "samples_seen": 784384,
          "online_loss": 0.0011544449953362346,
          "train_loss": 0.001183788408525288,
          "validation_loss": 0.0008757453178986907,
          "train_metric": 0.00011922682460863143,
          "validation_metric": 8.820187213132158e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.250281489001281
        },
        {
          "epoch": 767,
          "updates": 767,
          "samples_seen": 785408,
          "online_loss": 0.001183788408525288,
          "train_loss": 0.0012290503364056349,
          "validation_loss": 0.0010221396805718541,
          "train_metric": 0.00012378543033264577,
          "validation_metric": 0.00010294617095496505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.255378027000916
        },
        {
          "epoch": 768,
          "updates": 768,
          "samples_seen": 786432,
          "online_loss": 0.0012290503364056349,
          "train_loss": 0.0013320469297468662,
          "validation_loss": 0.0009325070423074067,
          "train_metric": 0.00013415889407042414,
          "validation_metric": 9.391867934027687e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.260567116001766
        },
        {
          "epoch": 769,
          "updates": 769,
          "samples_seen": 787456,
          "online_loss": 0.0013320469297468662,
          "train_loss": 0.001447569578886032,
          "validation_loss": 0.001210077665746212,
          "train_metric": 0.00014579389244318008,
          "validation_metric": 0.00012187458924017847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.265879363001659
        },
        {
          "epoch": 770,
          "updates": 770,
          "samples_seen": 788480,
          "online_loss": 0.001447569578886032,
          "train_loss": 0.0017056534998118877,
          "validation_loss": 0.0011190726654604077,
          "train_metric": 0.000171787163708359,
          "validation_metric": 0.0001127089126384817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.271074670001326
        },
        {
          "epoch": 771,
          "updates": 771,
          "samples_seen": 789504,
          "online_loss": 0.001705653266981244,
          "train_loss": 0.0017694695852696896,
          "validation_loss": 0.0014643807662650943,
          "train_metric": 0.00017821449728216976,
          "validation_metric": 0.00014748706598766148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.276260691000061
        },
        {
          "epoch": 772,
          "updates": 772,
          "samples_seen": 790528,
          "online_loss": 0.0017694695852696896,
          "train_loss": 0.0019188530277460814,
          "validation_loss": 0.0012297009816393256,
          "train_metric": 0.00019325983885210007,
          "validation_metric": 0.00012385098671074957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.281710613000541
        },
        {
          "epoch": 773,
          "updates": 773,
          "samples_seen": 791552,
          "online_loss": 0.0019188530277460814,
          "train_loss": 0.0015443939482793212,
          "validation_loss": 0.001284940168261528,
          "train_metric": 0.0001555456838104874,
          "validation_metric": 0.00012941447494085878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.286868518000119
        },
        {
          "epoch": 774,
          "updates": 774,
          "samples_seen": 792576,
          "online_loss": 0.0015443939482793212,
          "train_loss": 0.0012419045669957995,
          "validation_loss": 0.00087747722864151,
          "train_metric": 0.00012508008512668312,
          "validation_metric": 8.837628411129117e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.29207678900093
        },
        {
          "epoch": 775,
          "updates": 775,
          "samples_seen": 793600,
          "online_loss": 0.0012419045669957995,
          "train_loss": 0.0010886722011491656,
          "validation_loss": 0.0008532182546332479,
          "train_metric": 0.0001096470732591115,
          "validation_metric": 8.593301754444838e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.297911074001604
        },
        {
          "epoch": 776,
          "updates": 776,
          "samples_seen": 794624,
          "online_loss": 0.0010886722011491656,
          "train_loss": 0.001196759520098567,
          "validation_loss": 0.0009861482540145516,
          "train_metric": 0.00012053323007421568,
          "validation_metric": 9.932123793987557e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.303300348001358
        },
        {
          "epoch": 777,
          "updates": 777,
          "samples_seen": 795648,
          "online_loss": 0.001196759520098567,
          "train_loss": 0.0014250656822696328,
          "validation_loss": 0.0009742411784827709,
          "train_metric": 0.0001435274025425315,
          "validation_metric": 9.812200005399063e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.308541440001136
        },
        {
          "epoch": 778,
          "updates": 778,
          "samples_seen": 796672,
          "online_loss": 0.0014250659151002765,
          "train_loss": 0.001444288995116949,
          "validation_loss": 0.0012066807830706239,
          "train_metric": 0.00014546350575983524,
          "validation_metric": 0.00012153246643720195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.400611412000217
        },
        {
          "epoch": 779,
          "updates": 779,
          "samples_seen": 797696,
          "online_loss": 0.001444288995116949,
          "train_loss": 0.001352604478597641,
          "validation_loss": 0.0009447252959944308,
          "train_metric": 0.00013622936967294663,
          "validation_metric": 9.514927660347894e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.405772806001551
        },
        {
          "epoch": 780,
          "updates": 780,
          "samples_seen": 798720,
          "online_loss": 0.0013526043621823192,
          "train_loss": 0.0011354308808222413,
          "validation_loss": 0.0009405653108842671,
          "train_metric": 0.00011435644410084933,
          "validation_metric": 9.473028330830857e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.411051776000022
        },
        {
          "epoch": 781,
          "updates": 781,
          "samples_seen": 799744,
          "online_loss": 0.0011354308808222413,
          "train_loss": 0.0010728419292718172,
          "validation_loss": 0.0008598702261224389,
          "train_metric": 0.00010805271449498832,
          "validation_metric": 8.660297316964716e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.416216945000997
        },
        {
          "epoch": 782,
          "updates": 782,
          "samples_seen": 800768,
          "online_loss": 0.0010728419292718172,
          "train_loss": 0.001170840347185731,
          "validation_loss": 0.00084824173245579,
          "train_metric": 0.00011792274744948372,
          "validation_metric": 8.543180592823774e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.422016108001117
        },
        {
          "epoch": 783,
          "updates": 783,
          "samples_seen": 801792,
          "online_loss": 0.001170840347185731,
          "train_loss": 0.0012660787906497717,
          "validation_loss": 0.0010621469700708985,
          "train_metric": 0.00012751480971928686,
          "validation_metric": 0.00010697555262595415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.427365569001267
        },
        {
          "epoch": 784,
          "updates": 784,
          "samples_seen": 802816,
          "online_loss": 0.0012660787906497717,
          "train_loss": 0.0012825756566599011,
          "validation_loss": 0.0008919386309571564,
          "train_metric": 0.00012917631829623133,
          "validation_metric": 8.983279985841364e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.432528369001375
        },
        {
          "epoch": 785,
          "updates": 785,
          "samples_seen": 803840,
          "online_loss": 0.0012825755402445793,
          "train_loss": 0.0011488363379612565,
          "validation_loss": 0.0009559606551192701,
          "train_metric": 0.00011570658534765244,
          "validation_metric": 9.62808626354672e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.437663795000844
        },
        {
          "epoch": 786,
          "updates": 786,
          "samples_seen": 804864,
          "online_loss": 0.0011488362215459347,
          "train_loss": 0.0010592566104605794,
          "validation_loss": 0.0008241181494668126,
          "train_metric": 0.00010668442700989544,
          "validation_metric": 8.300215995404869e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.443749440000829
        },
        {
          "epoch": 787,
          "updates": 787,
          "samples_seen": 805888,
          "online_loss": 0.001059256261214614,
          "train_loss": 0.001081748865544796,
          "validation_loss": 0.0008185934275388718,
          "train_metric": 0.0001089498182409443,
          "validation_metric": 8.244573109550402e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.449539321001794
        },
        {
          "epoch": 788,
          "updates": 788,
          "samples_seen": 806912,
          "online_loss": 0.0010817492147907615,
          "train_loss": 0.0011551071656867862,
          "validation_loss": 0.0009565191576257348,
          "train_metric": 0.00011633815302047879,
          "validation_metric": 9.633710578782484e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.454858529001285
        },
        {
          "epoch": 789,
          "updates": 789,
          "samples_seen": 807936,
          "online_loss": 0.0011551071656867862,
          "train_loss": 0.0011955017689615488,
          "validation_loss": 0.0008626592461951077,
          "train_metric": 0.00012040655565215275,
          "validation_metric": 8.688386878930032e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.4599938960000145
        },
        {
          "epoch": 790,
          "updates": 790,
          "samples_seen": 808960,
          "online_loss": 0.0011955020017921925,
          "train_loss": 0.001133180339820683,
          "validation_loss": 0.0009442570735700428,
          "train_metric": 0.00011412977619329467,
          "validation_metric": 9.51021138462238e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.465126822000457
        },
        {
          "epoch": 791,
          "updates": 791,
          "samples_seen": 809984,
          "online_loss": 0.001133180339820683,
          "train_loss": 0.0010615414939820766,
          "validation_loss": 0.000820231216493994,
          "train_metric": 0.00010691455099731684,
          "validation_metric": 8.26106988824904e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.4702315280010225
        },
        {
          "epoch": 792,
          "updates": 792,
          "samples_seen": 811008,
          "online_loss": 0.001061541261151433,
          "train_loss": 0.001041037030518055,
          "validation_loss": 0.0008187565836124122,
          "train_metric": 0.00010484942322364077,
          "validation_metric": 8.246216748375446e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.475330826000572
        },
        {
          "epoch": 793,
          "updates": 793,
          "samples_seen": 812032,
          "online_loss": 0.0010410369141027331,
          "train_loss": 0.0010785022750496864,
          "validation_loss": 0.0008860885864123702,
          "train_metric": 0.00010862280760193244,
          "validation_metric": 8.924359281081706e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.480840813001123
        },
        {
          "epoch": 794,
          "updates": 794,
          "samples_seen": 813056,
          "online_loss": 0.0010785023914650083,
          "train_loss": 0.0011191718513146043,
          "validation_loss": 0.0008206183556467295,
          "train_metric": 0.00011271890252828598,
          "validation_metric": 8.264966891147196e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.486098595001749
        },
        {
          "epoch": 795,
          "updates": 795,
          "samples_seen": 814080,
          "online_loss": 0.0011191719677299261,
          "train_loss": 0.0011048588203266263,
          "validation_loss": 0.0009200095082633197,
          "train_metric": 0.00011127733159810305,
          "validation_metric": 9.265998960472643e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.491349720001381
        },
        {
          "epoch": 796,
          "updates": 796,
          "samples_seen": 815104,
          "online_loss": 0.0011048589367419481,
          "train_loss": 0.0010637189261615276,
          "validation_loss": 0.000803362054284662,
          "train_metric": 0.00010713386291172355,
          "validation_metric": 8.091169002000242e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.497256760001619
        },
        {
          "epoch": 797,
          "updates": 797,
          "samples_seen": 816128,
          "online_loss": 0.0010637188097462058,
          "train_loss": 0.001027604565024376,
          "validation_loss": 0.0008235177374444902,
          "train_metric": 0.00010349657532060519,
          "validation_metric": 8.294169674627483e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.5025166450013785
        },
        {
          "epoch": 798,
          "updates": 798,
          "samples_seen": 817152,
          "online_loss": 0.001027604565024376,
          "train_loss": 0.0010314199607819319,
          "validation_loss": 0.0008323512738570571,
          "train_metric": 0.00010388084047008306,
          "validation_metric": 8.383136446354911e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.5078147180011
        },
        {
          "epoch": 799,
          "updates": 799,
          "samples_seen": 818176,
          "online_loss": 0.0010314199607819319,
          "train_loss": 0.001058157067745924,
          "validation_loss": 0.000794737774413079,
          "train_metric": 0.00010657370876288041,
          "validation_metric": 8.004308620002121e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.513921796000432
        },
        {
          "epoch": 800,
          "updates": 800,
          "samples_seen": 819200,
          "online_loss": 0.0010581569513306022,
          "train_loss": 0.0010681835701689124,
          "validation_loss": 0.0008818783098831773,
          "train_metric": 0.00010758354619611055,
          "validation_metric": 8.881955727702007e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.519233322000218
        },
        {
          "epoch": 801,
          "updates": 801,
          "samples_seen": 820224,
          "online_loss": 0.0010681836865842342,
          "train_loss": 0.001056190230883658,
          "validation_loss": 0.000796620617620647,
          "train_metric": 0.00010637560626491904,
          "validation_metric": 8.023271220736206e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.524342724000235
        },
        {
          "epoch": 802,
          "updates": 802,
          "samples_seen": 821248,
          "online_loss": 0.001056190230883658,
          "train_loss": 0.0010265136370435357,
          "validation_loss": 0.0008354674209840596,
          "train_metric": 0.0001033867010846734,
          "validation_metric": 8.414522017119452e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.529882502001783
        },
        {
          "epoch": 803,
          "updates": 803,
          "samples_seen": 822272,
          "online_loss": 0.0010265137534588575,
          "train_loss": 0.001009210362099111,
          "validation_loss": 0.0007976507768034935,
          "train_metric": 0.00010164399282075465,
          "validation_metric": 8.033646736294031e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.535009490000448
        },
        {
          "epoch": 804,
          "updates": 804,
          "samples_seen": 823296,
          "online_loss": 0.001009210478514433,
          "train_loss": 0.0010140311205759645,
          "validation_loss": 0.0007859288016334176,
          "train_metric": 0.00010212949564447626,
          "validation_metric": 7.915585592854768e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.540826611000739
        },
        {
          "epoch": 805,
          "updates": 805,
          "samples_seen": 824320,
          "online_loss": 0.0010140311205759645,
          "train_loss": 0.001026876736432314,
          "validation_loss": 0.0008356301696039736,
          "train_metric": 0.00010342327004764229,
          "validation_metric": 8.416161290369928e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.54655411100066
        },
        {
          "epoch": 806,
          "updates": 806,
          "samples_seen": 825344,
          "online_loss": 0.001026876736432314,
          "train_loss": 0.0010329252108931541,
          "validation_loss": 0.0007811274845153093,
          "train_metric": 0.00010403244232293218,
          "validation_metric": 7.867230306146666e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.55231587200069
        },
        {
          "epoch": 807,
          "updates": 807,
          "samples_seen": 826368,
          "online_loss": 0.0010329250944778323,
          "train_loss": 0.00102230126503855,
          "validation_loss": 0.0008343576919287443,
          "train_metric": 0.00010296244727214798,
          "validation_metric": 8.403346146224067e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.557529862000592
        },
        {
          "epoch": 808,
          "updates": 808,
          "samples_seen": 827392,
          "online_loss": 0.00102230126503855,
          "train_loss": 0.0010052447905763984,
          "validation_loss": 0.0007762329187244177,
          "train_metric": 0.00010124457185156643,
          "validation_metric": 7.817934238119051e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.563470709001194
        },
        {
          "epoch": 809,
          "updates": 809,
          "samples_seen": 828416,
          "online_loss": 0.0010052449069917202,
          "train_loss": 0.0009936195565387607,
          "validation_loss": 0.0007885916857048869,
          "train_metric": 0.00010007372475229204,
          "validation_metric": 7.942406227812171e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.568793040001765
        },
        {
          "epoch": 810,
          "updates": 810,
          "samples_seen": 829440,
          "online_loss": 0.0009936195565387607,
          "train_loss": 0.000993420253507793,
          "validation_loss": 0.0007933712331578135,
          "train_metric": 0.00010005363583331928,
          "validation_metric": 7.99054469098337e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.573937082001066
        },
        {
          "epoch": 811,
          "updates": 811,
          "samples_seen": 830464,
          "online_loss": 0.0009934201370924711,
          "train_loss": 0.0009991388069465756,
          "validation_loss": 0.0007693846127949655,
          "train_metric": 0.00010062961518997326,
          "validation_metric": 7.748960342723876e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.580118051000682
        },
        {
          "epoch": 812,
          "updates": 812,
          "samples_seen": 831488,
          "online_loss": 0.0009991389233618975,
          "train_loss": 0.0010026107775047421,
          "validation_loss": 0.0008162833983078599,
          "train_metric": 0.00010097927588503808,
          "validation_metric": 8.221306779887527e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.585427397001695
        },
        {
          "epoch": 813,
          "updates": 813,
          "samples_seen": 832512,
          "online_loss": 0.0010026106610894203,
          "train_loss": 0.0009996660519391298,
          "validation_loss": 0.00076726742554456,
          "train_metric": 0.00010068268602481112,
          "validation_metric": 7.727636693743989e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.591101268000784
        },
        {
          "epoch": 814,
          "updates": 814,
          "samples_seen": 833536,
          "online_loss": 0.0009996660519391298,
          "train_loss": 0.0009896536357700825,
          "validation_loss": 0.0008003024850040674,
          "train_metric": 9.967429650714621e-05,
          "validation_metric": 8.060353866312653e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.596395418000611
        },
        {
          "epoch": 815,
          "updates": 815,
          "samples_seen": 834560,
          "online_loss": 0.0009896538686007261,
          "train_loss": 0.000980270211584866,
          "validation_loss": 0.0007655048975721002,
          "train_metric": 9.87292078207247e-05,
          "validation_metric": 7.709884812356904e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.602049316001285
        },
        {
          "epoch": 816,
          "updates": 816,
          "samples_seen": 835584,
          "online_loss": 0.0009802700951695442,
          "train_loss": 0.0009748269803822041,
          "validation_loss": 0.0007709423080086708,
          "train_metric": 9.818100079428405e-05,
          "validation_metric": 7.764648762531579e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.607364357001643
        },
        {
          "epoch": 817,
          "updates": 817,
          "samples_seen": 836608,
          "online_loss": 0.0009748268639668822,
          "train_loss": 0.0009741358808241785,
          "validation_loss": 0.0007784331683069468,
          "train_metric": 9.811139898374677e-05,
          "validation_metric": 7.840094622224569e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.697126679000576
        },
        {
          "epoch": 818,
          "updates": 818,
          "samples_seen": 837632,
          "online_loss": 0.0009741358808241785,
          "train_loss": 0.0009759974200278521,
          "validation_loss": 0.0007574241026304662,
          "train_metric": 9.829888585954905e-05,
          "validation_metric": 7.628498133271933e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.702989119999984
        },
        {
          "epoch": 819,
          "updates": 819,
          "samples_seen": 838656,
          "online_loss": 0.0009759974200278521,
          "train_loss": 0.0009760982356965542,
          "validation_loss": 0.0007877253228798509,
          "train_metric": 9.830903582042083e-05,
          "validation_metric": 7.933680899441242e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.708295595000891
        },
        {
          "epoch": 820,
          "updates": 820,
          "samples_seen": 839680,
          "online_loss": 0.0009760981774888933,
          "train_loss": 0.0009734651539474726,
          "validation_loss": 0.0007507947739213705,
          "train_metric": 9.804383444134146e-05,
          "validation_metric": 7.561729580629617e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.7142386230007105
        },
        {
          "epoch": 821,
          "updates": 821,
          "samples_seen": 840704,
          "online_loss": 0.0009734650375321507,
          "train_loss": 0.0009680193034000695,
          "validation_loss": 0.0007772344979457557,
          "train_metric": 9.749535820446908e-05,
          "validation_metric": 7.828021625755355e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.719468975001291
        },
        {
          "epoch": 822,
          "updates": 822,
          "samples_seen": 841728,
          "online_loss": 0.0009680192451924086,
          "train_loss": 0.0009619335178285837,
          "validation_loss": 0.0007504140958189964,
          "train_metric": 9.688241698313504e-05,
          "validation_metric": 7.557895878562704e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.725242922000689
        },
        {
          "epoch": 823,
          "updates": 823,
          "samples_seen": 842752,
          "online_loss": 0.0009619335178285837,
          "train_loss": 0.0009570996626280248,
          "validation_loss": 0.0007605706341564655,
          "train_metric": 9.639557538321242e-05,
          "validation_metric": 7.660188566660509e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.7307310800006235
        },
        {
          "epoch": 824,
          "updates": 824,
          "samples_seen": 843776,
          "online_loss": 0.0009570996044203639,
          "train_loss": 0.0009542727493681014,
          "validation_loss": 0.0007569299777969718,
          "train_metric": 9.611085260985419e-05,
          "validation_metric": 7.623521378263831e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.735746916001517
        },
        {
          "epoch": 825,
          "updates": 825,
          "samples_seen": 844800,
          "online_loss": 0.0009542728075757623,
          "train_loss": 0.0009529599919915199,
          "validation_loss": 0.0007467370014637709,
          "train_metric": 9.597864118404686e-05,
          "validation_metric": 7.520861981902272e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.741398342001048
        },
        {
          "epoch": 826,
          "updates": 826,
          "samples_seen": 845824,
          "online_loss": 0.0009529599919915199,
          "train_loss": 0.0009522997424937785,
          "validation_loss": 0.0007620205287821591,
          "train_metric": 9.591213893145323e-05,
          "validation_metric": 7.674791413592175e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.746993088001545
        },
        {
          "epoch": 827,
          "updates": 827,
          "samples_seen": 846848,
          "online_loss": 0.0009522996842861176,
          "train_loss": 0.0009511517127975821,
          "validation_loss": 0.0007389791426248848,
          "train_metric": 9.579650941304862e-05,
          "validation_metric": 7.44272765587084e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.752601422000225
        },
        {
          "epoch": 828,
          "updates": 828,
          "samples_seen": 847872,
          "online_loss": 0.0009511516545899212,
          "train_loss": 0.0009488220675848424,
          "validation_loss": 0.0007616137154400349,
          "train_metric": 9.556187433190644e-05,
          "validation_metric": 7.670693594263867e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.757902125000328
        },
        {
          "epoch": 829,
          "updates": 829,
          "samples_seen": 848896,
          "online_loss": 0.0009488218929618597,
          "train_loss": 0.0009456617408432066,
          "validation_loss": 0.0007357277208939195,
          "train_metric": 9.524358029011637e-05,
          "validation_metric": 7.409980753436685e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.764052685000934
        },
        {
          "epoch": 830,
          "updates": 830,
          "samples_seen": 849920,
          "online_loss": 0.0009456616826355457,
          "train_loss": 0.0009416943066753447,
          "validation_loss": 0.0007530621369369328,
          "train_metric": 9.484399924986064e-05,
          "validation_metric": 7.584565173601732e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.769219616000555
        },
        {
          "epoch": 831,
          "updates": 831,
          "samples_seen": 850944,
          "online_loss": 0.0009416944230906665,
          "train_loss": 0.0009377971873618662,
          "validation_loss": 0.000733937427867204,
          "train_metric": 9.445151226827875e-05,
          "validation_metric": 7.391947292489931e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.7750305740009935
        },
        {
          "epoch": 832,
          "updates": 832,
          "samples_seen": 851968,
          "online_loss": 0.0009377972455695271,
          "train_loss": 0.000934319687075913,
          "validation_loss": 0.0007413021521642804,
          "train_metric": 9.410125494468957e-05,
          "validation_metric": 7.46612495277077e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.780557552001483
        },
        {
          "epoch": 833,
          "updates": 833,
          "samples_seen": 852992,
          "online_loss": 0.000934319687075913,
          "train_loss": 0.0009314002236351371,
          "validation_loss": 0.0007355682319030166,
          "train_metric": 9.380721166962758e-05,
          "validation_metric": 7.408374221995473e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.785704936000911
        },
        {
          "epoch": 834,
          "updates": 834,
          "samples_seen": 854016,
          "online_loss": 0.000931400281842798,
          "train_loss": 0.0009290908346883953,
          "validation_loss": 0.000732520071323961,
          "train_metric": 9.357462113257498e-05,
          "validation_metric": 7.377674046438187e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.791588057001718
        },
        {
          "epoch": 835,
          "updates": 835,
          "samples_seen": 855040,
          "online_loss": 0.0009290907182730734,
          "train_loss": 0.0009271275484934449,
          "validation_loss": 0.0007379070739261806,
          "train_metric": 9.337688970845193e-05,
          "validation_metric": 7.43192940717563e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.797355642000184
        },
        {
          "epoch": 836,
          "updates": 836,
          "samples_seen": 856064,
          "online_loss": 0.0009271276067011058,
          "train_loss": 0.0009252981981262565,
          "validation_loss": 0.0007252930081449449,
          "train_metric": 9.319264790974557e-05,
          "validation_metric": 7.304885366465896e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.803101286001038
        },
        {
          "epoch": 837,
          "updates": 837,
          "samples_seen": 857088,
          "online_loss": 0.0009252981399185956,
          "train_loss": 0.0009234405588358641,
          "validation_loss": 0.0007375777931883931,
          "train_metric": 9.300554665969685e-05,
          "validation_metric": 7.428614480886608e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.808371283001179
        },
        {
          "epoch": 838,
          "updates": 838,
          "samples_seen": 858112,
          "online_loss": 0.000923440617043525,
          "train_loss": 0.0009214263409376144,
          "validation_loss": 0.0007199107785709202,
          "train_metric": 9.28026856854558e-05,
          "validation_metric": 7.250676571857184e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.814155038000536
        },
        {
          "epoch": 839,
          "updates": 839,
          "samples_seen": 859136,
          "online_loss": 0.0009214263409376144,
          "train_loss": 0.0009191870340146124,
          "validation_loss": 0.0007357128779403865,
          "train_metric": 9.257714555133134e-05,
          "validation_metric": 7.40982941351831e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.819381119001264
        },
        {
          "epoch": 840,
          "updates": 840,
          "samples_seen": 860160,
          "online_loss": 0.0009191869175992906,
          "train_loss": 0.0009168258984573185,
          "validation_loss": 0.0007159227970987558,
          "train_metric": 9.233933815266937e-05,
          "validation_metric": 7.21051255823113e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.82522378000067
        },
        {
          "epoch": 841,
          "updates": 841,
          "samples_seen": 861184,
          "online_loss": 0.0009168258402496576,
          "train_loss": 0.0009142394410446286,
          "validation_loss": 0.0007314187823794782,
          "train_metric": 9.20788588700816e-05,
          "validation_metric": 7.366581121459603e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.8308223680014635
        },
        {
          "epoch": 842,
          "updates": 842,
          "samples_seen": 862208,
          "online_loss": 0.0009142394992522895,
          "train_loss": 0.0009116401197388768,
          "validation_loss": 0.0007118477951735258,
          "train_metric": 9.181704081129283e-05,
          "validation_metric": 7.169468881329522e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.836556969001322
        },
        {
          "epoch": 843,
          "updates": 843,
          "samples_seen": 863232,
          "online_loss": 0.0009116401197388768,
          "train_loss": 0.0009089630329981446,
          "validation_loss": 0.000726240046788007,
          "train_metric": 9.154743020189926e-05,
          "validation_metric": 7.314424874493852e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.841842953001105
        },
        {
          "epoch": 844,
          "updates": 844,
          "samples_seen": 864256,
          "online_loss": 0.0009089630912058055,
          "train_loss": 0.0009063341422006488,
          "validation_loss": 0.000708867097273469,
          "train_metric": 9.128265810431913e-05,
          "validation_metric": 7.139451190596446e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.849147421000453
        },
        {
          "epoch": 845,
          "updates": 845,
          "samples_seen": 865280,
          "online_loss": 0.0009063342004083097,
          "train_loss": 0.0009037256822921336,
          "validation_loss": 0.0007218053215183318,
          "train_metric": 9.10199451027438e-05,
          "validation_metric": 7.269758498296142e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.854446930001359
        },
        {
          "epoch": 846,
          "updates": 846,
          "samples_seen": 866304,
          "online_loss": 0.0009037256822921336,
          "train_loss": 0.0009011822985485196,
          "validation_loss": 0.0007058145711198449,
          "train_metric": 9.076378046302125e-05,
          "validation_metric": 7.108705904101953e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.8600859210000635
        },
        {
          "epoch": 847,
          "updates": 847,
          "samples_seen": 867328,
          "online_loss": 0.0009011823567561805,
          "train_loss": 0.000898689788300544,
          "validation_loss": 0.0007171983597800136,
          "train_metric": 9.051273809745908e-05,
          "validation_metric": 7.223359716590494e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.865769560001354
        },
        {
          "epoch": 848,
          "updates": 848,
          "samples_seen": 868352,
          "online_loss": 0.0008986897300928831,
          "train_loss": 0.0008963024010881782,
          "validation_loss": 0.0007020133198238909,
          "train_metric": 9.027230407809839e-05,
          "validation_metric": 7.070421270327643e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.8715609530008805
        },
        {
          "epoch": 849,
          "updates": 849,
          "samples_seen": 869376,
          "online_loss": 0.000896303215995431,
          "train_loss": 0.0008939746185205877,
          "validation_loss": 0.0007136326166801155,
          "train_metric": 9.003785817185417e-05,
          "validation_metric": 7.187445589806885e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.876844195001468
        },
        {
          "epoch": 850,
          "updates": 850,
          "samples_seen": 870400,
          "online_loss": 0.0008939744438976049,
          "train_loss": 0.0008917977102100849,
          "validation_loss": 0.0006980680045671761,
          "train_metric": 8.981861174106598e-05,
          "validation_metric": 7.030685083009303e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.8827313570000115
        },
        {
          "epoch": 851,
          "updates": 851,
          "samples_seen": 871424,
          "online_loss": 0.0008917978266254067,
          "train_loss": 0.0008897353545762599,
          "validation_loss": 0.0007112654275260866,
          "train_metric": 8.961088315118104e-05,
          "validation_metric": 7.163605187088251e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.8887478900014685
        },
        {
          "epoch": 852,
          "updates": 852,
          "samples_seen": 872448,
          "online_loss": 0.0008897353545762599,
          "train_loss": 0.0008879226516000926,
          "validation_loss": 0.0006930834497325122,
          "train_metric": 8.942832937464118e-05,
          "validation_metric": 6.98048243066296e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.894779477001066
        },
        {
          "epoch": 853,
          "updates": 853,
          "samples_seen": 873472,
          "online_loss": 0.0008879227098077536,
          "train_loss": 0.0008863905677571893,
          "validation_loss": 0.0007103747338987887,
          "train_metric": 8.927400631364435e-05,
          "validation_metric": 7.154634658945724e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.90024001600068
        },
        {
          "epoch": 854,
          "updates": 854,
          "samples_seen": 874496,
          "online_loss": 0.0008863906259648502,
          "train_loss": 0.0008854441111907363,
          "validation_loss": 0.0006877043633721769,
          "train_metric": 8.917867671698332e-05,
          "validation_metric": 6.92630565026775e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.905906900001355
        },
        {
          "epoch": 855,
          "updates": 855,
          "samples_seen": 875520,
          "online_loss": 0.0008854441111907363,
          "train_loss": 0.0008851956226862967,
          "validation_loss": 0.0007133296458050609,
          "train_metric": 8.915366197470576e-05,
          "validation_metric": 7.184394780779257e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.911177016001602
        },
        {
          "epoch": 856,
          "updates": 856,
          "samples_seen": 876544,
          "online_loss": 0.0008851956226862967,
          "train_loss": 0.0008865473209880292,
          "validation_loss": 0.0006824967567808926,
          "train_metric": 8.928980241762474e-05,
          "validation_metric": 6.87385763740167e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.9170310360004805
        },
        {
          "epoch": 857,
          "updates": 857,
          "samples_seen": 877568,
          "online_loss": 0.0008865472627803683,
          "train_loss": 0.00088972452795133,
          "validation_loss": 0.0007230003830045462,
          "train_metric": 8.960980630945414e-05,
          "validation_metric": 7.281795114977285e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.029041830001006
        },
        {
          "epoch": 858,
          "updates": 858,
          "samples_seen": 878592,
          "online_loss": 0.0008897243533283472,
          "train_loss": 0.00089768081670627,
          "validation_loss": 0.0006792794447392225,
          "train_metric": 9.041113662533462e-05,
          "validation_metric": 6.841453432571143e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.03931614600151
        },
        {
          "epoch": 859,
          "updates": 859,
          "samples_seen": 879616,
          "online_loss": 0.0008976808749139309,
          "train_loss": 0.0009105730568990111,
          "validation_loss": 0.0007487751427106559,
          "train_metric": 9.170957491733134e-05,
          "validation_metric": 7.541388913523406e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.047312325001258
        },
        {
          "epoch": 860,
          "updates": 860,
          "samples_seen": 880640,
          "online_loss": 0.0009105729986913502,
          "train_loss": 0.0009385540033690631,
          "validation_loss": 0.0006881256704218686,
          "train_metric": 9.452771337237209e-05,
          "validation_metric": 6.930548988748342e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.05547819000094
        },
        {
          "epoch": 861,
          "updates": 861,
          "samples_seen": 881664,
          "online_loss": 0.0009385539451614022,
          "train_loss": 0.0009777449304237962,
          "validation_loss": 0.0008142952574416995,
          "train_metric": 9.847489127423614e-05,
          "validation_metric": 8.201284072129056e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.063738134000232
        },
        {
          "epoch": 862,
          "updates": 862,
          "samples_seen": 882688,
          "online_loss": 0.0009777449304237962,
          "train_loss": 0.001063765143044293,
          "validation_loss": 0.0007397630834020674,
          "train_metric": 0.00010713853407651186,
          "validation_metric": 7.45062279747799e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.071854485000586
        },
        {
          "epoch": 863,
          "updates": 863,
          "samples_seen": 883712,
          "online_loss": 0.0010637652594596148,
          "train_loss": 0.0011565065942704678,
          "validation_loss": 0.0009650971624068916,
          "train_metric": 0.00011647910287138075,
          "validation_metric": 9.720105299493298e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.079737242000192
        },
        {
          "epoch": 864,
          "updates": 864,
          "samples_seen": 884736,
          "online_loss": 0.0011565062450245023,
          "train_loss": 0.0013676888775080442,
          "validation_loss": 0.0008949417388066649,
          "train_metric": 0.0001377486187266186,
          "validation_metric": 9.013525414047763e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.088099721000617
        },
        {
          "epoch": 865,
          "updates": 865,
          "samples_seen": 885760,
          "online_loss": 0.0013676888775080442,
          "train_loss": 0.0014571655774489045,
          "validation_loss": 0.0011995971435680985,
          "train_metric": 0.00014676038699690253,
          "validation_metric": 0.0001208190296893008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.095609894000518
        },
        {
          "epoch": 866,
          "updates": 866,
          "samples_seen": 886784,
          "online_loss": 0.0014571654610335827,
          "train_loss": 0.001683729817159474,
          "validation_loss": 0.0010678606340661645,
          "train_metric": 0.00016957911429926753,
          "validation_metric": 0.00010755100811365992,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.103200306000872
        },
        {
          "epoch": 867,
          "updates": 867,
          "samples_seen": 887808,
          "online_loss": 0.001683729817159474,
          "train_loss": 0.0014524662401527166,
          "validation_loss": 0.0011932491324841976,
          "train_metric": 0.0001462870859540999,
          "validation_metric": 0.00012017967674182728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.110452986000382
        },
        {
          "epoch": 868,
          "updates": 868,
          "samples_seen": 888832,
          "online_loss": 0.0014524663565680385,
          "train_loss": 0.0012599823530763388,
          "validation_loss": 0.0008273270796053112,
          "train_metric": 0.00012690080620814115,
          "validation_metric": 8.332534343935549e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.118079644000318
        },
        {
          "epoch": 869,
          "updates": 869,
          "samples_seen": 889856,
          "online_loss": 0.0012599823530763388,
          "train_loss": 0.0009519692393951118,
          "validation_loss": 0.0007821714389137924,
          "train_metric": 9.587885870132595e-05,
          "validation_metric": 7.8777433373034e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.12565688200084
        },
        {
          "epoch": 870,
          "updates": 870,
          "samples_seen": 890880,
          "online_loss": 0.0009519691811874509,
          "train_loss": 0.0008497372036799788,
          "validation_loss": 0.0006672486197203398,
          "train_metric": 8.558241825085133e-05,
          "validation_metric": 6.720283272443339e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.13468802100033
        },
        {
          "epoch": 871,
          "updates": 871,
          "samples_seen": 891904,
          "online_loss": 0.0008497372036799788,
          "train_loss": 0.0009479537256993353,
          "validation_loss": 0.0006820193375460804,
          "train_metric": 9.547442459734157e-05,
          "validation_metric": 6.869048957014456e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.142224088000148
        },
        {
          "epoch": 872,
          "updates": 872,
          "samples_seen": 892928,
          "online_loss": 0.0009479537839069963,
          "train_loss": 0.0011079920222982764,
          "validation_loss": 0.0009225671528838575,
          "train_metric": 0.00011159291170770302,
          "validation_metric": 9.29175948840566e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.150029679
        },
        {
          "epoch": 873,
          "updates": 873,
          "samples_seen": 893952,
          "online_loss": 0.0011079921387135983,
          "train_loss": 0.00125860923435539,
          "validation_loss": 0.0008471477776765823,
          "train_metric": 0.00012676250480581075,
          "validation_metric": 8.532161882612854e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.157667249000951
        },
        {
          "epoch": 874,
          "updates": 874,
          "samples_seen": 894976,
          "online_loss": 0.001258609350770712,
          "train_loss": 0.0011483781272545457,
          "validation_loss": 0.0009610809502191842,
          "train_metric": 0.00011566042667254806,
          "validation_metric": 9.679655340733007e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.165764923000097
        },
        {
          "epoch": 875,
          "updates": 875,
          "samples_seen": 896000,
          "online_loss": 0.0011483780108392239,
          "train_loss": 0.0010066316463053226,
          "validation_loss": 0.0007089154096320271,
          "train_metric": 0.0001013842411339283,
          "validation_metric": 7.139935769373551e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.173333843000364
        },
        {
          "epoch": 876,
          "updates": 876,
          "samples_seen": 897024,
          "online_loss": 0.0010066315298900008,
          "train_loss": 0.0008611077209934592,
          "validation_loss": 0.0007037981413304806,
          "train_metric": 8.672763215145096e-05,
          "validation_metric": 7.088397978805006e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.181693043999985
        },
        {
          "epoch": 877,
          "updates": 877,
          "samples_seen": 898048,
          "online_loss": 0.0008611077792011201,
          "train_loss": 0.0008436394855380058,
          "validation_loss": 0.0006798666436225176,
          "train_metric": 8.496828377246857e-05,
          "validation_metric": 6.847368058515713e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.1891033410011
        },
        {
          "epoch": 878,
          "updates": 878,
          "samples_seen": 899072,
          "online_loss": 0.0008436395437456667,
          "train_loss": 0.0009280474623665214,
          "validation_loss": 0.0006645897519774735,
          "train_metric": 9.346954175271094e-05,
          "validation_metric": 6.693504838040099e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.197781967000992
        },
        {
          "epoch": 879,
          "updates": 879,
          "samples_seen": 900096,
          "online_loss": 0.0009280475205741823,
          "train_loss": 0.0010001023765653372,
          "validation_loss": 0.0008366559050045907,
          "train_metric": 0.00010072665463667363,
          "validation_metric": 8.426490967394784e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.205543973001113
        },
        {
          "epoch": 880,
          "updates": 880,
          "samples_seen": 901120,
          "online_loss": 0.001000102492980659,
          "train_loss": 0.0010198407107964158,
          "validation_loss": 0.0007096895133145154,
          "train_metric": 0.0001027146281558089,
          "validation_metric": 7.147734140744433e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.213776522001353
        },
        {
          "epoch": 881,
          "updates": 881,
          "samples_seen": 902144,
          "online_loss": 0.0010198408272117376,
          "train_loss": 0.0009269552538171411,
          "validation_loss": 0.000770614598877728,
          "train_metric": 9.335954382549971e-05,
          "validation_metric": 7.761348388157785e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.22190497200063
        },
        {
          "epoch": 882,
          "updates": 882,
          "samples_seen": 903168,
          "online_loss": 0.0009269553702324629,
          "train_loss": 0.0008469929452985525,
          "validation_loss": 0.0006457192357629538,
          "train_metric": 8.530601189704612e-05,
          "validation_metric": 6.503447366412729e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.231808102000286
        },
        {
          "epoch": 883,
          "updates": 883,
          "samples_seen": 904192,
          "online_loss": 0.0008469927706755698,
          "train_loss": 0.0008231452666223049,
          "validation_loss": 0.0006494449335150421,
          "train_metric": 8.29041819088161e-05,
          "validation_metric": 6.540971662616357e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.23969590300112
        },
        {
          "epoch": 884,
          "updates": 884,
          "samples_seen": 905216,
          "online_loss": 0.0008231452666223049,
          "train_loss": 0.0008590120123699307,
          "validation_loss": 0.0007092936430126429,
          "train_metric": 8.651654206914827e-05,
          "validation_metric": 7.143745460780337e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.247626891001346
        },
        {
          "epoch": 885,
          "updates": 885,
          "samples_seen": 906240,
          "online_loss": 0.0008590120705775917,
          "train_loss": 0.0009129978134296834,
          "validation_loss": 0.0006702881073579192,
          "train_metric": 9.195379971060902e-05,
          "validation_metric": 6.750896136509255e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.25517472500178
        },
        {
          "epoch": 886,
          "updates": 886,
          "samples_seen": 907264,
          "online_loss": 0.0009129978134296834,
          "train_loss": 0.0009200727799907327,
          "validation_loss": 0.0007704133167862892,
          "train_metric": 9.266636334359646e-05,
          "validation_metric": 7.759321306366473e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.262912867001432
        },
        {
          "epoch": 887,
          "updates": 887,
          "samples_seen": 908288,
          "online_loss": 0.0009200727799907327,
          "train_loss": 0.0008948168833740056,
          "validation_loss": 0.0006554453284479678,
          "train_metric": 9.012268128572032e-05,
          "validation_metric": 6.601405038964003e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.270465823999984
        },
        {
          "epoch": 888,
          "updates": 888,
          "samples_seen": 909312,
          "online_loss": 0.0008948167669586837,
          "train_loss": 0.0008420287631452084,
          "validation_loss": 0.0006931263487786055,
          "train_metric": 8.480605902150273e-05,
          "validation_metric": 6.980915350141004e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.278019833000144
        },
        {
          "epoch": 889,
          "updates": 889,
          "samples_seen": 910336,
          "online_loss": 0.0008420287631452084,
          "train_loss": 0.0008119730046018958,
          "validation_loss": 0.0006360951811075211,
          "train_metric": 8.17789405118674e-05,
          "validation_metric": 6.406517059076577e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.286448659000598
        },
        {
          "epoch": 890,
          "updates": 890,
          "samples_seen": 911360,
          "online_loss": 0.0008119730046018958,
          "train_loss": 0.0008162672747857869,
          "validation_loss": 0.0006301674293354154,
          "train_metric": 8.221144526032731e-05,
          "validation_metric": 6.346815644064918e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.295027944001049
        },
        {
          "epoch": 891,
          "updates": 891,
          "samples_seen": 912384,
          "online_loss": 0.0008162672747857869,
          "train_loss": 0.0008403171086683869,
          "validation_loss": 0.0006925458437763155,
          "train_metric": 8.463365520583466e-05,
          "validation_metric": 6.975068390602246e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.302690303000418
        },
        {
          "epoch": 892,
          "updates": 892,
          "samples_seen": 913408,
          "online_loss": 0.0008403169922530651,
          "train_loss": 0.0008610229124315083,
          "validation_loss": 0.0006358936079777777,
          "train_metric": 8.671907562529668e-05,
          "validation_metric": 6.404487794497982e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.310290494000583
        },
        {
          "epoch": 893,
          "updates": 893,
          "samples_seen": 914432,
          "online_loss": 0.0008610228542238474,
          "train_loss": 0.0008542059222236276,
          "validation_loss": 0.0007074377499520779,
          "train_metric": 8.60325017129071e-05,
          "validation_metric": 7.125054253265262e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.318166924000252
        },
        {
          "epoch": 894,
          "updates": 894,
          "samples_seen": 915456,
          "online_loss": 0.0008542059804312885,
          "train_loss": 0.0008339945925399661,
          "validation_loss": 0.0006294915219768882,
          "train_metric": 8.399687067139894e-05,
          "validation_metric": 6.340007530525327e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.327047254000718
        },
        {
          "epoch": 895,
          "updates": 895,
          "samples_seen": 916480,
          "online_loss": 0.0008339944761246443,
          "train_loss": 0.0008089957409538329,
          "validation_loss": 0.0006567153614014387,
          "train_metric": 8.147908374667168e-05,
          "validation_metric": 6.614196172449738e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.33475964200079
        },
        {
          "epoch": 896,
          "updates": 896,
          "samples_seen": 917504,
          "online_loss": 0.000808995682746172,
          "train_loss": 0.0007971076411195099,
          "validation_loss": 0.0006318636005744338,
          "train_metric": 8.028175216168165e-05,
          "validation_metric": 6.363898137351498e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.437750787001278
        },
        {
          "epoch": 897,
          "updates": 897,
          "samples_seen": 918528,
          "online_loss": 0.0007971076411195099,
          "train_loss": 0.000800608191639185,
          "validation_loss": 0.000623608473688364,
          "train_metric": 8.063433051574975e-05,
          "validation_metric": 6.280756497289985e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.44662790700022
        },
        {
          "epoch": 898,
          "updates": 898,
          "samples_seen": 919552,
          "online_loss": 0.000800608191639185,
          "train_loss": 0.0008115301607176661,
          "validation_loss": 0.0006612673169001937,
          "train_metric": 8.173434616765007e-05,
          "validation_metric": 6.660041253780946e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.454589094000767
        },
        {
          "epoch": 899,
          "updates": 899,
          "samples_seen": 920576,
          "online_loss": 0.0008115301607176661,
          "train_loss": 0.0008203480392694473,
          "validation_loss": 0.0006194963352754712,
          "train_metric": 8.262245682999492e-05,
          "validation_metric": 6.239341018954292e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.463123763000112
        },
        {
          "epoch": 900,
          "updates": 900,
          "samples_seen": 921600,
          "online_loss": 0.0008203479810617864,
          "train_loss": 0.0008174561662599444,
          "validation_loss": 0.0006689684232696891,
          "train_metric": 8.233119297074154e-05,
          "validation_metric": 6.737604417139664e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.470907164000891
        },
        {
          "epoch": 901,
          "updates": 901,
          "samples_seen": 922624,
          "online_loss": 0.0008174561662599444,
          "train_loss": 0.0008077211095951498,
          "validation_loss": 0.0006138905300758779,
          "train_metric": 8.13506921986118e-05,
          "validation_metric": 6.182881043059751e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.479409524001312
        },
        {
          "epoch": 902,
          "updates": 902,
          "samples_seen": 923648,
          "online_loss": 0.0008077209931798279,
          "train_loss": 0.0007941419607959688,
          "validation_loss": 0.0006414054078049958,
          "train_metric": 7.998306682566181e-05,
          "validation_metric": 6.460000440711156e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.48722819500108
        },
        {
          "epoch": 903,
          "updates": 903,
          "samples_seen": 924672,
          "online_loss": 0.0007941419607959688,
          "train_loss": 0.0007849346147850156,
          "validation_loss": 0.0006156045710667968,
          "train_metric": 7.905574602773413e-05,
          "validation_metric": 6.200143980095163e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.495310586000414
        },
        {
          "epoch": 904,
          "updates": 904,
          "samples_seen": 925696,
          "online_loss": 0.0007849344983696938,
          "train_loss": 0.0007824752829037607,
          "validation_loss": 0.000615886936429888,
          "train_metric": 7.880805060267448e-05,
          "validation_metric": 6.202987424330786e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.502991202001795
        },
        {
          "epoch": 905,
          "updates": 905,
          "samples_seen": 926720,
          "online_loss": 0.0007824753993190825,
          "train_loss": 0.0007852755952626467,
          "validation_loss": 0.0006329832831397653,
          "train_metric": 7.909007399575785e-05,
          "validation_metric": 6.375175871653482e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.51101583900163
        },
        {
          "epoch": 906,
          "updates": 906,
          "samples_seen": 927744,
          "online_loss": 0.0007852755952626467,
          "train_loss": 0.0007897132309153676,
          "validation_loss": 0.0006089680246077478,
          "train_metric": 7.953702152008191e-05,
          "validation_metric": 6.133304123068228e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.519836620000206
        },
        {
          "epoch": 907,
          "updates": 907,
          "samples_seen": 928768,
          "online_loss": 0.0007897131727077067,
          "train_loss": 0.0007914191228337586,
          "validation_loss": 0.0006446284241974354,
          "train_metric": 7.970884325914085e-05,
          "validation_metric": 6.492461398011073e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.527674202001435
        },
        {
          "epoch": 908,
          "updates": 908,
          "samples_seen": 929792,
          "online_loss": 0.0007914191228337586,
          "train_loss": 0.0007899187621660531,
          "validation_loss": 0.000606082146987319,
          "train_metric": 7.955772889545187e-05,
          "validation_metric": 6.104237399995327e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.536094149001656
        },
        {
          "epoch": 909,
          "updates": 909,
          "samples_seen": 930816,
          "online_loss": 0.000789918820373714,
          "train_loss": 0.0007842397317290306,
          "validation_loss": 0.0006360754487104714,
          "train_metric": 7.898575131548569e-05,
          "validation_metric": 6.40631842543371e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.543655680001393
        },
        {
          "epoch": 910,
          "updates": 910,
          "samples_seen": 931840,
          "online_loss": 0.0007842396153137088,
          "train_loss": 0.0007776214624755085,
          "validation_loss": 0.000602449057623744,
          "train_metric": 7.831918628653511e-05,
          "validation_metric": 6.067645881557837e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.55245481700149
        },
        {
          "epoch": 911,
          "updates": 911,
          "samples_seen": 932864,
          "online_loss": 0.0007776214042678475,
          "train_loss": 0.0007714284583926201,
          "validation_loss": 0.0006175193120725453,
          "train_metric": 7.769545482005924e-05,
          "validation_metric": 6.219428178155795e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.560063270000683
        },
        {
          "epoch": 912,
          "updates": 912,
          "samples_seen": 933888,
          "online_loss": 0.0007714284583926201,
          "train_loss": 0.0007673747604712844,
          "validation_loss": 0.0006047316128388047,
          "train_metric": 7.728717173449695e-05,
          "validation_metric": 6.090635361033492e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.567485545001546
        },
        {
          "epoch": 913,
          "updates": 913,
          "samples_seen": 934912,
          "online_loss": 0.0007673747604712844,
          "train_loss": 0.0007656480884179473,
          "validation_loss": 0.0006027659401297569,
          "train_metric": 7.711327634751797e-05,
          "validation_metric": 6.070838571758941e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.575062697000249
        },
        {
          "epoch": 914,
          "updates": 914,
          "samples_seen": 935936,
          "online_loss": 0.0007656480302102864,
          "train_loss": 0.00076558708678931,
          "validation_loss": 0.0006122006452642381,
          "train_metric": 7.710714271524921e-05,
          "validation_metric": 6.165860395412892e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.583146162000048
        },
        {
          "epoch": 915,
          "updates": 915,
          "samples_seen": 936960,
          "online_loss": 0.00076558708678931,
          "train_loss": 0.0007661862182430923,
          "validation_loss": 0.0005957770626991987,
          "train_metric": 7.716747495578602e-05,
          "validation_metric": 6.000448047416285e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.591593917000864
        },
        {
          "epoch": 916,
          "updates": 916,
          "samples_seen": 937984,
          "online_loss": 0.000766186392866075,
          "train_loss": 0.0007662898860871792,
          "validation_loss": 0.0006179222837090492,
          "train_metric": 7.717791595496237e-05,
          "validation_metric": 6.223486707312986e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.599902130001283
        },
        {
          "epoch": 917,
          "updates": 917,
          "samples_seen": 939008,
          "online_loss": 0.0007662898860871792,
          "train_loss": 0.0007656533853150904,
          "validation_loss": 0.0005927462480030954,
          "train_metric": 7.711380021646619e-05,
          "validation_metric": 5.9699239500332624e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.60837968800115
        },
        {
          "epoch": 918,
          "updates": 918,
          "samples_seen": 940032,
          "online_loss": 0.0007656533271074295,
          "train_loss": 0.0007636320660822093,
          "validation_loss": 0.0006166725070215762,
          "train_metric": 7.691023347433656e-05,
          "validation_metric": 6.21090002823621e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.616386506000708
        },
        {
          "epoch": 919,
          "updates": 919,
          "samples_seen": 941056,
          "online_loss": 0.0007636320078745484,
          "train_loss": 0.0007609737222082913,
          "validation_loss": 0.0005898843519389629,
          "train_metric": 7.664248551009223e-05,
          "validation_metric": 5.9410991525510326e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.624686759001634
        },
        {
          "epoch": 920,
          "updates": 920,
          "samples_seen": 942080,
          "online_loss": 0.0007609736640006304,
          "train_loss": 0.0007575752679258585,
          "validation_loss": 0.0006093693664297462,
          "train_metric": 7.630020991200581e-05,
          "validation_metric": 6.137344462331384e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.632517361000282
        },
        {
          "epoch": 921,
          "updates": 921,
          "samples_seen": 943104,
          "online_loss": 0.0007575752679258585,
          "train_loss": 0.000754204229451716,
          "validation_loss": 0.000587794347666204,
          "train_metric": 7.596069190185517e-05,
          "validation_metric": 5.920049443375319e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.640971954000634
        },
        {
          "epoch": 922,
          "updates": 922,
          "samples_seen": 944128,
          "online_loss": 0.0007542042876593769,
          "train_loss": 0.0007509430288337171,
          "validation_loss": 0.0006002910085953772,
          "train_metric": 7.563223334727809e-05,
          "validation_metric": 6.045911504770629e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.649167262999981
        },
        {
          "epoch": 923,
          "updates": 923,
          "samples_seen": 945152,
          "online_loss": 0.0007509429706260562,
          "train_loss": 0.0007481007487513125,
          "validation_loss": 0.0005874696071259677,
          "train_metric": 7.534596079494804e-05,
          "validation_metric": 5.9167781728319824e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.657689277000827
        },
        {
          "epoch": 924,
          "updates": 924,
          "samples_seen": 946176,
          "online_loss": 0.0007481006905436516,
          "train_loss": 0.000745686877053231,
          "validation_loss": 0.0005922058480791748,
          "train_metric": 7.510286377510056e-05,
          "validation_metric": 5.96448007854633e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.665731126000537
        },
        {
          "epoch": 925,
          "updates": 925,
          "samples_seen": 947200,
          "online_loss": 0.0007456869352608919,
          "train_loss": 0.0007436861633323133,
          "validation_loss": 0.0005885413847863674,
          "train_metric": 7.49013343011029e-05,
          "validation_metric": 5.927573511144146e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.673303800000213
        },
        {
          "epoch": 926,
          "updates": 926,
          "samples_seen": 948224,
          "online_loss": 0.0007436860469169915,
          "train_loss": 0.0007420204346999526,
          "validation_loss": 0.0005862450925633311,
          "train_metric": 7.47335798223503e-05,
          "validation_metric": 5.904446152271703e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.682094486001006
        },
        {
          "epoch": 927,
          "updates": 927,
          "samples_seen": 949248,
          "online_loss": 0.0007420203764922917,
          "train_loss": 0.0007405999349430203,
          "validation_loss": 0.0005900724790990353,
          "train_metric": 7.45905126677826e-05,
          "validation_metric": 5.9429945395095274e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.690127100000609
        },
        {
          "epoch": 928,
          "updates": 928,
          "samples_seen": 950272,
          "online_loss": 0.0007405999349430203,
          "train_loss": 0.0007393354899249971,
          "validation_loss": 0.0005816345801576972,
          "train_metric": 7.446316885761917e-05,
          "validation_metric": 5.858010990777984e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.698775327000476
        },
        {
          "epoch": 929,
          "updates": 929,
          "samples_seen": 951296,
          "online_loss": 0.0007393354899249971,
          "train_loss": 0.000738160393666476,
          "validation_loss": 0.0005908160819672048,
          "train_metric": 7.434482540702447e-05,
          "validation_metric": 5.9504836826818064e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.706730496000091
        },
        {
          "epoch": 930,
          "updates": 930,
          "samples_seen": 952320,
          "online_loss": 0.000738160393666476,
          "train_loss": 0.000737086171284318,
          "validation_loss": 0.0005772592267021537,
          "train_metric": 7.423662464134395e-05,
          "validation_metric": 5.8139437896898016e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.715484155000013
        },
        {
          "epoch": 931,
          "updates": 931,
          "samples_seen": 953344,
          "online_loss": 0.0007370862294919789,
          "train_loss": 0.0007360572926700115,
          "validation_loss": 0.0005909715546295047,
          "train_metric": 7.413299317704514e-05,
          "validation_metric": 5.9520494687603787e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.724930193000546
        },
        {
          "epoch": 932,
          "updates": 932,
          "samples_seen": 954368,
          "online_loss": 0.0007360572926700115,
          "train_loss": 0.0007352177053689957,
          "validation_loss": 0.0005731600685976446,
          "train_metric": 7.404842472169548e-05,
          "validation_metric": 5.772658187197521e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.739312972000334
        },
        {
          "epoch": 933,
          "updates": 933,
          "samples_seen": 955392,
          "online_loss": 0.0007352177635766566,
          "train_loss": 0.0007344633340835571,
          "validation_loss": 0.0005914487410336733,
          "train_metric": 7.397244917228818e-05,
          "validation_metric": 5.9568545111687854e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.753135969000141
        },
        {
          "epoch": 934,
          "updates": 934,
          "samples_seen": 956416,
          "online_loss": 0.0007344633340835571,
          "train_loss": 0.000734131783246994,
          "validation_loss": 0.000569400261156261,
          "train_metric": 7.393905980279669e-05,
          "validation_metric": 5.7347908295923844e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.882255393000378
        },
        {
          "epoch": 935,
          "updates": 935,
          "samples_seen": 957440,
          "online_loss": 0.000734131783246994,
          "train_loss": 0.0007340317824855447,
          "validation_loss": 0.0005933118518441916,
          "train_metric": 7.392900442937389e-05,
          "validation_metric": 5.9756192058557644e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.890176408000116
        },
        {
          "epoch": 936,
          "updates": 936,
          "samples_seen": 958464,
          "online_loss": 0.0007340318989008665,
          "train_loss": 0.0007348862127400935,
          "validation_loss": 0.000566351693123579,
          "train_metric": 7.4015035352204e-05,
          "validation_metric": 5.704086652258411e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.89899938400049
        },
        {
          "epoch": 937,
          "updates": 937,
          "samples_seen": 959488,
          "online_loss": 0.0007348861545324326,
          "train_loss": 0.0007363254553638399,
          "validation_loss": 0.0005984459421597421,
          "train_metric": 7.416000153170899e-05,
          "validation_metric": 6.027328345226124e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.907153542000742
        },
        {
          "epoch": 938,
          "updates": 938,
          "samples_seen": 960512,
          "online_loss": 0.0007363254553638399,
          "train_loss": 0.000739985320251435,
          "validation_loss": 0.0005646100617013872,
          "train_metric": 7.452861609635875e-05,
          "validation_metric": 5.686545046046376e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.916104810001343
        },
        {
          "epoch": 939,
          "updates": 939,
          "samples_seen": 961536,
          "online_loss": 0.0007399850874207914,
          "train_loss": 0.0007449895492754877,
          "validation_loss": 0.0006098579615354538,
          "train_metric": 7.503262895625085e-05,
          "validation_metric": 6.14226664765738e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.924101486001746
        },
        {
          "epoch": 940,
          "updates": 940,
          "samples_seen": 962560,
          "online_loss": 0.0007449895492754877,
          "train_loss": 0.0007554253679700196,
          "validation_loss": 0.0005663487827405334,
          "train_metric": 7.60836701374501e-05,
          "validation_metric": 5.7040571846300736e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.932383217001188
        },
        {
          "epoch": 941,
          "updates": 941,
          "samples_seen": 963584,
          "online_loss": 0.0007554253679700196,
          "train_loss": 0.0007685967721045017,
          "validation_loss": 0.0006345350411720574,
          "train_metric": 7.741025910945609e-05,
          "validation_metric": 6.390804628608748e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.94230927100034
        },
        {
          "epoch": 942,
          "updates": 942,
          "samples_seen": 964608,
          "online_loss": 0.0007685968303121626,
          "train_loss": 0.0007955303881317377,
          "validation_loss": 0.0005792609881609678,
          "train_metric": 8.01229034550488e-05,
          "validation_metric": 5.834104376845062e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.951688990000548
        },
        {
          "epoch": 943,
          "updates": 943,
          "samples_seen": 965632,
          "online_loss": 0.0007955303299240768,
          "train_loss": 0.0008258366724476218,
          "validation_loss": 0.0006866189069114625,
          "train_metric": 8.31752477097325e-05,
          "validation_metric": 6.91537352395244e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.958912861000499
        },
        {
          "epoch": 944,
          "updates": 944,
          "samples_seen": 966656,
          "online_loss": 0.0008258367306552827,
          "train_loss": 0.0008900795364752412,
          "validation_loss": 0.000621283776126802,
          "train_metric": 8.964556036517024e-05,
          "validation_metric": 6.257343193283305e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.966398154001581
        },
        {
          "epoch": 945,
          "updates": 945,
          "samples_seen": 967680,
          "online_loss": 0.000890079652890563,
          "train_loss": 0.0009461540612392128,
          "validation_loss": 0.0007859401521272957,
          "train_metric": 9.529318049317226e-05,
          "validation_metric": 7.915700552985072e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.973840079001093
        },
        {
          "epoch": 946,
          "updates": 946,
          "samples_seen": 968704,
          "online_loss": 0.0009461541776545346,
          "train_loss": 0.0010735040996223688,
          "validation_loss": 0.0007161048706620932,
          "train_metric": 0.00010811939864652231,
          "validation_metric": 7.212346827145666e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.981655343000966
        },
        {
          "epoch": 947,
          "updates": 947,
          "samples_seen": 969728,
          "online_loss": 0.0010735043324530125,
          "train_loss": 0.0011237392900511622,
          "validation_loss": 0.0009234369499608874,
          "train_metric": 0.0001131789103965275,
          "validation_metric": 9.300518286181614e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.989518593001776
        },
        {
          "epoch": 948,
          "updates": 948,
          "samples_seen": 970752,
          "online_loss": 0.0011237392900511622,
          "train_loss": 0.0012673642486333847,
          "validation_loss": 0.0008226206409744918,
          "train_metric": 0.00012764427810907364,
          "validation_metric": 8.28513439046219e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.997280661000332
        },
        {
          "epoch": 949,
          "updates": 949,
          "samples_seen": 971776,
          "online_loss": 0.0012673641322180629,
          "train_loss": 0.0011823745444417,
          "validation_loss": 0.000965452054515481,
          "train_metric": 0.00011908443411812186,
          "validation_metric": 9.723679977469146e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.004922181000438
        },
        {
          "epoch": 950,
          "updates": 950,
          "samples_seen": 972800,
          "online_loss": 0.0011823746608570218,
          "train_loss": 0.0011482968693599105,
          "validation_loss": 0.0007525308174081147,
          "train_metric": 0.00011565226304810494,
          "validation_metric": 7.579215161968023e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.01290557900029
        },
        {
          "epoch": 951,
          "updates": 951,
          "samples_seen": 973824,
          "online_loss": 0.0011482968693599105,
          "train_loss": 0.0009332078043371439,
          "validation_loss": 0.0007684349548071623,
          "train_metric": 9.398927795700729e-05,
          "validation_metric": 7.739394641248509e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.020636197001295
        },
        {
          "epoch": 952,
          "updates": 952,
          "samples_seen": 974848,
          "online_loss": 0.0009332078625448048,
          "train_loss": 0.0007859008619561791,
          "validation_loss": 0.0005647771758958697,
          "train_metric": 7.915306196082383e-05,
          "validation_metric": 5.688228702638298e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.028857262001111
        },
        {
          "epoch": 953,
          "updates": 953,
          "samples_seen": 975872,
          "online_loss": 0.0007859007455408573,
          "train_loss": 0.0007035542512312531,
          "validation_loss": 0.0005616492708213627,
          "train_metric": 7.085940887918696e-05,
          "validation_metric": 5.656725988956168e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.037551807001364
        },
        {
          "epoch": 954,
          "updates": 954,
          "samples_seen": 976896,
          "online_loss": 0.0007035543094389141,
          "train_loss": 0.000716482347343117,
          "validation_loss": 0.0005827420391142368,
          "train_metric": 7.216147787403315e-05,
          "validation_metric": 5.869163942406885e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.045331600000281
        },
        {
          "epoch": 955,
          "updates": 955,
          "samples_seen": 977920,
          "online_loss": 0.000716482347343117,
          "train_loss": 0.0007962165982462466,
          "validation_loss": 0.0005783130181953311,
          "train_metric": 8.019201050046831e-05,
          "validation_metric": 5.82455686526373e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.053003211000032
        },
        {
          "epoch": 956,
          "updates": 956,
          "samples_seen": 978944,
          "online_loss": 0.0007962164236232638,
          "train_loss": 0.0008774519665166736,
          "validation_loss": 0.0007324303151108325,
          "train_metric": 8.837375207804143e-05,
          "validation_metric": 7.376769644906744e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.06082129600145
        },
        {
          "epoch": 957,
          "updates": 957,
          "samples_seen": 979968,
          "online_loss": 0.0008774517918936908,
          "train_loss": 0.0009626243263483047,
          "validation_loss": 0.0006611835560761392,
          "train_metric": 9.695199696579948e-05,
          "validation_metric": 6.659199425484985e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.06898990100126
        },
        {
          "epoch": 958,
          "updates": 958,
          "samples_seen": 980992,
          "online_loss": 0.0009626244427636266,
          "train_loss": 0.0009341123513877392,
          "validation_loss": 0.0007791662937961519,
          "train_metric": 9.408038022229448e-05,
          "validation_metric": 7.847477536415681e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.076565266001126
        },
        {
          "epoch": 959,
          "updates": 959,
          "samples_seen": 982016,
          "online_loss": 0.0009341122931800783,
          "train_loss": 0.0008961156709119678,
          "validation_loss": 0.0006168427062220871,
          "train_metric": 9.02534811757505e-05,
          "validation_metric": 6.21261351625435e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.084166289001587
        },
        {
          "epoch": 960,
          "updates": 960,
          "samples_seen": 983040,
          "online_loss": 0.0008961155544966459,
          "train_loss": 0.0007894502487033606,
          "validation_loss": 0.0006548339733853936,
          "train_metric": 7.951053703436628e-05,
          "validation_metric": 6.595248851226643e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.091606894000506
        },
        {
          "epoch": 961,
          "updates": 961,
          "samples_seen": 984064,
          "online_loss": 0.0007894502487033606,
          "train_loss": 0.0007150525925680995,
          "validation_loss": 0.0005347655387595296,
          "train_metric": 7.201747939689085e-05,
          "validation_metric": 5.3859621402807534e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.10025166700143
        },
        {
          "epoch": 962,
          "updates": 962,
          "samples_seen": 985088,
          "online_loss": 0.0007150524761527777,
          "train_loss": 0.0006860254216007888,
          "validation_loss": 0.0005399042856879532,
          "train_metric": 6.909396324772388e-05,
          "validation_metric": 5.4377178457798436e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.107942715001627
        },
        {
          "epoch": 963,
          "updates": 963,
          "samples_seen": 986112,
          "online_loss": 0.0006860254798084497,
          "train_loss": 0.000706770340912044,
          "validation_loss": 0.000575842394027859,
          "train_metric": 7.118331996025518e-05,
          "validation_metric": 5.7996738178189844e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.116146096001103
        },
        {
          "epoch": 964,
          "updates": 964,
          "samples_seen": 987136,
          "online_loss": 0.000706770399119705,
          "train_loss": 0.0007552370079793036,
          "validation_loss": 0.000557074963580817,
          "train_metric": 7.606470899190754e-05,
          "validation_metric": 5.610656080534682e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.123660875000496
        },
        {
          "epoch": 965,
          "updates": 965,
          "samples_seen": 988160,
          "online_loss": 0.0007552370079793036,
          "train_loss": 0.0007900033378973603,
          "validation_loss": 0.0006572356214746833,
          "train_metric": 7.95662563177757e-05,
          "validation_metric": 6.619435589527711e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.131433615000788
        },
        {
          "epoch": 966,
          "updates": 966,
          "samples_seen": 989184,
          "online_loss": 0.0007900033961050212,
          "train_loss": 0.0008129992638714612,
          "validation_loss": 0.000588933820836246,
          "train_metric": 8.188229548977688e-05,
          "validation_metric": 5.931525811320171e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.139494191000267
        },
        {
          "epoch": 967,
          "updates": 967,
          "samples_seen": 990208,
          "online_loss": 0.0008129992056638002,
          "train_loss": 0.0007851410191506147,
          "validation_loss": 0.0006557408487424254,
          "train_metric": 7.907652616268024e-05,
          "validation_metric": 6.604380905628204e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.147145894001369
        },
        {
          "epoch": 968,
          "updates": 968,
          "samples_seen": 991232,
          "online_loss": 0.0007851409609429538,
          "train_loss": 0.0007513494929298759,
          "validation_loss": 0.0005539866979233921,
          "train_metric": 7.567316788481548e-05,
          "validation_metric": 5.579551725531928e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.154871590000766
        },
        {
          "epoch": 969,
          "updates": 969,
          "samples_seen": 992256,
          "online_loss": 0.0007513494929298759,
          "train_loss": 0.000706662074662745,
          "validation_loss": 0.0005828255088999867,
          "train_metric": 7.11724060238339e-05,
          "validation_metric": 5.8700054069049656e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.16264172800038
        },
        {
          "epoch": 970,
          "updates": 970,
          "samples_seen": 993280,
          "online_loss": 0.0007066619582474232,
          "train_loss": 0.0006794193177483976,
          "validation_loss": 0.0005269600078463554,
          "train_metric": 6.84286278556101e-05,
          "validation_metric": 5.307348328642547e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.17135527100072
        },
        {
          "epoch": 971,
          "updates": 971,
          "samples_seen": 994304,
          "online_loss": 0.0006794193759560585,
          "train_loss": 0.0006731379544362426,
          "validation_loss": 0.000528636563103646,
          "train_metric": 6.779599061701447e-05,
          "validation_metric": 5.3242336434777826e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.179060906000814
        },
        {
          "epoch": 972,
          "updates": 972,
          "samples_seen": 995328,
          "online_loss": 0.0006731379544362426,
          "train_loss": 0.000684308004565537,
          "validation_loss": 0.0005562677979469299,
          "train_metric": 6.892098463140428e-05,
          "validation_metric": 5.602525197900832e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.186584121000124
        },
        {
          "epoch": 973,
          "updates": 973,
          "samples_seen": 996352,
          "online_loss": 0.0006843079463578761,
          "train_loss": 0.0007034307345747948,
          "validation_loss": 0.0005270064575597644,
          "train_metric": 7.084695971570909e-05,
          "validation_metric": 5.307815445121378e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.194752227000208
        },
        {
          "epoch": 974,
          "updates": 974,
          "samples_seen": 997376,
          "online_loss": 0.0007034306763671339,
          "train_loss": 0.0007162638939917088,
          "validation_loss": 0.0005920497933402658,
          "train_metric": 7.213947537820786e-05,
          "validation_metric": 5.9629088354995474e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.202889064001283
        },
        {
          "epoch": 975,
          "updates": 975,
          "samples_seen": 998400,
          "online_loss": 0.0007162638939917088,
          "train_loss": 0.0007232408970594406,
          "validation_loss": 0.0005375344189815223,
          "train_metric": 7.284218008862808e-05,
          "validation_metric": 5.413849794422276e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.308408067001437
        },
        {
          "epoch": 976,
          "updates": 976,
          "samples_seen": 999424,
          "online_loss": 0.0007232410134747624,
          "train_loss": 0.0007135169580578804,
          "validation_loss": 0.0005896688671782613,
          "train_metric": 7.186282164184377e-05,
          "validation_metric": 5.938928734394722e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.31651591300033
        },
        {
          "epoch": 977,
          "updates": 977,
          "samples_seen": 1000448,
          "online_loss": 0.0007135170162655413,
          "train_loss": 0.0007004098151810467,
          "validation_loss": 0.0005299310432747006,
          "train_metric": 7.05427082721144e-05,
          "validation_metric": 5.337272159522399e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.32414826600143
        },
        {
          "epoch": 978,
          "updates": 978,
          "samples_seen": 1001472,
          "online_loss": 0.0007004096987657249,
          "train_loss": 0.0006821553688496351,
          "validation_loss": 0.0005576400435529649,
          "train_metric": 6.870419019833207e-05,
          "validation_metric": 5.616345879388973e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.331892662001337
        },
        {
          "epoch": 979,
          "updates": 979,
          "samples_seen": 1002496,
          "online_loss": 0.0006821553106419742,
          "train_loss": 0.0006682903040200472,
          "validation_loss": 0.0005191012169234455,
          "train_metric": 6.730775930918753e-05,
          "validation_metric": 5.2281971875345334e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.340398787000595
        },
        {
          "epoch": 980,
          "updates": 980,
          "samples_seen": 1003520,
          "online_loss": 0.0006682903040200472,
          "train_loss": 0.0006601907080039382,
          "validation_loss": 0.0005271228146739304,
          "train_metric": 6.64919862174429e-05,
          "validation_metric": 5.3089879656909034e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.348508273000334
        },
        {
          "epoch": 981,
          "updates": 981,
          "samples_seen": 1004544,
          "online_loss": 0.0006601905915886164,
          "train_loss": 0.0006586291128769517,
          "validation_loss": 0.0005249512032605708,
          "train_metric": 6.633471639361233e-05,
          "validation_metric": 5.287116437102668e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.356119682001008
        },
        {
          "epoch": 982,
          "updates": 982,
          "samples_seen": 1005568,
          "online_loss": 0.0006586290546692908,
          "train_loss": 0.0006618484039790928,
          "validation_loss": 0.0005139770219102502,
          "train_metric": 6.665895489277318e-05,
          "validation_metric": 5.1765888201771304e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.364514296001289
        },
        {
          "epoch": 983,
          "updates": 983,
          "samples_seen": 1006592,
          "online_loss": 0.0006618483457714319,
          "train_loss": 0.0006669853464700282,
          "validation_loss": 0.0005413395119830966,
          "train_metric": 6.71763118589297e-05,
          "validation_metric": 5.4521733545698225e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.372144856000887
        },
        {
          "epoch": 984,
          "updates": 984,
          "samples_seen": 1007616,
          "online_loss": 0.0006669852882623672,
          "train_loss": 0.0006723259575664997,
          "validation_loss": 0.0005132388323545456,
          "train_metric": 6.771420885343105e-05,
          "validation_metric": 5.1691535190911964e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.381091654000556
        },
        {
          "epoch": 985,
          "updates": 985,
          "samples_seen": 1008640,
          "online_loss": 0.0006723259575664997,
          "train_loss": 0.0006747024017386138,
          "validation_loss": 0.0005524601438082755,
          "train_metric": 6.795356603106484e-05,
          "validation_metric": 5.564176171901636e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.388968120001664
        },
        {
          "epoch": 986,
          "updates": 986,
          "samples_seen": 1009664,
          "online_loss": 0.0006747024017386138,
          "train_loss": 0.000675826973747462,
          "validation_loss": 0.0005143481539562345,
          "train_metric": 6.806680903537199e-05,
          "validation_metric": 5.180325752007775e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.396664002000762
        },
        {
          "epoch": 987,
          "updates": 987,
          "samples_seen": 1010688,
          "online_loss": 0.000675826973747462,
          "train_loss": 0.0006728560547344387,
          "validation_loss": 0.0005514714284799993,
          "train_metric": 6.776759983040392e-05,
          "validation_metric": 5.554218660108745e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.404956117001348
        },
        {
          "epoch": 988,
          "updates": 988,
          "samples_seen": 1011712,
          "online_loss": 0.0006728560547344387,
          "train_loss": 0.0006693972973152995,
          "validation_loss": 0.00051166518824175,
          "train_metric": 6.741924153175205e-05,
          "validation_metric": 5.153305392013863e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.413410901001043
        },
        {
          "epoch": 989,
          "updates": 989,
          "samples_seen": 1012736,
          "online_loss": 0.0006693972391076386,
          "train_loss": 0.0006635253666900098,
          "validation_loss": 0.0005410865996964276,
          "train_metric": 6.682784442091361e-05,
          "validation_metric": 5.449626041809097e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.421044433000134
        },
        {
          "epoch": 990,
          "updates": 990,
          "samples_seen": 1013760,
          "online_loss": 0.0006635253666900098,
          "train_loss": 0.000658184930216521,
          "validation_loss": 0.0005069210892543197,
          "train_metric": 6.628997653024271e-05,
          "validation_metric": 5.105523450765759e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.429784736001238
        },
        {
          "epoch": 991,
          "updates": 991,
          "samples_seen": 1014784,
          "online_loss": 0.000658184930216521,
          "train_loss": 0.0006525017088279128,
          "validation_loss": 0.0005277899326756597,
          "train_metric": 6.571758422069252e-05,
          "validation_metric": 5.315707312547602e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.43760618700071
        },
        {
          "epoch": 992,
          "updates": 992,
          "samples_seen": 1015808,
          "online_loss": 0.0006525016506202519,
          "train_loss": 0.0006478198920376599,
          "validation_loss": 0.0005040006944909692,
          "train_metric": 6.524604395963252e-05,
          "validation_metric": 5.0761100283125415e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.44635136100078
        },
        {
          "epoch": 993,
          "updates": 993,
          "samples_seen": 1016832,
          "online_loss": 0.0006478199502453208,
          "train_loss": 0.0006438592099584639,
          "validation_loss": 0.000516285770572722,
          "train_metric": 6.484714685939252e-05,
          "validation_metric": 5.1998416893184185e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.454496423000819
        },
        {
          "epoch": 994,
          "updates": 994,
          "samples_seen": 1017856,
          "online_loss": 0.0006438592099584639,
          "train_loss": 0.0006408768240362406,
          "validation_loss": 0.0005042643169872463,
          "train_metric": 6.454676622524858e-05,
          "validation_metric": 5.078765389043838e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.462123450000945
        },
        {
          "epoch": 995,
          "updates": 995,
          "samples_seen": 1018880,
          "online_loss": 0.0006408767658285797,
          "train_loss": 0.0006387077737599611,
          "validation_loss": 0.0005084198201075196,
          "train_metric": 6.432830559788272e-05,
          "validation_metric": 5.120619243825786e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.469813896001142
        },
        {
          "epoch": 996,
          "updates": 996,
          "samples_seen": 1019904,
          "online_loss": 0.000638707831967622,
          "train_loss": 0.0006372057250700891,
          "validation_loss": 0.0005065763252787292,
          "train_metric": 6.417703843908384e-05,
          "validation_metric": 5.102050636196509e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.477292334000595
        },
        {
          "epoch": 997,
          "updates": 997,
          "samples_seen": 1020928,
          "online_loss": 0.0006372059579007328,
          "train_loss": 0.0006361986743286252,
          "validation_loss": 0.0005032154731452465,
          "train_metric": 6.40755970380269e-05,
          "validation_metric": 5.068202153779566e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.4856790190006
        },
        {
          "epoch": 998,
          "updates": 998,
          "samples_seen": 1021952,
          "online_loss": 0.0006361985579133034,
          "train_loss": 0.0006355317891575396,
          "validation_loss": 0.000508865574374795,
          "train_metric": 6.400843267329037e-05,
          "validation_metric": 5.125107054482214e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.493642493000152
        },
        {
          "epoch": 999,
          "updates": 999,
          "samples_seen": 1022976,
          "online_loss": 0.0006355317891575396,
          "train_loss": 0.0006351592601276934,
          "validation_loss": 0.000498867710120976,
          "train_metric": 6.397091055987403e-05,
          "validation_metric": 5.024412166676484e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.502299887000845
        },
        {
          "epoch": 1000,
          "updates": 1000,
          "samples_seen": 1024000,
          "online_loss": 0.0006351593183353543,
          "train_loss": 0.000635014905128628,
          "validation_loss": 0.0005107403849251568,
          "train_metric": 6.395637319656089e-05,
          "validation_metric": 5.143989983480424e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.509896975001539
        },
        {
          "epoch": 1001,
          "updates": 1001,
          "samples_seen": 1025024,
          "online_loss": 0.0006350149633362889,
          "train_loss": 0.0006352690979838371,
          "validation_loss": 0.0004950868315063417,
          "train_metric": 6.398198456736282e-05,
          "validation_metric": 4.986333442502655e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.51844968200021
        },
        {
          "epoch": 1002,
          "updates": 1002,
          "samples_seen": 1026048,
          "online_loss": 0.000635269156191498,
          "train_loss": 0.000635801930911839,
          "validation_loss": 0.0005134421517141163,
          "train_metric": 6.40356374788098e-05,
          "validation_metric": 5.171201701159589e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.526247599000271
        },
        {
          "epoch": 1003,
          "updates": 1003,
          "samples_seen": 1027072,
          "online_loss": 0.000635801930911839,
          "train_loss": 0.0006371396011672914,
          "validation_loss": 0.0004924579989165068,
          "train_metric": 6.417035910999402e-05,
          "validation_metric": 4.9598562327446416e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.534863896000388
        },
        {
          "epoch": 1004,
          "updates": 1004,
          "samples_seen": 1028096,
          "online_loss": 0.0006371396011672914,
          "train_loss": 0.0006389608606696129,
          "validation_loss": 0.0005187432980164886,
          "train_metric": 6.435380055336282e-05,
          "validation_metric": 5.224592314334586e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.542777118000231
        },
        {
          "epoch": 1005,
          "updates": 1005,
          "samples_seen": 1029120,
          "online_loss": 0.0006389608606696129,
          "train_loss": 0.0006426576292142272,
          "validation_loss": 0.0004919053171761334,
          "train_metric": 6.472612585639581e-05,
          "validation_metric": 4.95428939757403e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.551683437000975
        },
        {
          "epoch": 1006,
          "updates": 1006,
          "samples_seen": 1030144,
          "online_loss": 0.0006426576292142272,
          "train_loss": 0.0006473846733570099,
          "validation_loss": 0.000529381912201643,
          "train_metric": 6.520222086692229e-05,
          "validation_metric": 5.331740976544097e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.55949179900017
        },
        {
          "epoch": 1007,
          "updates": 1007,
          "samples_seen": 1031168,
          "online_loss": 0.0006473847315646708,
          "train_loss": 0.0006566757801920176,
          "validation_loss": 0.0004951779264956713,
          "train_metric": 6.61379744997248e-05,
          "validation_metric": 4.9872509407578036e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.567358880000029
        },
        {
          "epoch": 1008,
          "updates": 1008,
          "samples_seen": 1032192,
          "online_loss": 0.0006566757219843566,
          "train_loss": 0.0006680529331788421,
          "validation_loss": 0.0005505089648067951,
          "train_metric": 6.72838359605521e-05,
          "validation_metric": 5.5445249017793685e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.57502480200128
        },
        {
          "epoch": 1009,
          "updates": 1009,
          "samples_seen": 1033216,
          "online_loss": 0.0006680528749711812,
          "train_loss": 0.0006907910574227571,
          "validation_loss": 0.0005076448433101177,
          "train_metric": 6.957395089557394e-05,
          "validation_metric": 5.112812505103648e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.584579705000579
        },
        {
          "epoch": 1010,
          "updates": 1010,
          "samples_seen": 1034240,
          "online_loss": 0.0006907911156304181,
          "train_loss": 0.0007163554546423256,
          "validation_loss": 0.000593390897847712,
          "train_metric": 7.214868674054742e-05,
          "validation_metric": 5.976414468022995e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.592403151000326
        },
        {
          "epoch": 1011,
          "updates": 1011,
          "samples_seen": 1035264,
          "online_loss": 0.0007163553382270038,
          "train_loss": 0.0007697967230342329,
          "validation_loss": 0.0005439639207907021,
          "train_metric": 7.753110548947006e-05,
          "validation_metric": 5.478605453390628e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.601633563001087
        },
        {
          "epoch": 1012,
          "updates": 1012,
          "samples_seen": 1036288,
          "online_loss": 0.0007697968394495547,
          "train_loss": 0.000818698899820447,
          "validation_loss": 0.000676262890920043,
          "train_metric": 8.245634671766311e-05,
          "validation_metric": 6.81107267155312e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.609191593001015
        },
        {
          "epoch": 1013,
          "updates": 1013,
          "samples_seen": 1037312,
          "online_loss": 0.000818698899820447,
          "train_loss": 0.0009282850078307092,
          "validation_loss": 0.0006273422040976584,
          "train_metric": 9.349345782538876e-05,
          "validation_metric": 6.31836082902737e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.617333740001413
        },
        {
          "epoch": 1014,
          "updates": 1014,
          "samples_seen": 1038336,
          "online_loss": 0.0009282850660383701,
          "train_loss": 0.0009848650079220533,
          "validation_loss": 0.0008027541916817427,
          "train_metric": 9.919199510477483e-05,
          "validation_metric": 8.085046283667907e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.625008393000826
        },
        {
          "epoch": 1015,
          "updates": 1015,
          "samples_seen": 1039360,
          "online_loss": 0.0009848650079220533,
          "train_loss": 0.001135219936259091,
          "validation_loss": 0.0007438641041517258,
          "train_metric": 0.0001143351910286583,
          "validation_metric": 7.491926953662187e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.633495510001012
        },
        {
          "epoch": 1016,
          "updates": 1016,
          "samples_seen": 1040384,
          "online_loss": 0.001135219819843769,
          "train_loss": 0.0011002946412190795,
          "validation_loss": 0.0008862048853188753,
          "train_metric": 0.00011081764387199655,
          "validation_metric": 8.92553071025759e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.758260610000434
        },
        {
          "epoch": 1017,
          "updates": 1017,
          "samples_seen": 1041408,
          "online_loss": 0.0011002946412190795,
          "train_loss": 0.001136608305387199,
          "validation_loss": 0.0007424845243804157,
          "train_metric": 0.00011447501310613006,
          "validation_metric": 7.478033512597904e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.767402295001375
        },
        {
          "epoch": 1018,
          "updates": 1018,
          "samples_seen": 1042432,
          "online_loss": 0.0011366093531250954,
          "train_loss": 0.000943946884945035,
          "validation_loss": 0.0007650591433048248,
          "train_metric": 9.507086360827088e-05,
          "validation_metric": 7.705395546508953e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.775381587000084
        },
        {
          "epoch": 1019,
          "updates": 1019,
          "samples_seen": 1043456,
          "online_loss": 0.0009439469431526959,
          "train_loss": 0.0008047684677876532,
          "validation_loss": 0.0005544520681723952,
          "train_metric": 8.105334563879296e-05,
          "validation_metric": 5.5842385336291045e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.784386601000733
        },
        {
          "epoch": 1020,
          "updates": 1020,
          "samples_seen": 1044480,
          "online_loss": 0.0008047685259953141,
          "train_loss": 0.0006615233141928911,
          "validation_loss": 0.0005379451322369277,
          "train_metric": 6.662620580755174e-05,
          "validation_metric": 5.41798654012382e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.793199343001106
        },
        {
          "epoch": 1021,
          "updates": 1021,
          "samples_seen": 1045504,
          "online_loss": 0.000661523372400552,
          "train_loss": 0.000607120047789067,
          "validation_loss": 0.00047587574226781726,
          "train_metric": 6.114691495895386e-05,
          "validation_metric": 4.79284644825384e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.803451119001693
        },
        {
          "epoch": 1022,
          "updates": 1022,
          "samples_seen": 1046528,
          "online_loss": 0.000607120047789067,
          "train_loss": 0.0006326356087811291,
          "validation_loss": 0.0004805134376510978,
          "train_metric": 6.371673953253776e-05,
          "validation_metric": 4.8395548219559714e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.812490910000633
        },
        {
          "epoch": 1023,
          "updates": 1023,
          "samples_seen": 1047552,
          "online_loss": 0.000632635725196451,
          "train_loss": 0.0007051856955513358,
          "validation_loss": 0.0005852221511304379,
          "train_metric": 7.102372910594568e-05,
          "validation_metric": 5.894143396290019e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.821210720001545
        },
        {
          "epoch": 1024,
          "updates": 1024,
          "samples_seen": 1048576,
          "online_loss": 0.0007051856955513358,
          "train_loss": 0.0008039958192966878,
          "validation_loss": 0.0005656669964082539,
          "train_metric": 8.097550744423643e-05,
          "validation_metric": 5.6971897720359266e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.829972328001531
        },
        {
          "epoch": 1025,
          "updates": 1025,
          "samples_seen": 1049600,
          "online_loss": 0.0008039957610890269,
          "train_loss": 0.0008423904655501246,
          "validation_loss": 0.0006977069424465299,
          "train_metric": 8.484248246531934e-05,
          "validation_metric": 7.027050014585257e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.838228993001394
        },
        {
          "epoch": 1026,
          "updates": 1026,
          "samples_seen": 1050624,
          "online_loss": 0.0008423904073424637,
          "train_loss": 0.0008804461103864014,
          "validation_loss": 0.0005992785445414484,
          "train_metric": 8.86753186932765e-05,
          "validation_metric": 6.03571388637647e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.84653339200122
        },
        {
          "epoch": 1027,
          "updates": 1027,
          "samples_seen": 1051648,
          "online_loss": 0.0008804461685940623,
          "train_loss": 0.0008042146801017225,
          "validation_loss": 0.0006647268310189247,
          "train_metric": 8.09975535958074e-05,
          "validation_metric": 6.694886542391032e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.855683566000152
        },
        {
          "epoch": 1028,
          "updates": 1028,
          "samples_seen": 1052672,
          "online_loss": 0.0008042146218940616,
          "train_loss": 0.0007327759522013366,
          "validation_loss": 0.0005149298813194036,
          "train_metric": 7.380251190625131e-05,
          "validation_metric": 5.186185080674477e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.864819100001114
        },
        {
          "epoch": 1029,
          "updates": 1029,
          "samples_seen": 1053696,
          "online_loss": 0.0007327759522013366,
          "train_loss": 0.0006438099662773311,
          "validation_loss": 0.0005253504496067762,
          "train_metric": 6.484217738034204e-05,
          "validation_metric": 5.2911374950781465e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.873645776000558
        },
        {
          "epoch": 1030,
          "updates": 1030,
          "samples_seen": 1054720,
          "online_loss": 0.0006438099080696702,
          "train_loss": 0.0006000847206450999,
          "validation_loss": 0.00046633410966023803,
          "train_metric": 6.043833491276018e-05,
          "validation_metric": 4.6967466914793476e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.88336300300034
        },
        {
          "epoch": 1031,
          "updates": 1031,
          "samples_seen": 1055744,
          "online_loss": 0.000600084662437439,
          "train_loss": 0.0006049369112588465,
          "validation_loss": 0.00046797501272521913,
          "train_metric": 6.092702460591681e-05,
          "validation_metric": 4.713272937806323e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.892074238001442
        },
        {
          "epoch": 1032,
          "updates": 1032,
          "samples_seen": 1056768,
          "online_loss": 0.0006049368530511856,
          "train_loss": 0.0006427670596167445,
          "validation_loss": 0.0005299485637806356,
          "train_metric": 6.473714165622368e-05,
          "validation_metric": 5.337447873898782e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.9008693830001
        },
        {
          "epoch": 1033,
          "updates": 1033,
          "samples_seen": 1057792,
          "online_loss": 0.0006427670596167445,
          "train_loss": 0.0006939682061783969,
          "validation_loss": 0.0005125193856656551,
          "train_metric": 6.98939329595305e-05,
          "validation_metric": 5.161907392903231e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.909340868000072
        },
        {
          "epoch": 1034,
          "updates": 1034,
          "samples_seen": 1058816,
          "online_loss": 0.0006939682061783969,
          "train_loss": 0.0007174920756369829,
          "validation_loss": 0.0005973008228465915,
          "train_metric": 7.22631739336066e-05,
          "validation_metric": 6.015795224811882e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.917531413000688
        },
        {
          "epoch": 1035,
          "updates": 1035,
          "samples_seen": 1059840,
          "online_loss": 0.0007174920756369829,
          "train_loss": 0.0007323011523112655,
          "validation_loss": 0.0005289211985655129,
          "train_metric": 7.375467248493806e-05,
          "validation_metric": 5.3271007345756516e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.925761299000442
        },
        {
          "epoch": 1036,
          "updates": 1036,
          "samples_seen": 1060864,
          "online_loss": 0.0007323010941036046,
          "train_loss": 0.000699036056175828,
          "validation_loss": 0.0005829540896229446,
          "train_metric": 7.04043559380807e-05,
          "validation_metric": 5.8713001635624096e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.93366463700113
        },
        {
          "epoch": 1037,
          "updates": 1037,
          "samples_seen": 1061888,
          "online_loss": 0.0006990362307988107,
          "train_loss": 0.0006647815462201834,
          "validation_loss": 0.0004848484240937978,
          "train_metric": 6.695436604786664e-05,
          "validation_metric": 4.883215297013521e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.94163152100009
        },
        {
          "epoch": 1038,
          "updates": 1038,
          "samples_seen": 1062912,
          "online_loss": 0.0006647814880125225,
          "train_loss": 0.0006207380210980773,
          "validation_loss": 0.0005098038818687201,
          "train_metric": 6.251846207305789e-05,
          "validation_metric": 5.134557795827277e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.950308932000553
        },
        {
          "epoch": 1039,
          "updates": 1039,
          "samples_seen": 1063936,
          "online_loss": 0.0006207379628904164,
          "train_loss": 0.0005934975342825055,
          "validation_loss": 0.00045711221173405647,
          "train_metric": 5.977490218356252e-05,
          "validation_metric": 4.603866909747012e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.95909580100124
        },
        {
          "epoch": 1040,
          "updates": 1040,
          "samples_seen": 1064960,
          "online_loss": 0.0005934975924901664,
          "train_loss": 0.000585964706260711,
          "validation_loss": 0.00046012975508347154,
          "train_metric": 5.9016219893237576e-05,
          "validation_metric": 4.634258584701456e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.968356720000884
        },
        {
          "epoch": 1041,
          "updates": 1041,
          "samples_seen": 1065984,
          "online_loss": 0.00058596464805305,
          "train_loss": 0.0005958301480859518,
          "validation_loss": 0.00048314008745364845,
          "train_metric": 6.000982830300927e-05,
          "validation_metric": 4.8660098400432616e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.976353952000864
        },
        {
          "epoch": 1042,
          "updates": 1042,
          "samples_seen": 1067008,
          "online_loss": 0.0005958301480859518,
          "train_loss": 0.0006151775014586747,
          "validation_loss": 0.00046625640243291855,
          "train_metric": 6.195842433953658e-05,
          "validation_metric": 4.695963434642181e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.984734845000276
        },
        {
          "epoch": 1043,
          "updates": 1043,
          "samples_seen": 1068032,
          "online_loss": 0.0006151775596663356,
          "train_loss": 0.0006311066681519151,
          "validation_loss": 0.000520823523402214,
          "train_metric": 6.356275116559118e-05,
          "validation_metric": 5.245542706688866e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.992712936000316
        },
        {
          "epoch": 1044,
          "updates": 1044,
          "samples_seen": 1069056,
          "online_loss": 0.000631106726359576,
          "train_loss": 0.0006435986142605543,
          "validation_loss": 0.0004827489610761404,
          "train_metric": 6.482088792836294e-05,
          "validation_metric": 4.8620720917824656e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.001201036000566
        },
        {
          "epoch": 1045,
          "updates": 1045,
          "samples_seen": 1070080,
          "online_loss": 0.0006435986142605543,
          "train_loss": 0.0006391261704266071,
          "validation_loss": 0.0005278319003991783,
          "train_metric": 6.437044794438407e-05,
          "validation_metric": 5.3161293180892244e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.009156487001746
        },
        {
          "epoch": 1046,
          "updates": 1046,
          "samples_seen": 1071104,
          "online_loss": 0.0006391261704266071,
          "train_loss": 0.0006312921177595854,
          "validation_loss": 0.0004735301772598177,
          "train_metric": 6.358142854878679e-05,
          "validation_metric": 4.769222141476348e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.018032130001302
        },
        {
          "epoch": 1047,
          "updates": 1047,
          "samples_seen": 1072128,
          "online_loss": 0.0006312921177595854,
          "train_loss": 0.0006130668916739523,
          "validation_loss": 0.000502920476719737,
          "train_metric": 6.174584268592298e-05,
          "validation_metric": 5.065231380285695e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.02632995500062
        },
        {
          "epoch": 1048,
          "updates": 1048,
          "samples_seen": 1073152,
          "online_loss": 0.0006130668334662914,
          "train_loss": 0.0005971269565634429,
          "validation_loss": 0.00045457290252670646,
          "train_metric": 6.014043538016267e-05,
          "validation_metric": 4.5782915549352765e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.035566413000197
        },
        {
          "epoch": 1049,
          "updates": 1049,
          "samples_seen": 1074176,
          "online_loss": 0.0005971269565634429,
          "train_loss": 0.0005831989692524076,
          "validation_loss": 0.0004703050071839243,
          "train_metric": 5.8737670769914985e-05,
          "validation_metric": 4.736739720101468e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.044131255001048
        },
        {
          "epoch": 1050,
          "updates": 1050,
          "samples_seen": 1075200,
          "online_loss": 0.0005831989692524076,
          "train_loss": 0.0005753448931500316,
          "validation_loss": 0.0004513487801887095,
          "train_metric": 5.794663593405858e-05,
          "validation_metric": 4.545820047496818e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.053841863000343
        },
        {
          "epoch": 1051,
          "updates": 1051,
          "samples_seen": 1076224,
          "online_loss": 0.0005753448931500316,
          "train_loss": 0.000573448371142149,
          "validation_loss": 0.00045238216989673674,
          "train_metric": 5.775563113274984e-05,
          "validation_metric": 4.556227213470265e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.062748931000897
        },
        {
          "epoch": 1052,
          "updates": 1052,
          "samples_seen": 1077248,
          "online_loss": 0.0005734484293498099,
          "train_loss": 0.0005762676591984928,
          "validation_loss": 0.00046525633661076427,
          "train_metric": 5.803957537864335e-05,
          "validation_metric": 4.685891326516867e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.071285829000772
        },
        {
          "epoch": 1053,
          "updates": 1053,
          "samples_seen": 1078272,
          "online_loss": 0.0005762677174061537,
          "train_loss": 0.0005819368525408208,
          "validation_loss": 0.0004521634546108544,
          "train_metric": 5.8610552514437586e-05,
          "validation_metric": 4.554024781100452e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.079214433000743
        },
        {
          "epoch": 1054,
          "updates": 1054,
          "samples_seen": 1079296,
          "online_loss": 0.0005819369107484818,
          "train_loss": 0.0005877042422071099,
          "validation_loss": 0.00048160774167627096,
          "train_metric": 5.9191421314608306e-05,
          "validation_metric": 4.850576806347817e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.087841785001729
        },
        {
          "epoch": 1055,
          "updates": 1055,
          "samples_seen": 1080320,
          "online_loss": 0.0005877042422071099,
          "train_loss": 0.0005934929940849543,
          "validation_loss": 0.000455892295576632,
          "train_metric": 5.977443652227521e-05,
          "validation_metric": 4.5915800001239404e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.096433045000595
        },
        {
          "epoch": 1056,
          "updates": 1056,
          "samples_seen": 1081344,
          "online_loss": 0.0005934929940849543,
          "train_loss": 0.0005958566907793283,
          "validation_loss": 0.0004893350414931774,
          "train_metric": 6.0012502217432484e-05,
          "validation_metric": 4.928403359372169e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.104917493001267
        },
        {
          "epoch": 1057,
          "updates": 1057,
          "samples_seen": 1082368,
          "online_loss": 0.0005958566907793283,
          "train_loss": 0.0005980036803521216,
          "validation_loss": 0.0004542845126707107,
          "train_metric": 6.02287364017684e-05,
          "validation_metric": 4.575387356453575e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.216735216001325
        },
        {
          "epoch": 1058,
          "updates": 1058,
          "samples_seen": 1083392,
          "online_loss": 0.0005980036221444607,
          "train_loss": 0.0005957200773991644,
          "validation_loss": 0.0004875592130701989,
          "train_metric": 5.999874338158406e-05,
          "validation_metric": 4.910517964162864e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.225504105001164
        },
        {
          "epoch": 1059,
          "updates": 1059,
          "samples_seen": 1084416,
          "online_loss": 0.0005957200191915035,
          "train_loss": 0.0005936277448199689,
          "validation_loss": 0.0004488716076593846,
          "train_metric": 5.978800618322566e-05,
          "validation_metric": 4.520870788837783e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.234333229000185
        },
        {
          "epoch": 1060,
          "updates": 1060,
          "samples_seen": 1085440,
          "online_loss": 0.000593627686612308,
          "train_loss": 0.0005879410891793668,
          "validation_loss": 0.00047906526015140116,
          "train_metric": 5.92152719036676e-05,
          "validation_metric": 4.82496980112046e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.242556454000805
        },
        {
          "epoch": 1061,
          "updates": 1061,
          "samples_seen": 1086464,
          "online_loss": 0.0005879407981410623,
          "train_loss": 0.0005829784786328673,
          "validation_loss": 0.0004440387128852308,
          "train_metric": 5.871545727131888e-05,
          "validation_metric": 4.4721949961967766e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.251692643001661
        },
        {
          "epoch": 1062,
          "updates": 1062,
          "samples_seen": 1087488,
          "online_loss": 0.0005829785368405282,
          "train_loss": 0.0005766745307482779,
          "validation_loss": 0.0004686071479227394,
          "train_metric": 5.808054993394762e-05,
          "validation_metric": 4.719639400718734e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.260640514001352
        },
        {
          "epoch": 1063,
          "updates": 1063,
          "samples_seen": 1088512,
          "online_loss": 0.0005766745307482779,
          "train_loss": 0.0005717018502764404,
          "validation_loss": 0.00044221626012586057,
          "train_metric": 5.757971666753292e-05,
          "validation_metric": 4.453840301721357e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.270339935001175
        },
        {
          "epoch": 1064,
          "updates": 1064,
          "samples_seen": 1089536,
          "online_loss": 0.0005717018502764404,
          "train_loss": 0.0005670237587764859,
          "validation_loss": 0.0004599198291543871,
          "train_metric": 5.710856567020528e-05,
          "validation_metric": 4.632144555216655e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.27829386600024
        },
        {
          "epoch": 1065,
          "updates": 1065,
          "samples_seen": 1090560,
          "online_loss": 0.0005670238169841468,
          "train_loss": 0.0005634740809909999,
          "validation_loss": 0.0004415612493176013,
          "train_metric": 5.675104330293834e-05,
          "validation_metric": 4.447242827154696e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.288402485000915
        },
        {
          "epoch": 1066,
          "updates": 1066,
          "samples_seen": 1091584,
          "online_loss": 0.0005634741391986609,
          "train_loss": 0.0005604400648735464,
          "validation_loss": 0.000452667532954365,
          "train_metric": 5.6445482186973095e-05,
          "validation_metric": 4.559101216727868e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.296738571001697
        },
        {
          "epoch": 1067,
          "updates": 1067,
          "samples_seen": 1092608,
          "online_loss": 0.0005604401812888682,
          "train_loss": 0.000558049010578543,
          "validation_loss": 0.0004399519239086658,
          "train_metric": 5.620465162792243e-05,
          "validation_metric": 4.431034903973341e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.306144696000047
        },
        {
          "epoch": 1068,
          "updates": 1068,
          "samples_seen": 1093632,
          "online_loss": 0.000558049010578543,
          "train_loss": 0.0005560601712204516,
          "validation_loss": 0.0004461958596948534,
          "train_metric": 5.6004340876825154e-05,
          "validation_metric": 4.493921005632728e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.314083397001014
        },
        {
          "epoch": 1069,
          "updates": 1069,
          "samples_seen": 1094656,
          "online_loss": 0.0005560601712204516,
          "train_loss": 0.0005544484010897577,
          "validation_loss": 0.00043795534293167293,
          "train_metric": 5.5842010624473915e-05,
          "validation_metric": 4.410926339915022e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.323448011000437
        },
        {
          "epoch": 1070,
          "updates": 1070,
          "samples_seen": 1095680,
          "online_loss": 0.0005544484010897577,
          "train_loss": 0.0005530636990442872,
          "validation_loss": 0.00044119282392784953,
          "train_metric": 5.5702545068925247e-05,
          "validation_metric": 4.443533543962985e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.332431337001253
        },
        {
          "epoch": 1071,
          "updates": 1071,
          "samples_seen": 1096704,
          "online_loss": 0.0005530636408366263,
          "train_loss": 0.000551823410205543,
          "validation_loss": 0.0004369606904219836,
          "train_metric": 5.557763506658375e-05,
          "validation_metric": 4.400908073876053e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.34130286800064
        },
        {
          "epoch": 1072,
          "updates": 1072,
          "samples_seen": 1097728,
          "online_loss": 0.000551823468413204,
          "train_loss": 0.000550652330275625,
          "validation_loss": 0.00043840898433700204,
          "train_metric": 5.545969179365784e-05,
          "validation_metric": 4.4154956412967294e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.349234944000273
        },
        {
          "epoch": 1073,
          "updates": 1073,
          "samples_seen": 1098752,
          "online_loss": 0.0005506525631062686,
          "train_loss": 0.0005495473160408437,
          "validation_loss": 0.0004372734110802412,
          "train_metric": 5.534839874599129e-05,
          "validation_metric": 4.404057835927233e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.356922548000512
        },
        {
          "epoch": 1074,
          "updates": 1074,
          "samples_seen": 1099776,
          "online_loss": 0.0005495473742485046,
          "train_loss": 0.0005485076690092683,
          "validation_loss": 0.00043710845056921244,
          "train_metric": 5.5243690439965576e-05,
          "validation_metric": 4.402396734803915e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.367304587000035
        },
        {
          "epoch": 1075,
          "updates": 1075,
          "samples_seen": 1100800,
          "online_loss": 0.0005485077272169292,
          "train_loss": 0.0005475180805660784,
          "validation_loss": 0.00043764186557382345,
          "train_metric": 5.514401345863007e-05,
          "validation_metric": 4.4077689381083474e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.37517438300165
        },
        {
          "epoch": 1076,
          "updates": 1076,
          "samples_seen": 1101824,
          "online_loss": 0.0005475180805660784,
          "train_loss": 0.00054655980784446,
          "validation_loss": 0.000435498368460685,
          "train_metric": 5.504750879481435e-05,
          "validation_metric": 4.386179716675542e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.383708434001164
        },
        {
          "epoch": 1077,
          "updates": 1077,
          "samples_seen": 1102848,
          "online_loss": 0.0005465598660521209,
          "train_loss": 0.0005456203361973166,
          "validation_loss": 0.0004370447713881731,
          "train_metric": 5.4952881328063086e-05,
          "validation_metric": 4.4017542677465826e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.391107538000142
        },
        {
          "epoch": 1078,
          "updates": 1078,
          "samples_seen": 1103872,
          "online_loss": 0.0005456202779896557,
          "train_loss": 0.0005447324947454035,
          "validation_loss": 0.00043275117059238255,
          "train_metric": 5.486345980898477e-05,
          "validation_metric": 4.358511068858206e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.399331945000085
        },
        {
          "epoch": 1079,
          "updates": 1079,
          "samples_seen": 1104896,
          "online_loss": 0.0005447324947454035,
          "train_loss": 0.000543933070730418,
          "validation_loss": 0.00043590241693891585,
          "train_metric": 5.4782947700005025e-05,
          "validation_metric": 4.3902498873649165e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.406864893000602
        },
        {
          "epoch": 1080,
          "updates": 1080,
          "samples_seen": 1105920,
          "online_loss": 0.0005439331289380789,
          "train_loss": 0.0005432839971035719,
          "validation_loss": 0.0004294195678085089,
          "train_metric": 5.4717573220841587e-05,
          "validation_metric": 4.324957262724638e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.415129216000423
        },
        {
          "epoch": 1081,
          "updates": 1081,
          "samples_seen": 1106944,
          "online_loss": 0.000543283938895911,
          "train_loss": 0.0005428496515378356,
          "validation_loss": 0.00043575803283602,
          "train_metric": 5.46738228877075e-05,
          "validation_metric": 4.3887950596399605e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.423377105000327
        },
        {
          "epoch": 1082,
          "updates": 1082,
          "samples_seen": 1107968,
          "online_loss": 0.0005428496515378356,
          "train_loss": 0.0005428019212558866,
          "validation_loss": 0.00042670531547628343,
          "train_metric": 5.466903166961856e-05,
          "validation_metric": 4.297619670978747e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.432242379001764
        },
        {
          "epoch": 1083,
          "updates": 1083,
          "samples_seen": 1108992,
          "online_loss": 0.0005428019212558866,
          "train_loss": 0.0005433327169157565,
          "validation_loss": 0.00043837004341185093,
          "train_metric": 5.4722477216273546e-05,
          "validation_metric": 4.415102375787683e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.44136604400046
        },
        {
          "epoch": 1084,
          "updates": 1084,
          "samples_seen": 1110016,
          "online_loss": 0.0005433326587080956,
          "train_loss": 0.0005450890166684985,
          "validation_loss": 0.0004253672668710351,
          "train_metric": 5.489937393576838e-05,
          "validation_metric": 4.284143142285757e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.451271391000773
        },
        {
          "epoch": 1085,
          "updates": 1085,
          "samples_seen": 1111040,
          "online_loss": 0.0005450890748761594,
          "train_loss": 0.0005485917208716273,
          "validation_loss": 0.00044610226177610457,
          "train_metric": 5.525214510271326e-05,
          "validation_metric": 4.492979132919572e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.45953062800072
        },
        {
          "epoch": 1086,
          "updates": 1086,
          "samples_seen": 1112064,
          "online_loss": 0.0005485917208716273,
          "train_loss": 0.0005561554571613669,
          "validation_loss": 0.00042737688636407256,
          "train_metric": 5.6013945140875876e-05,
          "validation_metric": 4.304383764974773e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.467167329001313
        },
        {
          "epoch": 1087,
          "updates": 1087,
          "samples_seen": 1113088,
          "online_loss": 0.0005561555153690279,
          "train_loss": 0.0005689498502761126,
          "validation_loss": 0.00046665023546665907,
          "train_metric": 5.730254270019941e-05,
          "validation_metric": 4.6999302867334336e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.474967051000931
        },
        {
          "epoch": 1088,
          "updates": 1088,
          "samples_seen": 1114112,
          "online_loss": 0.0005689497920684516,
          "train_loss": 0.0005960134440101683,
          "validation_loss": 0.0004430152475833893,
          "train_metric": 6.0028287407476455e-05,
          "validation_metric": 4.461887147044763e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.482633483001337
        },
        {
          "epoch": 1089,
          "updates": 1089,
          "samples_seen": 1115136,
          "online_loss": 0.0005960134440101683,
          "train_loss": 0.0006373354699462652,
          "validation_loss": 0.0005243425839580595,
          "train_metric": 6.419009150704369e-05,
          "validation_metric": 5.280986806610599e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.490391576000548
        },
        {
          "epoch": 1090,
          "updates": 1090,
          "samples_seen": 1116160,
          "online_loss": 0.0006373354117386043,
          "train_loss": 0.000729315506760031,
          "validation_loss": 0.000511220539920032,
          "train_metric": 7.345397170865908e-05,
          "validation_metric": 5.148825948708691e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.498109141000896
        },
        {
          "epoch": 1091,
          "updates": 1091,
          "samples_seen": 1117184,
          "online_loss": 0.000729315506760031,
          "train_loss": 0.0008433882030658424,
          "validation_loss": 0.0006818558322265744,
          "train_metric": 8.494297071592882e-05,
          "validation_metric": 6.867402407806367e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.506023064001056
        },
        {
          "epoch": 1092,
          "updates": 1092,
          "samples_seen": 1118208,
          "online_loss": 0.0008433881448581815,
          "train_loss": 0.0011168336495757103,
          "validation_loss": 0.0007357660797424614,
          "train_metric": 0.0001124834016081877,
          "validation_metric": 7.410366379190236e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.513793045000057
        },
        {
          "epoch": 1093,
          "updates": 1093,
          "samples_seen": 1119232,
          "online_loss": 0.0011168336495757103,
          "train_loss": 0.0012912546517327428,
          "validation_loss": 0.001004980644211173,
          "train_metric": 0.0001300504372920841,
          "validation_metric": 0.00010121798550244421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.521119936000105
        },
        {
          "epoch": 1094,
          "updates": 1094,
          "samples_seen": 1120256,
          "online_loss": 0.001291254535317421,
          "train_loss": 0.0017595004756003618,
          "validation_loss": 0.001127383904531598,
          "train_metric": 0.0001772104442352429,
          "validation_metric": 0.0001135459533543326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.528719841000566
        },
        {
          "epoch": 1095,
          "updates": 1095,
          "samples_seen": 1121280,
          "online_loss": 0.0017595004756003618,
          "train_loss": 0.0015253603924065828,
          "validation_loss": 0.0011661016615107656,
          "train_metric": 0.00015362871636170894,
          "validation_metric": 0.00011744549556169659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.535907640000005
        },
        {
          "epoch": 1096,
          "updates": 1096,
          "samples_seen": 1122304,
          "online_loss": 0.0015253603924065828,
          "train_loss": 0.0014255907153710723,
          "validation_loss": 0.0009104657219722867,
          "train_metric": 0.00014358026965055615,
          "validation_metric": 9.169877012027428e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.54355359100009
        },
        {
          "epoch": 1097,
          "updates": 1097,
          "samples_seen": 1123328,
          "online_loss": 0.0014255907153710723,
          "train_loss": 0.0008352425647899508,
          "validation_loss": 0.0006572218262590468,
          "train_metric": 8.412257739109918e-05,
          "validation_metric": 6.619297346333042e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.553095763001693
        },
        {
          "epoch": 1098,
          "updates": 1098,
          "samples_seen": 1124352,
          "online_loss": 0.0008352425647899508,
          "train_loss": 0.0005516766104847193,
          "validation_loss": 0.00040508201345801353,
          "train_metric": 5.556285395869054e-05,
          "validation_metric": 4.0798375266604125e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.65946654200161
        },
        {
          "epoch": 1099,
          "updates": 1099,
          "samples_seen": 1125376,
          "online_loss": 0.0005516766104847193,
          "train_loss": 0.0006071479292586446,
          "validation_loss": 0.00043536207522265613,
          "train_metric": 6.114971620263532e-05,
          "validation_metric": 4.384807834867388e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.66477446800127
        },
        {
          "epoch": 1100,
          "updates": 1100,
          "samples_seen": 1126400,
          "online_loss": 0.0006071479292586446,
          "train_loss": 0.0008664295310154557,
          "validation_loss": 0.0006886804476380348,
          "train_metric": 8.726360829314217e-05,
          "validation_metric": 6.936136924196035e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.66999613400003
        },
        {
          "epoch": 1101,
          "updates": 1101,
          "samples_seen": 1127424,
          "online_loss": 0.0008664295310154557,
          "train_loss": 0.0012134042335674167,
          "validation_loss": 0.0008164847968146205,
          "train_metric": 0.00012220963253639638,
          "validation_metric": 8.223336044466123e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.675250216001587
        },
        {
          "epoch": 1102,
          "updates": 1102,
          "samples_seen": 1128448,
          "online_loss": 0.0012134042335674167,
          "train_loss": 0.0011204201728105545,
          "validation_loss": 0.0008862435934133828,
          "train_metric": 0.00011284460924798623,
          "validation_metric": 8.925922156777233e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.680614998000237
        },
        {
          "epoch": 1103,
          "updates": 1103,
          "samples_seen": 1129472,
          "online_loss": 0.0011204195907339454,
          "train_loss": 0.0009801356354728341,
          "validation_loss": 0.0006578545435331762,
          "train_metric": 9.87156672636047e-05,
          "validation_metric": 6.625670357607305e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.68598912500056
        },
        {
          "epoch": 1104,
          "updates": 1104,
          "samples_seen": 1130496,
          "online_loss": 0.0009801355190575123,
          "train_loss": 0.0006523407646454871,
          "validation_loss": 0.0005344303790479898,
          "train_metric": 6.570137338712811e-05,
          "validation_metric": 5.382586823543534e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.691481881000072
        },
        {
          "epoch": 1105,
          "updates": 1105,
          "samples_seen": 1131520,
          "online_loss": 0.0006523407646454871,
          "train_loss": 0.000525153533089906,
          "validation_loss": 0.0004055248573422432,
          "train_metric": 5.2891537052346393e-05,
          "validation_metric": 4.0842969610821456e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.697265870001502
        },
        {
          "epoch": 1106,
          "updates": 1106,
          "samples_seen": 1132544,
          "online_loss": 0.0005251534748822451,
          "train_loss": 0.0006108134984970093,
          "validation_loss": 0.0004323215689510107,
          "train_metric": 6.15189055679366e-05,
          "validation_metric": 4.354184784460813e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.70243093099998
        },
        {
          "epoch": 1107,
          "updates": 1107,
          "samples_seen": 1133568,
          "online_loss": 0.0006108135567046702,
          "train_loss": 0.0007589889573864639,
          "validation_loss": 0.0006129867979325354,
          "train_metric": 7.644259312655777e-05,
          "validation_metric": 6.173778820084408e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.70818789200166
        },
        {
          "epoch": 1108,
          "updates": 1108,
          "samples_seen": 1134592,
          "online_loss": 0.0007589890155941248,
          "train_loss": 0.0008555800304748118,
          "validation_loss": 0.0005811947048641741,
          "train_metric": 8.617089042672887e-05,
          "validation_metric": 5.853580296388827e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.713466735000111
        },
        {
          "epoch": 1109,
          "updates": 1109,
          "samples_seen": 1135616,
          "online_loss": 0.0008555801468901336,
          "train_loss": 0.0007201050175353885,
          "validation_loss": 0.000584394030738622,
          "train_metric": 7.2526338044554e-05,
          "validation_metric": 5.885801510885358e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.719071158000588
        },
        {
          "epoch": 1110,
          "updates": 1110,
          "samples_seen": 1136640,
          "online_loss": 0.0007201050175353885,
          "train_loss": 0.0005840222002007067,
          "validation_loss": 0.0004394689458422363,
          "train_metric": 5.8820569392992184e-05,
          "validation_metric": 4.4261694711167365e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.724413700001605
        },
        {
          "epoch": 1111,
          "updates": 1111,
          "samples_seen": 1137664,
          "online_loss": 0.0005840221419930458,
          "train_loss": 0.0005181426531635225,
          "validation_loss": 0.00041983305709436536,
          "train_metric": 5.218543083174154e-05,
          "validation_metric": 4.228405305184424e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.730011257001024
        },
        {
          "epoch": 1112,
          "updates": 1112,
          "samples_seen": 1138688,
          "online_loss": 0.0005181426531635225,
          "train_loss": 0.0005625157500617206,
          "validation_loss": 0.0004670513153541833,
          "train_metric": 5.66545277251862e-05,
          "validation_metric": 4.703969898400828e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.735208616000818
        },
        {
          "epoch": 1113,
          "updates": 1113,
          "samples_seen": 1139712,
          "online_loss": 0.0005625158082693815,
          "train_loss": 0.0006578855100087821,
          "validation_loss": 0.00048313644947484136,
          "train_metric": 6.625982496188954e-05,
          "validation_metric": 4.86597309645731e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.741073454000798
        },
        {
          "epoch": 1114,
          "updates": 1114,
          "samples_seen": 1140736,
          "online_loss": 0.000657885626424104,
          "train_loss": 0.0006854824023321271,
          "validation_loss": 0.0005686357617378235,
          "train_metric": 6.903927715029567e-05,
          "validation_metric": 5.727091775042936e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.746791839001162
        },
        {
          "epoch": 1115,
          "updates": 1115,
          "samples_seen": 1141760,
          "online_loss": 0.0006854822277091444,
          "train_loss": 0.0006612495635636151,
          "validation_loss": 0.00047029092092998326,
          "train_metric": 6.659864448010921e-05,
          "validation_metric": 4.736597838927992e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.752184616001614
        },
        {
          "epoch": 1116,
          "updates": 1116,
          "samples_seen": 1142784,
          "online_loss": 0.0006612495635636151,
          "train_loss": 0.0005711359553970397,
          "validation_loss": 0.0004736571863759309,
          "train_metric": 5.752272409154102e-05,
          "validation_metric": 4.770501618622802e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.757405178001136
        },
        {
          "epoch": 1117,
          "updates": 1117,
          "samples_seen": 1143808,
          "online_loss": 0.0005711358971893787,
          "train_loss": 0.0005173153476789594,
          "validation_loss": 0.00040185663965530694,
          "train_metric": 5.210211020312272e-05,
          "validation_metric": 4.047352558700368e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.7635163190007
        },
        {
          "epoch": 1118,
          "updates": 1118,
          "samples_seen": 1144832,
          "online_loss": 0.0005173152894712985,
          "train_loss": 0.0005247793742455542,
          "validation_loss": 0.0004016666207462549,
          "train_metric": 5.2853858505841345e-05,
          "validation_metric": 4.045439345645718e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.769321296000271
        },
        {
          "epoch": 1119,
          "updates": 1119,
          "samples_seen": 1145856,
          "online_loss": 0.0005247793742455542,
          "train_loss": 0.0005686547374352813,
          "validation_loss": 0.00046830979408696294,
          "train_metric": 5.727282405132428e-05,
          "validation_metric": 4.7166449803626165e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.774745453001742
        },
        {
          "epoch": 1120,
          "updates": 1120,
          "samples_seen": 1146880,
          "online_loss": 0.0005686547374352813,
          "train_loss": 0.0006064728368073702,
          "validation_loss": 0.00044294120743870735,
          "train_metric": 6.108172237873077e-05,
          "validation_metric": 4.46114172518719e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.780079532001764
        },
        {
          "epoch": 1121,
          "updates": 1121,
          "samples_seen": 1147904,
          "online_loss": 0.0006064728368073702,
          "train_loss": 0.000590617535635829,
          "validation_loss": 0.00048655876889824867,
          "train_metric": 5.9484835219336674e-05,
          "validation_metric": 4.900441490462981e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.785542104000342
        },
        {
          "epoch": 1122,
          "updates": 1122,
          "samples_seen": 1148928,
          "online_loss": 0.0005906175938434899,
          "train_loss": 0.0005547189502976835,
          "validation_loss": 0.00042275807936675847,
          "train_metric": 5.586926272371784e-05,
          "validation_metric": 4.25786420237273e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.791034378000404
        },
        {
          "epoch": 1123,
          "updates": 1123,
          "samples_seen": 1149952,
          "online_loss": 0.0005547188920900226,
          "train_loss": 0.000517113774549216,
          "validation_loss": 0.00042208307422697544,
          "train_metric": 5.2081803005421534e-05,
          "validation_metric": 4.251066275173798e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.79622731100062
        },
        {
          "epoch": 1124,
          "updates": 1124,
          "samples_seen": 1150976,
          "online_loss": 0.000517113774549216,
          "train_loss": 0.0005078414687886834,
          "validation_loss": 0.0004094902833458036,
          "train_metric": 5.114792656968348e-05,
          "validation_metric": 4.1242357838200405e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.801343165001526
        },
        {
          "epoch": 1125,
          "updates": 1125,
          "samples_seen": 1152000,
          "online_loss": 0.0005078414687886834,
          "train_loss": 0.0005247495719231665,
          "validation_loss": 0.0004053451702930033,
          "train_metric": 5.285086444928311e-05,
          "validation_metric": 4.0824881580192596e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.806686470001296
        },
        {
          "epoch": 1126,
          "updates": 1126,
          "samples_seen": 1153024,
          "online_loss": 0.0005247495719231665,
          "train_loss": 0.0005457812803797424,
          "validation_loss": 0.00044607085874304175,
          "train_metric": 5.4969095799606293e-05,
          "validation_metric": 4.4926622649654746e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.812228222001067
        },
        {
          "epoch": 1127,
          "updates": 1127,
          "samples_seen": 1154048,
          "online_loss": 0.000545781513210386,
          "train_loss": 0.000555328675545752,
          "validation_loss": 0.000413857982493937,
          "train_metric": 5.593067180598155e-05,
          "validation_metric": 4.1682262235553935e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.817585684000733
        },
        {
          "epoch": 1128,
          "updates": 1128,
          "samples_seen": 1155072,
          "online_loss": 0.000555328675545752,
          "train_loss": 0.0005400212830863893,
          "validation_loss": 0.0004412956186570227,
          "train_metric": 5.4388965509133413e-05,
          "validation_metric": 4.444568185135722e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.823428628000329
        },
        {
          "epoch": 1129,
          "updates": 1129,
          "samples_seen": 1156096,
          "online_loss": 0.0005400212830863893,
          "train_loss": 0.0005189072107896209,
          "validation_loss": 0.0003978439199272543,
          "train_metric": 5.2262432291172445e-05,
          "validation_metric": 4.006938615930267e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.829356630001712
        },
        {
          "epoch": 1130,
          "updates": 1130,
          "samples_seen": 1157120,
          "online_loss": 0.0005189072107896209,
          "train_loss": 0.0005035869544371963,
          "validation_loss": 0.0004040980420541018,
          "train_metric": 5.07194381498266e-05,
          "validation_metric": 4.069927672389895e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.834799652000584
        },
        {
          "epoch": 1131,
          "updates": 1131,
          "samples_seen": 1158144,
          "online_loss": 0.0005035870126448572,
          "train_loss": 0.0005029639578424394,
          "validation_loss": 0.0004051018913742155,
          "train_metric": 5.065669392934069e-05,
          "validation_metric": 4.080038343090564e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.839952984000774
        },
        {
          "epoch": 1132,
          "updates": 1132,
          "samples_seen": 1159168,
          "online_loss": 0.0005029640160501003,
          "train_loss": 0.0005130066419951618,
          "validation_loss": 0.00040000563603825867,
          "train_metric": 5.166815026313998e-05,
          "validation_metric": 4.0287108276970685e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.85065118500097
        },
        {
          "epoch": 1133,
          "updates": 1133,
          "samples_seen": 1160192,
          "online_loss": 0.0005130066419951618,
          "train_loss": 0.0005228096270002425,
          "validation_loss": 0.00042881001718342304,
          "train_metric": 5.265547224553302e-05,
          "validation_metric": 4.3188170820940286e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.858779705000416
        },
        {
          "epoch": 1134,
          "updates": 1134,
          "samples_seen": 1161216,
          "online_loss": 0.0005228096270002425,
          "train_loss": 0.0005263753118924797,
          "validation_loss": 0.00040596918552182615,
          "train_metric": 5.301459168549627e-05,
          "validation_metric": 4.088773130206391e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.867289956000604
        },
        {
          "epoch": 1135,
          "updates": 1135,
          "samples_seen": 1162240,
          "online_loss": 0.0005263753118924797,
          "train_loss": 0.0005188863724470139,
          "validation_loss": 0.00042461458360776305,
          "train_metric": 5.226033681537956e-05,
          "validation_metric": 4.276562322047539e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.87463146600021
        },
        {
          "epoch": 1136,
          "updates": 1136,
          "samples_seen": 1163264,
          "online_loss": 0.0005188863724470139,
          "train_loss": 0.0005081527051515877,
          "validation_loss": 0.00039617379661649466,
          "train_metric": 5.1179278671042994e-05,
          "validation_metric": 3.990117693319917e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.883401693999986
        },
        {
          "epoch": 1137,
          "updates": 1137,
          "samples_seen": 1164288,
          "online_loss": 0.0005081526469439268,
          "train_loss": 0.0004987790598534048,
          "validation_loss": 0.0004014149308204651,
          "train_metric": 5.023520498070866e-05,
          "validation_metric": 4.042904038215056e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.891397548000896
        },
        {
          "epoch": 1138,
          "updates": 1138,
          "samples_seen": 1165312,
          "online_loss": 0.0004987790598534048,
          "train_loss": 0.00049554742872715,
          "validation_loss": 0.0003948249504901469,
          "train_metric": 4.990972229279578e-05,
          "validation_metric": 3.976532752858475e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.90091916700112
        },
        {
          "epoch": 1139,
          "updates": 1139,
          "samples_seen": 1166336,
          "online_loss": 0.00049554742872715,
          "train_loss": 0.0004980481462553144,
          "validation_loss": 0.0003905814082827419,
          "train_metric": 5.0161579565610737e-05,
          "validation_metric": 3.933793050237e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.01490333000038
        },
        {
          "epoch": 1140,
          "updates": 1140,
          "samples_seen": 1167360,
          "online_loss": 0.0004980481462553144,
          "train_loss": 0.0005026599392294884,
          "validation_loss": 0.00040656072087585926,
          "train_metric": 5.062606942374259e-05,
          "validation_metric": 4.0947306843008846e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.023180820000562
        },
        {
          "epoch": 1141,
          "updates": 1141,
          "samples_seen": 1168384,
          "online_loss": 0.0005026599392294884,
          "train_loss": 0.0005059438990429044,
          "validation_loss": 0.0003919019945897162,
          "train_metric": 5.095681262901053e-05,
          "validation_metric": 3.9470931369578466e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.031232833000104
        },
        {
          "epoch": 1142,
          "updates": 1142,
          "samples_seen": 1169408,
          "online_loss": 0.0005059438990429044,
          "train_loss": 0.0005048910970799625,
          "validation_loss": 0.0004106954438611865,
          "train_metric": 5.0850776460720226e-05,
          "validation_metric": 4.136373900109902e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.03810581200014
        },
        {
          "epoch": 1143,
          "updates": 1143,
          "samples_seen": 1170432,
          "online_loss": 0.0005048908060416579,
          "train_loss": 0.0005011153989471495,
          "validation_loss": 0.00039142052992247045,
          "train_metric": 5.0470505811972544e-05,
          "validation_metric": 3.942244438803755e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.046167631000571
        },
        {
          "epoch": 1144,
          "updates": 1144,
          "samples_seen": 1171456,
          "online_loss": 0.0005011153407394886,
          "train_loss": 0.0004956224584020674,
          "validation_loss": 0.00040127953980118036,
          "train_metric": 4.99172747367993e-05,
          "validation_metric": 4.041540523758158e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.053952915000991
        },
        {
          "epoch": 1145,
          "updates": 1145,
          "samples_seen": 1172480,
          "online_loss": 0.0004956224001944065,
          "train_loss": 0.0004912373842671514,
          "validation_loss": 0.00038947572465986013,
          "train_metric": 4.9475638661533594e-05,
          "validation_metric": 3.922656469512731e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.063343313000587
        },
        {
          "epoch": 1146,
          "updates": 1146,
          "samples_seen": 1173504,
          "online_loss": 0.0004912374424748123,
          "train_loss": 0.0004890579148195684,
          "validation_loss": 0.0003905762860085815,
          "train_metric": 4.9256130296271294e-05,
          "validation_metric": 3.933741390937939e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.071511729000122
        },
        {
          "epoch": 1147,
          "updates": 1147,
          "samples_seen": 1174528,
          "online_loss": 0.0004890580312348902,
          "train_loss": 0.0004891252610832453,
          "validation_loss": 0.0003923048498108983,
          "train_metric": 4.926290057483129e-05,
          "validation_metric": 3.951150574721396e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.079335879001519
        },
        {
          "epoch": 1148,
          "updates": 1148,
          "samples_seen": 1175552,
          "online_loss": 0.0004891252610832453,
          "train_loss": 0.0004904764355160296,
          "validation_loss": 0.00038603832945227623,
          "train_metric": 4.9398982810089365e-05,
          "validation_metric": 3.888036735588685e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.087817642001028
        },
        {
          "epoch": 1149,
          "updates": 1149,
          "samples_seen": 1176576,
          "online_loss": 0.0004904764355160296,
          "train_loss": 0.0004916677135042846,
          "validation_loss": 0.00039675107109360397,
          "train_metric": 4.9518970627104864e-05,
          "validation_metric": 3.995931911049411e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.096108121000725
        },
        {
          "epoch": 1150,
          "updates": 1150,
          "samples_seen": 1177600,
          "online_loss": 0.0004916675970889628,
          "train_loss": 0.0004919504281133413,
          "validation_loss": 0.0003848500200547278,
          "train_metric": 4.954744508722797e-05,
          "validation_metric": 3.8760681491112337e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.104699999001241
        },
        {
          "epoch": 1151,
          "updates": 1151,
          "samples_seen": 1178624,
          "online_loss": 0.0004919504863210022,
          "train_loss": 0.0004906926769763231,
          "validation_loss": 0.00039673352148383856,
          "train_metric": 4.942076702718623e-05,
          "validation_metric": 3.995754013885744e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.112934928001778
        },
        {
          "epoch": 1152,
          "updates": 1152,
          "samples_seen": 1179648,
          "online_loss": 0.0004906926769763231,
          "train_loss": 0.0004885996459051967,
          "validation_loss": 0.0003843975136987865,
          "train_metric": 4.9209971621166915e-05,
          "validation_metric": 3.8715112168574706e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.121457029001249
        },
        {
          "epoch": 1153,
          "updates": 1153,
          "samples_seen": 1180672,
          "online_loss": 0.0004885996459051967,
          "train_loss": 0.00048603402683511376,
          "validation_loss": 0.0003924999909941107,
          "train_metric": 4.8951558710541576e-05,
          "validation_metric": 3.953116538468748e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.129396070000439
        },
        {
          "epoch": 1154,
          "updates": 1154,
          "samples_seen": 1181696,
          "online_loss": 0.0004860339977312833,
          "train_loss": 0.0004837859305553138,
          "validation_loss": 0.00038463925011456013,
          "train_metric": 4.872515273746103e-05,
          "validation_metric": 3.873946116073057e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.137069016000169
        },
        {
          "epoch": 1155,
          "updates": 1155,
          "samples_seen": 1182720,
          "online_loss": 0.0004837859305553138,
          "train_loss": 0.00048215934657491744,
          "validation_loss": 0.0003872064989991486,
          "train_metric": 4.8561327275820076e-05,
          "validation_metric": 3.8998015952529386e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.14482343200143
        },
        {
          "epoch": 1156,
          "updates": 1156,
          "samples_seen": 1183744,
          "online_loss": 0.000482159317471087,
          "train_loss": 0.00048124577733688056,
          "validation_loss": 0.00038583960849791765,
          "train_metric": 4.846931551583111e-05,
          "validation_metric": 3.886035119649023e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.152561364000576
        },
        {
          "epoch": 1157,
          "updates": 1157,
          "samples_seen": 1184768,
          "online_loss": 0.000481245806440711,
          "train_loss": 0.00048090226482599974,
          "validation_loss": 0.0003831196518149227,
          "train_metric": 4.843471833737567e-05,
          "validation_metric": 3.8586407754337415e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.161442837001232
        },
        {
          "epoch": 1158,
          "updates": 1158,
          "samples_seen": 1185792,
          "online_loss": 0.00048090226482599974,
          "train_loss": 0.00048083733418025076,
          "validation_loss": 0.0003871876106131822,
          "train_metric": 4.842817361350171e-05,
          "validation_metric": 3.899612056557089e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.169308266000371
        },
        {
          "epoch": 1159,
          "updates": 1159,
          "samples_seen": 1186816,
          "online_loss": 0.00048083733418025076,
          "train_loss": 0.0004808112280443311,
          "validation_loss": 0.00038052842137403786,
          "train_metric": 4.8425543354824185e-05,
          "validation_metric": 3.8325433706631884e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.177864370001771
        },
        {
          "epoch": 1160,
          "updates": 1160,
          "samples_seen": 1187840,
          "online_loss": 0.00048081116983667016,
          "train_loss": 0.00048053235514089465,
          "validation_loss": 0.00038752483669668436,
          "train_metric": 4.839745815843344e-05,
          "validation_metric": 3.9030088373692706e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.185829206000562
        },
        {
          "epoch": 1161,
          "updates": 1161,
          "samples_seen": 1188864,
          "online_loss": 0.00048053235514089465,
          "train_loss": 0.0004799978050868958,
          "validation_loss": 0.0003790954942815006,
          "train_metric": 4.834361971006729e-05,
          "validation_metric": 3.8181104173418134e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.194531190001726
        },
        {
          "epoch": 1162,
          "updates": 1162,
          "samples_seen": 1189888,
          "online_loss": 0.00047999777598306537,
          "train_loss": 0.00047910131979733706,
          "validation_loss": 0.00038659217534586787,
          "train_metric": 4.8253332352032885e-05,
          "validation_metric": 3.8936148484935984e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.202686839000307
        },
        {
          "epoch": 1163,
          "updates": 1163,
          "samples_seen": 1190912,
          "online_loss": 0.0004791012906935066,
          "train_loss": 0.00047802517656236887,
          "validation_loss": 0.0003783393185585737,
          "train_metric": 4.814494968741201e-05,
          "validation_metric": 3.810495400102809e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.211434503000419
        },
        {
          "epoch": 1164,
          "updates": 1164,
          "samples_seen": 1191936,
          "online_loss": 0.0004780251474585384,
          "train_loss": 0.0004767837526742369,
          "validation_loss": 0.00038442076765932143,
          "train_metric": 4.8019915993791074e-05,
          "validation_metric": 3.871745502692647e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.21926366200023
        },
        {
          "epoch": 1165,
          "updates": 1165,
          "samples_seen": 1192960,
          "online_loss": 0.00047678378177806735,
          "train_loss": 0.0004755458212457597,
          "validation_loss": 0.0003776201920118183,
          "train_metric": 4.7895235184114426e-05,
          "validation_metric": 3.8032521842978895e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.228076313000201
        },
        {
          "epoch": 1166,
          "updates": 1166,
          "samples_seen": 1193984,
          "online_loss": 0.0004755458212457597,
          "train_loss": 0.00047433824511244893,
          "validation_loss": 0.00038160235271789134,
          "train_metric": 4.777361391461454e-05,
          "validation_metric": 3.843359081656672e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.23598963600125
        },
        {
          "epoch": 1167,
          "updates": 1167,
          "samples_seen": 1195008,
          "online_loss": 0.0004743382742162794,
          "train_loss": 0.00047323177568614483,
          "validation_loss": 0.00037690912722609937,
          "train_metric": 4.766217534779571e-05,
          "validation_metric": 3.7960908230161294e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.24453580700174
        },
        {
          "epoch": 1168,
          "updates": 1168,
          "samples_seen": 1196032,
          "online_loss": 0.00047323183389380574,
          "train_loss": 0.0004722304583992809,
          "validation_loss": 0.00037869653897359967,
          "train_metric": 4.7561323299305514e-05,
          "validation_metric": 3.8140926335472614e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.252458475000822
        },
        {
          "epoch": 1169,
          "updates": 1169,
          "samples_seen": 1197056,
          "online_loss": 0.00047223042929545045,
          "train_loss": 0.0004713415401056409,
          "validation_loss": 0.0003762126434594393,
          "train_metric": 4.7471796278841794e-05,
          "validation_metric": 3.789075344684534e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.261242213000514
        },
        {
          "epoch": 1170,
          "updates": 1170,
          "samples_seen": 1198080,
          "online_loss": 0.0004713415401056409,
          "train_loss": 0.0004705425526481122,
          "validation_loss": 0.0003760965773835778,
          "train_metric": 4.739132054965012e-05,
          "validation_metric": 3.787907553487457e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.26982992700141
        },
        {
          "epoch": 1171,
          "updates": 1171,
          "samples_seen": 1199104,
          "online_loss": 0.00047054258175194263,
          "train_loss": 0.0004698150441981852,
          "validation_loss": 0.0003757852246053517,
          "train_metric": 4.7318048018496484e-05,
          "validation_metric": 3.7847712519578636e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.278749138000421
        },
        {
          "epoch": 1172,
          "updates": 1172,
          "samples_seen": 1200128,
          "online_loss": 0.0004698150441981852,
          "train_loss": 0.00046913535334169865,
          "validation_loss": 0.0003742525295820087,
          "train_metric": 4.7249595809262246e-05,
          "validation_metric": 3.769334580283612e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.287520515001233
        },
        {
          "epoch": 1173,
          "updates": 1173,
          "samples_seen": 1201152,
          "online_loss": 0.00046913529513403773,
          "train_loss": 0.0004684953310061246,
          "validation_loss": 0.0003756596997845918,
          "train_metric": 4.718513446277939e-05,
          "validation_metric": 3.7835066905245185e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.29597532400112
        },
        {
          "epoch": 1174,
          "updates": 1174,
          "samples_seen": 1202176,
          "online_loss": 0.00046849530190229416,
          "train_loss": 0.00046787969768047333,
          "validation_loss": 0.00037287722807377577,
          "train_metric": 4.712312875199132e-05,
          "validation_metric": 3.75548297597561e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.304954957000518
        },
        {
          "epoch": 1175,
          "updates": 1175,
          "samples_seen": 1203200,
          "online_loss": 0.0004678790282923728,
          "train_loss": 0.00046728586312383413,
          "validation_loss": 0.00037546904059126973,
          "train_metric": 4.706332038040273e-05,
          "validation_metric": 3.781586201512255e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.313673811000626
        },
        {
          "epoch": 1176,
          "updates": 1176,
          "samples_seen": 1204224,
          "online_loss": 0.00046728586312383413,
          "train_loss": 0.00046671810559928417,
          "validation_loss": 0.0003715385973919183,
          "train_metric": 4.700613862951286e-05,
          "validation_metric": 3.742000990314409e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.322176069001216
        },
        {
          "epoch": 1177,
          "updates": 1177,
          "samples_seen": 1205248,
          "online_loss": 0.00046671810559928417,
          "train_loss": 0.0004661712155211717,
          "validation_loss": 0.00037510410766117275,
          "train_metric": 4.695105963037349e-05,
          "validation_metric": 3.77791075152345e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.330244577000485
        },
        {
          "epoch": 1178,
          "updates": 1178,
          "samples_seen": 1206272,
          "online_loss": 0.0004661712737288326,
          "train_loss": 0.0004656808450818062,
          "validation_loss": 0.00037002330645918846,
          "train_metric": 4.690166679210961e-05,
          "validation_metric": 3.72674003301654e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.339136860000508
        },
        {
          "epoch": 1179,
          "updates": 1179,
          "samples_seen": 1207296,
          "online_loss": 0.0004656808159779757,
          "train_loss": 0.0004652368661481887,
          "validation_loss": 0.00037473515840247273,
          "train_metric": 4.6856952394591644e-05,
          "validation_metric": 3.774195647565648e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.446587954000279
        },
        {
          "epoch": 1180,
          "updates": 1180,
          "samples_seen": 1208320,
          "online_loss": 0.0004652368661481887,
          "train_loss": 0.00046491133980453014,
          "validation_loss": 0.00036844343412667513,
          "train_metric": 4.682416329160333e-05,
          "validation_metric": 3.71082714991644e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.45523975300057
        },
        {
          "epoch": 1181,
          "updates": 1181,
          "samples_seen": 1209344,
          "online_loss": 0.0004649113689083606,
          "train_loss": 0.00046470059896819293,
          "validation_loss": 0.0003748824237845838,
          "train_metric": 4.680294659920037e-05,
          "validation_metric": 3.7756784877274185e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.463156352001533
        },
        {
          "epoch": 1182,
          "updates": 1182,
          "samples_seen": 1210368,
          "online_loss": 0.0004647006280720234,
          "train_loss": 0.0004647431196644902,
          "validation_loss": 0.00036714968155138195,
          "train_metric": 4.680722486227751e-05,
          "validation_metric": 3.6977973650209606e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.471612119001293
        },
        {
          "epoch": 1183,
          "updates": 1183,
          "samples_seen": 1211392,
          "online_loss": 0.0004647431196644902,
          "train_loss": 0.0004650347400456667,
          "validation_loss": 0.0003760948311537504,
          "train_metric": 4.683660154114477e-05,
          "validation_metric": 3.7878897273913026e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.47976233100053
        },
        {
          "epoch": 1184,
          "updates": 1184,
          "samples_seen": 1212416,
          "online_loss": 0.00046503476914949715,
          "train_loss": 0.0004659255500882864,
          "validation_loss": 0.00036636850563809276,
          "train_metric": 4.692631773650646e-05,
          "validation_metric": 3.689929508254863e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.488688824001656
        },
        {
          "epoch": 1185,
          "updates": 1185,
          "samples_seen": 1213440,
          "online_loss": 0.00046592552098445594,
          "train_loss": 0.00046738958917558193,
          "validation_loss": 0.000379278848413378,
          "train_metric": 4.7073768655536696e-05,
          "validation_metric": 3.819958146777935e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.497315052001795
        },
        {
          "epoch": 1186,
          "updates": 1186,
          "samples_seen": 1214464,
          "online_loss": 0.0004673896182794124,
          "train_loss": 0.0004703104787040502,
          "validation_loss": 0.0003668500285129994,
          "train_metric": 4.736795017379336e-05,
          "validation_metric": 3.694778570206836e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.505630430001474
        },
        {
          "epoch": 1187,
          "updates": 1187,
          "samples_seen": 1215488,
          "online_loss": 0.00047031050780788064,
          "train_loss": 0.00047455180902034044,
          "validation_loss": 0.0003866229089908302,
          "train_metric": 4.779511800734326e-05,
          "validation_metric": 3.893924440490082e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.51375829600147
        },
        {
          "epoch": 1188,
          "updates": 1188,
          "samples_seen": 1216512,
          "online_loss": 0.00047455180902034044,
          "train_loss": 0.0004825882497243583,
          "validation_loss": 0.0003709583543241024,
          "train_metric": 4.8604524636175483e-05,
          "validation_metric": 3.7361562135629356e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.521528204000788
        },
        {
          "epoch": 1189,
          "updates": 1189,
          "samples_seen": 1217536,
          "online_loss": 0.00048258822062052786,
          "train_loss": 0.000493546889629215,
          "validation_loss": 0.0004033244331367314,
          "train_metric": 4.970824011252262e-05,
          "validation_metric": 4.0621362131787464e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.529396768000879
        },
        {
          "epoch": 1190,
          "updates": 1190,
          "samples_seen": 1218560,
          "online_loss": 0.000493546889629215,
          "train_loss": 0.000514829414896667,
          "validation_loss": 0.0003857785777654499,
          "train_metric": 5.185172994970344e-05,
          "validation_metric": 3.885420665028505e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.5368432940013
        },
        {
          "epoch": 1191,
          "updates": 1191,
          "samples_seen": 1219584,
          "online_loss": 0.0005148293566890061,
          "train_loss": 0.0005415985942818224,
          "validation_loss": 0.00044179256656207144,
          "train_metric": 5.454782512970269e-05,
          "validation_metric": 4.4495725887827575e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.544755399001588
        },
        {
          "epoch": 1192,
          "updates": 1192,
          "samples_seen": 1220608,
          "online_loss": 0.0005415986524894834,
          "train_loss": 0.0005969011108390987,
          "validation_loss": 0.0004298588028177619,
          "train_metric": 6.011769437463954e-05,
          "validation_metric": 4.3293803173583e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.552461125000264
        },
        {
          "epoch": 1193,
          "updates": 1193,
          "samples_seen": 1221632,
          "online_loss": 0.0005969011690467596,
          "train_loss": 0.0006556124426424503,
          "validation_loss": 0.0005272681592032313,
          "train_metric": 6.603088695555925e-05,
          "validation_metric": 5.310451888362877e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.560303103000479
        },
        {
          "epoch": 1194,
          "updates": 1194,
          "samples_seen": 1222656,
          "online_loss": 0.0006556124426424503,
          "train_loss": 0.0007879160111770034,
          "validation_loss": 0.0005410934099927545,
          "train_metric": 7.935602479847148e-05,
          "validation_metric": 5.44969443581067e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.567984129000251
        },
        {
          "epoch": 1195,
          "updates": 1195,
          "samples_seen": 1223680,
          "online_loss": 0.0007879160693846643,
          "train_loss": 0.0008785460377112031,
          "validation_loss": 0.0006869803764857352,
          "train_metric": 8.848393918015063e-05,
          "validation_metric": 6.919013685546815e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.576069370001278
        },
        {
          "epoch": 1196,
          "updates": 1196,
          "samples_seen": 1224704,
          "online_loss": 0.000878546095918864,
          "train_loss": 0.0011133832158520818,
          "validation_loss": 0.0007383015472441912,
          "train_metric": 0.00011213588004466146,
          "validation_metric": 7.435902080032974e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.583915522000098
        },
        {
          "epoch": 1197,
          "updates": 1197,
          "samples_seen": 1225728,
          "online_loss": 0.0011133833322674036,
          "train_loss": 0.0011127123143523932,
          "validation_loss": 0.0008493346394971013,
          "train_metric": 0.00011206830822629854,
          "validation_metric": 8.554186933906749e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.591764249000335
        },
        {
          "epoch": 1198,
          "updates": 1198,
          "samples_seen": 1226752,
          "online_loss": 0.0011127123143523932,
          "train_loss": 0.0012543312041088939,
          "validation_loss": 0.0008210506057366729,
          "train_metric": 0.00012633163714781404,
          "validation_metric": 8.269319368992001e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.59952888800035
        },
        {
          "epoch": 1199,
          "updates": 1199,
          "samples_seen": 1227776,
          "online_loss": 0.001254331087693572,
          "train_loss": 0.0009803958237171173,
          "validation_loss": 0.0007505550747737288,
          "train_metric": 9.87418825388886e-05,
          "validation_metric": 7.559316145488992e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.607343799001683
        },
        {
          "epoch": 1200,
          "updates": 1200,
          "samples_seen": 1228800,
          "online_loss": 0.0009803959401324391,
          "train_loss": 0.0007958422647789121,
          "validation_loss": 0.0005333485314622521,
          "train_metric": 8.015431376406923e-05,
          "validation_metric": 5.371691440814175e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.615278663000936
        },
        {
          "epoch": 1201,
          "updates": 1201,
          "samples_seen": 1229824,
          "online_loss": 0.0007958421483635902,
          "train_loss": 0.0005521299317479134,
          "validation_loss": 0.00043552476563490927,
          "train_metric": 5.5608492402825505e-05,
          "validation_metric": 4.386445652926341e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.622708991000763
        },
        {
          "epoch": 1202,
          "updates": 1202,
          "samples_seen": 1230848,
          "online_loss": 0.0005521299317479134,
          "train_loss": 0.0004534095642156899,
          "validation_loss": 0.00034928316017612815,
          "train_metric": 4.566574716591276e-05,
          "validation_metric": 3.517852019285783e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.63124839000011
        },
        {
          "epoch": 1203,
          "updates": 1203,
          "samples_seen": 1231872,
          "online_loss": 0.0004534095642156899,
          "train_loss": 0.0004884613445028663,
          "validation_loss": 0.000369668792700395,
          "train_metric": 4.919604543829337e-05,
          "validation_metric": 3.723168629221618e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.639299712000138
        },
        {
          "epoch": 1204,
          "updates": 1204,
          "samples_seen": 1232896,
          "online_loss": 0.0004884614027105272,
          "train_loss": 0.0006056923302821815,
          "validation_loss": 0.0004909002454951406,
          "train_metric": 6.1003112932667136e-05,
          "validation_metric": 4.944166721543297e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.647501527000713
        },
        {
          "epoch": 1205,
          "updates": 1205,
          "samples_seen": 1233920,
          "online_loss": 0.0006056923302821815,
          "train_loss": 0.0007759132422506809,
          "validation_loss": 0.0005445850547403097,
          "train_metric": 7.814715354470536e-05,
          "validation_metric": 5.484860957949422e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.65517598200131
        },
        {
          "epoch": 1206,
          "updates": 1206,
          "samples_seen": 1234944,
          "online_loss": 0.0007759133586660028,
          "train_loss": 0.0008310142438858747,
          "validation_loss": 0.0006597677711397409,
          "train_metric": 8.369670831598341e-05,
          "validation_metric": 6.644938548561186e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.663377881000997
        },
        {
          "epoch": 1207,
          "updates": 1207,
          "samples_seen": 1235968,
          "online_loss": 0.0008310141856782138,
          "train_loss": 0.0009005595929920673,
          "validation_loss": 0.0006061754538677633,
          "train_metric": 9.07010689843446e-05,
          "validation_metric": 6.105176726123318e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.67145671400067
        },
        {
          "epoch": 1208,
          "updates": 1208,
          "samples_seen": 1236992,
          "online_loss": 0.0009005595929920673,
          "train_loss": 0.0007486989488825202,
          "validation_loss": 0.000595257559325546,
          "train_metric": 7.540622027590871e-05,
          "validation_metric": 5.995215178700164e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.679573314000663
        },
        {
          "epoch": 1209,
          "updates": 1209,
          "samples_seen": 1238016,
          "online_loss": 0.0007486990070901811,
          "train_loss": 0.0006183963851071894,
          "validation_loss": 0.00042723145452328026,
          "train_metric": 6.228261918295175e-05,
          "validation_metric": 4.302918750909157e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.68702913800007
        },
        {
          "epoch": 1210,
          "updates": 1210,
          "samples_seen": 1239040,
          "online_loss": 0.0006183964433148503,
          "train_loss": 0.0004846436495427042,
          "validation_loss": 0.00038529056473635137,
          "train_metric": 4.881153654423542e-05,
          "validation_metric": 3.8805053918622434e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.694819840000491
        },
        {
          "epoch": 1211,
          "updates": 1211,
          "samples_seen": 1240064,
          "online_loss": 0.0004846436786465347,
          "train_loss": 0.0004495670727919787,
          "validation_loss": 0.00035324040800333023,
          "train_metric": 4.527874989435077e-05,
          "validation_metric": 3.5577078961068764e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.702259556001081
        },
        {
          "epoch": 1212,
          "updates": 1212,
          "samples_seen": 1241088,
          "online_loss": 0.0004495671601034701,
          "train_loss": 0.0005034357309341431,
          "validation_loss": 0.00037688546581193805,
          "train_metric": 5.07042022945825e-05,
          "validation_metric": 3.795852171606384e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.709819335001157
        },
        {
          "epoch": 1213,
          "updates": 1213,
          "samples_seen": 1242112,
          "online_loss": 0.0005034357309341431,
          "train_loss": 0.0005863127880729735,
          "validation_loss": 0.00047452235594391823,
          "train_metric": 5.905128637095913e-05,
          "validation_metric": 4.779215669259429e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.717242682001597
        },
        {
          "epoch": 1214,
          "updates": 1214,
          "samples_seen": 1243136,
          "online_loss": 0.0005863128462806344,
          "train_loss": 0.0006725527346134186,
          "validation_loss": 0.0004909621202386916,
          "train_metric": 6.773704080842435e-05,
          "validation_metric": 4.944791362504475e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.724986710001758
        },
        {
          "epoch": 1215,
          "updates": 1215,
          "samples_seen": 1244160,
          "online_loss": 0.000672552443575114,
          "train_loss": 0.0006589710828848183,
          "validation_loss": 0.0005347104743123055,
          "train_metric": 6.636916077695787e-05,
          "validation_metric": 5.385408439906314e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.733015856001657
        },
        {
          "epoch": 1216,
          "updates": 1216,
          "samples_seen": 1245184,
          "online_loss": 0.0006589711410924792,
          "train_loss": 0.000631301780231297,
          "validation_loss": 0.00045680691255256534,
          "train_metric": 6.358239625114948e-05,
          "validation_metric": 4.600791726261377e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.741168876000302
        },
        {
          "epoch": 1217,
          "updates": 1217,
          "samples_seen": 1246208,
          "online_loss": 0.0006313016638159752,
          "train_loss": 0.0005401275120675564,
          "validation_loss": 0.0004447867104317993,
          "train_metric": 5.439966480480507e-05,
          "validation_metric": 4.479728886508383e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.748713200000566
        },
        {
          "epoch": 1218,
          "updates": 1218,
          "samples_seen": 1247232,
          "online_loss": 0.0005401275120675564,
          "train_loss": 0.0004738116404041648,
          "validation_loss": 0.00035573745844885707,
          "train_metric": 4.7720572183607146e-05,
          "validation_metric": 3.582856879802421e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.84401450400037
        },
        {
          "epoch": 1219,
          "updates": 1219,
          "samples_seen": 1248256,
          "online_loss": 0.0004738116404041648,
          "train_loss": 0.0004421171033754945,
          "validation_loss": 0.0003499231534078717,
          "train_metric": 4.45284167653881e-05,
          "validation_metric": 3.524297790136188e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.85202827500143
        },
        {
          "epoch": 1220,
          "updates": 1220,
          "samples_seen": 1249280,
          "online_loss": 0.00044211713247932494,
          "train_loss": 0.0004537542990874499,
          "validation_loss": 0.0003644040261860937,
          "train_metric": 4.570047167362645e-05,
          "validation_metric": 3.6701436329167336e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.857562978000715
        },
        {
          "epoch": 1221,
          "updates": 1221,
          "samples_seen": 1250304,
          "online_loss": 0.00045375432819128036,
          "train_loss": 0.0004908616538159549,
          "validation_loss": 0.00036275156890042126,
          "train_metric": 4.9437781854067e-05,
          "validation_metric": 3.6535006074700505e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.86340651500177
        },
        {
          "epoch": 1222,
          "updates": 1222,
          "samples_seen": 1251328,
          "online_loss": 0.0004908616538159549,
          "train_loss": 0.0005199082661420107,
          "validation_loss": 0.0004233973450027406,
          "train_metric": 5.23632625117898e-05,
          "validation_metric": 4.264302697265521e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.868767645000844
        }
      ],
      "best": {
        "epoch": 1202,
        "updates": 1202
      },
      "best_validation_metric": 3.517852019285783e-05,
      "config": {
        "label": "Full batch",
        "batch": 1024,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.00593114830305716
    },
    "9d4c0f0efefb_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7652762699872255,
          "train_loss": 0.42042335867881775,
          "validation_loss": 0.2850041389465332,
          "train_metric": 0.04234350472688675,
          "validation_metric": 0.0287045706063509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.020139460000791587
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.22368359565734863,
          "train_loss": 0.13688752055168152,
          "validation_loss": 0.1255667358636856,
          "train_metric": 0.013786809518933296,
          "validation_metric": 0.012646622955799103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03855141300118703
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.11294215638190508,
          "train_loss": 0.08564957976341248,
          "validation_loss": 0.07503470033407211,
          "train_metric": 0.008626311086118221,
          "validation_metric": 0.007557219825685024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05712496300111525
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.0772611484862864,
          "train_loss": 0.0631377175450325,
          "validation_loss": 0.05167700722813606,
          "train_metric": 0.006359000224620104,
          "validation_metric": 0.0052047185599803925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07475882600010664
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.061356538673862815,
          "train_loss": 0.05692056939005852,
          "validation_loss": 0.05108335614204407,
          "train_metric": 0.005732830613851547,
          "validation_metric": 0.005144929047673941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0931782679999742
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.05140789214055985,
          "train_loss": 0.04872747138142586,
          "validation_loss": 0.04551690071821213,
          "train_metric": 0.004907652735710144,
          "validation_metric": 0.004584295675158501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11154592100137961
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.046517667127773166,
          "train_loss": 0.04431310296058655,
          "validation_loss": 0.042578257620334625,
          "train_metric": 0.004463053308427334,
          "validation_metric": 0.004288326483219862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1292116590011574
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04272751940879971,
          "train_loss": 0.03902768716216087,
          "validation_loss": 0.03809235244989395,
          "train_metric": 0.003930726088583469,
          "validation_metric": 0.003836521878838539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1463869489998615
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.03780989674851298,
          "train_loss": 0.03432001173496246,
          "validation_loss": 0.032857973128557205,
          "train_metric": 0.0034565855748951435,
          "validation_metric": 0.003309335093945265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1645147610015556
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.031064621172845364,
          "train_loss": 0.03347982466220856,
          "validation_loss": 0.03712276741862297,
          "train_metric": 0.003371965140104294,
          "validation_metric": 0.0037388689815998077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18136900200079253
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.031844150566030294,
          "train_loss": 0.028499653562903404,
          "validation_loss": 0.029775334522128105,
          "train_metric": 0.002870380412787199,
          "validation_metric": 0.002998862648382783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.199475239000094
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.02805367941618897,
          "train_loss": 0.026206981390714645,
          "validation_loss": 0.02643156796693802,
          "train_metric": 0.0026394713204354048,
          "validation_metric": 0.0026620905846357346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21779961500033096
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.025128397130174562,
          "train_loss": 0.024445120245218277,
          "validation_loss": 0.02698550373315811,
          "train_metric": 0.002462022937834263,
          "validation_metric": 0.002717881463468075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3521893930010265
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.024175066268071532,
          "train_loss": 0.02489951066672802,
          "validation_loss": 0.028000233694911003,
          "train_metric": 0.0025077874306589365,
          "validation_metric": 0.0028200806118547916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3716081270013092
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.02421150080044754,
          "train_loss": 0.03799642622470856,
          "validation_loss": 0.04064636677503586,
          "train_metric": 0.0038268608041107655,
          "validation_metric": 0.004093753173947334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38943587700123317
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.024743812944507226,
          "train_loss": 0.02022702433168888,
          "validation_loss": 0.023470282554626465,
          "train_metric": 0.0020371919963508844,
          "validation_metric": 0.00236384104937315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40942606099997647
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.02018856190261431,
          "train_loss": 0.020980266854166985,
          "validation_loss": 0.021665973588824272,
          "train_metric": 0.0021130559034645557,
          "validation_metric": 0.002182117197662592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4286989390002418
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.019975117058493197,
          "train_loss": 0.01583278365433216,
          "validation_loss": 0.018246788531541824,
          "train_metric": 0.0015946200583130121,
          "validation_metric": 0.0018377496162429452,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44751515999996627
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.018019937502685934,
          "train_loss": 0.018257061019539833,
          "validation_loss": 0.020841917023062706,
          "train_metric": 0.0018387841992080212,
          "validation_metric": 0.0020991216879338026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4684557400014455
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.016130378571688198,
          "train_loss": 0.014833329245448112,
          "validation_loss": 0.01784461922943592,
          "train_metric": 0.0014939584070816636,
          "validation_metric": 0.0017972447676584125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4888205220013333
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.016955066967057064,
          "train_loss": 0.02115985006093979,
          "validation_loss": 0.018467724323272705,
          "train_metric": 0.0021311426535248756,
          "validation_metric": 0.0018600014736875892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5093293750014709
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.015326663153246045,
          "train_loss": 0.01697344332933426,
          "validation_loss": 0.020991913974285126,
          "train_metric": 0.0017095031216740608,
          "validation_metric": 0.0021142286714166403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5282386779999797
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01588170396280475,
          "train_loss": 0.012263892218470573,
          "validation_loss": 0.014492574147880077,
          "train_metric": 0.001235174247995019,
          "validation_metric": 0.0014596389373764396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5476914210012183
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.01254026428796351,
          "train_loss": 0.011447018012404442,
          "validation_loss": 0.012168483808636665,
          "train_metric": 0.0011529017938300967,
          "validation_metric": 0.00122556509450078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5687155520008673
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.012048857679474168,
          "train_loss": 0.01243788842111826,
          "validation_loss": 0.014209525659680367,
          "train_metric": 0.0012526987120509148,
          "validation_metric": 0.0014311315026134253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5873051390008186
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.013750464626355097,
          "train_loss": 0.010670033283531666,
          "validation_loss": 0.012816783972084522,
          "train_metric": 0.0010746467160061002,
          "validation_metric": 0.0012908594217151403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6049290920000203
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.012473276641685516,
          "train_loss": 0.014790358953177929,
          "validation_loss": 0.016849016770720482,
          "train_metric": 0.0014896307839080691,
          "validation_metric": 0.0016969713615253568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6237631240001065
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.011476379659143277,
          "train_loss": 0.012335777282714844,
          "validation_loss": 0.014566931873559952,
          "train_metric": 0.0012424145825207233,
          "validation_metric": 0.0014671280514448881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6432453710003756
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.012046405303408392,
          "train_loss": 0.01248689740896225,
          "validation_loss": 0.012682496570050716,
          "train_metric": 0.0012576347216963768,
          "validation_metric": 0.0012773345224559307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6637872720002633
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.011282033665338531,
          "train_loss": 0.009161850437521935,
          "validation_loss": 0.010782088153064251,
          "train_metric": 0.0009227480040863156,
          "validation_metric": 0.0010859325993806124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6901235730001645
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.012668340845266357,
          "train_loss": 0.01739412546157837,
          "validation_loss": 0.01528378576040268,
          "train_metric": 0.001751872361637652,
          "validation_metric": 0.0015393270878121257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7090584180004953
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.0133912053424865,
          "train_loss": 0.014109708368778229,
          "validation_loss": 0.011819601990282536,
          "train_metric": 0.0014210781082510948,
          "validation_metric": 0.0011904268758371472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7307135290011502
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.011047368810977787,
          "train_loss": 0.009357995353639126,
          "validation_loss": 0.008963565342128277,
          "train_metric": 0.0009425030439160764,
          "validation_metric": 0.0009027774794958532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7434043290013506
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.012644548391108401,
          "train_loss": 0.012444963678717613,
          "validation_loss": 0.009428474120795727,
          "train_metric": 0.0012534112902358174,
          "validation_metric": 0.000949601293541491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7556844230002753
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.010892855127167422,
          "train_loss": 0.007740549743175507,
          "validation_loss": 0.0094137042760849,
          "train_metric": 0.0007795998244546354,
          "validation_metric": 0.0009481139713898301,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7674699060007697
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.011691184074152261,
          "train_loss": 0.00874636322259903,
          "validation_loss": 0.009288237430155277,
          "train_metric": 0.0008809016435407102,
          "validation_metric": 0.0009354772046208382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7793024640013755
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.012017411660053767,
          "train_loss": 0.009930570609867573,
          "validation_loss": 0.010600082576274872,
          "train_metric": 0.0010001707123592496,
          "validation_metric": 0.0010676016099750996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.790747064000243
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.010487652019946836,
          "train_loss": 0.007947818376123905,
          "validation_loss": 0.009073385037481785,
          "train_metric": 0.0008004752453416586,
          "validation_metric": 0.0009138380992226303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8025824750002357
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.008720675177755766,
          "train_loss": 0.008866284042596817,
          "validation_loss": 0.008035572245717049,
          "train_metric": 0.0008929796749725938,
          "validation_metric": 0.0008093133801594377,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8149221889998444
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.010380977859313134,
          "train_loss": 0.012286576442420483,
          "validation_loss": 0.008569505997002125,
          "train_metric": 0.0012374588986858726,
          "validation_metric": 0.0008630892261862755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8267277860013564
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010750436398666352,
          "train_loss": 0.007281068246811628,
          "validation_loss": 0.006733343005180359,
          "train_metric": 0.0007333225221373141,
          "validation_metric": 0.0006781577249057591,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8392488360004791
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.008277384302346036,
          "train_loss": 0.007872648537158966,
          "validation_loss": 0.008370221592485905,
          "train_metric": 0.0007929043495096266,
          "validation_metric": 0.0008430181187577546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8515129860006709
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.007766476846882142,
          "train_loss": 0.008425431326031685,
          "validation_loss": 0.007813382893800735,
          "train_metric": 0.0008485785219818354,
          "validation_metric": 0.0007869352702982724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.863558491000731
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.007163673566537909,
          "train_loss": 0.0058179195038974285,
          "validation_loss": 0.006076711229979992,
          "train_metric": 0.0005859595257788897,
          "validation_metric": 0.0006120241014286876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8756259970014071
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.011814785524620675,
          "train_loss": 0.0059189461171627045,
          "validation_loss": 0.006169567815959454,
          "train_metric": 0.0005961346323601902,
          "validation_metric": 0.0006213763263076544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8876481170009356
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.01034527469164459,
          "train_loss": 0.006979392841458321,
          "validation_loss": 0.006189743522554636,
          "train_metric": 0.000702938879840076,
          "validation_metric": 0.0006234082393348217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8996150490002037
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.006988057924900204,
          "train_loss": 0.006034561432898045,
          "validation_loss": 0.006086136680096388,
          "train_metric": 0.0006077789003029466,
          "validation_metric": 0.0006129734683781862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9125481940009195
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.006713563569064718,
          "train_loss": 0.011150868609547615,
          "validation_loss": 0.010636470280587673,
          "train_metric": 0.0011230746749788523,
          "validation_metric": 0.0010712662478908896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9252639670012286
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.009696730674477294,
          "train_loss": 0.006723904982209206,
          "validation_loss": 0.006750838831067085,
          "train_metric": 0.0006772070191800594,
          "validation_metric": 0.0006799197290092707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9380366559998947
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009931101929396391,
          "train_loss": 0.007089352235198021,
          "validation_loss": 0.005562353879213333,
          "train_metric": 0.0007140135858207941,
          "validation_metric": 0.0005602199817076325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9507772370016028
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.006415045660105534,
          "train_loss": 0.006310243159532547,
          "validation_loss": 0.006061024032533169,
          "train_metric": 0.0006355445366352797,
          "validation_metric": 0.0006104441708885133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9629136020012083
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0064065639016916975,
          "train_loss": 0.007004227954894304,
          "validation_loss": 0.006366886664181948,
          "train_metric": 0.0007054401212371886,
          "validation_metric": 0.0006412494694814086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9747803099999146
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.007003329184954055,
          "train_loss": 0.0064489999786019325,
          "validation_loss": 0.005463524721562862,
          "train_metric": 0.000649519613943994,
          "validation_metric": 0.000550266238860786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9872867640006007
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.007192634715465829,
          "train_loss": 0.009942920878529549,
          "validation_loss": 0.009080067276954651,
          "train_metric": 0.0010014144936576486,
          "validation_metric": 0.0009145111544057727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9993397230009577
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.012678841536398977,
          "train_loss": 0.010058409534394741,
          "validation_loss": 0.007470104843378067,
          "train_metric": 0.0010130461305379868,
          "validation_metric": 0.0007523616659455001,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.01156098499996
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0074841548630502075,
          "train_loss": 0.0088228490203619,
          "validation_loss": 0.007785570342093706,
          "train_metric": 0.0008886050782166421,
          "validation_metric": 0.0007841342012397945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.023690906000411
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.007798420439939946,
          "train_loss": 0.004910251125693321,
          "validation_loss": 0.0042976499535143375,
          "train_metric": 0.000494542473461479,
          "validation_metric": 0.0004328435461502522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0361783290009043
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.008617057064839173,
          "train_loss": 0.004746794234961271,
          "validation_loss": 0.004256707616150379,
          "train_metric": 0.00047807974624447525,
          "validation_metric": 0.0004287200281396508,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0485712740010058
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.005919049930525944,
          "train_loss": 0.004768840037286282,
          "validation_loss": 0.003860761411488056,
          "train_metric": 0.0004803001065738499,
          "validation_metric": 0.00038884172681719065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0617080700012593
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0045982688388903625,
          "train_loss": 0.005283898673951626,
          "validation_loss": 0.004355849698185921,
          "train_metric": 0.0005321749486029148,
          "validation_metric": 0.0004387052031233907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0742460270012089
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0055537172956974246,
          "train_loss": 0.005246427375823259,
          "validation_loss": 0.005048526916652918,
          "train_metric": 0.0005284009966999292,
          "validation_metric": 0.0005084691802039742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0866848320001736
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.005480020838149358,
          "train_loss": 0.0043077412992715836,
          "validation_loss": 0.0036414219066500664,
          "train_metric": 0.0004338599101174623,
          "validation_metric": 0.0003667506971396506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0990336049999314
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.00464105586434016,
          "train_loss": 0.004129167180508375,
          "validation_loss": 0.0033857915550470352,
          "train_metric": 0.0004158745869062841,
          "validation_metric": 0.00034100451739504933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1123941260011634
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0064185140072368085,
          "train_loss": 0.0041029006242752075,
          "validation_loss": 0.0037063572090119123,
          "train_metric": 0.00041322916513308883,
          "validation_metric": 0.00037329073529690504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1250633920008113
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.00585170355043374,
          "train_loss": 0.004980285186320543,
          "validation_loss": 0.004809828475117683,
          "train_metric": 0.0005015960778109729,
          "validation_metric": 0.00048442831030115485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1384221380012605
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005007249521440826,
          "train_loss": 0.007144916336983442,
          "validation_loss": 0.006564212031662464,
          "train_metric": 0.0007196097867563367,
          "validation_metric": 0.0006611233111470938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.16648751700086
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.004626823676517233,
          "train_loss": 0.004026313778012991,
          "validation_loss": 0.002965265652164817,
          "train_metric": 0.0004055155150126666,
          "validation_metric": 0.0002986506442539394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.17995703500128
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.006068435264751315,
          "train_loss": 0.003705250797793269,
          "validation_loss": 0.002774723106995225,
          "train_metric": 0.00037317926762625575,
          "validation_metric": 0.0002794599859043956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1927773309998884
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.009226883616065606,
          "train_loss": 0.005627352744340897,
          "validation_loss": 0.005550009198486805,
          "train_metric": 0.0005667663644999266,
          "validation_metric": 0.000558976549655199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2052350180001667
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0051749997364822775,
          "train_loss": 0.005725057329982519,
          "validation_loss": 0.005376391112804413,
          "train_metric": 0.0005766067770309746,
          "validation_metric": 0.0005414903862401843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2177808670003287
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.005788034679426346,
          "train_loss": 0.017407048493623734,
          "validation_loss": 0.015044525265693665,
          "train_metric": 0.0017531742341816425,
          "validation_metric": 0.0015152294654399157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2295587990010972
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.012850459286710247,
          "train_loss": 0.01382659561932087,
          "validation_loss": 0.012793240137398243,
          "train_metric": 0.0013925640378147364,
          "validation_metric": 0.0012884882744401693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2415740460000961
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.008258260786533356,
          "train_loss": 0.006541796028614044,
          "validation_loss": 0.0052823470905423164,
          "train_metric": 0.0006588657270185649,
          "validation_metric": 0.0005320186028257012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2538187360005395
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.005467409624543507,
          "train_loss": 0.005280505400151014,
          "validation_loss": 0.005102162249386311,
          "train_metric": 0.0005318331532180309,
          "validation_metric": 0.0005138710839673877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2657424520002678
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.004311830118240323,
          "train_loss": 0.00373004749417305,
          "validation_loss": 0.0026033874601125717,
          "train_metric": 0.00037567675462923944,
          "validation_metric": 0.00026220365543849766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2784840300009819
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.006244776726816781,
          "train_loss": 0.004198379814624786,
          "validation_loss": 0.003795134834945202,
          "train_metric": 0.0004228454199619591,
          "validation_metric": 0.0003822321305051446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.290902944001573
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.008559952409996185,
          "train_loss": 0.00789130199700594,
          "validation_loss": 0.007644448895007372,
          "train_metric": 0.0007947830017656088,
          "validation_metric": 0.0007699209381826222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3033798510005
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.010795303081977181,
          "train_loss": 0.007206970825791359,
          "validation_loss": 0.008518384769558907,
          "train_metric": 0.0007258597761392593,
          "validation_metric": 0.0008579405257478356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.315465605001009
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.008597618623753078,
          "train_loss": 0.007892834022641182,
          "validation_loss": 0.00703678373247385,
          "train_metric": 0.0007949374266900122,
          "validation_metric": 0.0007087191333994269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3285653729999467
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.00527198170311749,
          "train_loss": 0.003689192933961749,
          "validation_loss": 0.003060741350054741,
          "train_metric": 0.00037156202597543597,
          "validation_metric": 0.00030826666625216603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.340715258000273
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.004612095221091295,
          "train_loss": 0.0034868940711021423,
          "validation_loss": 0.0030472921207547188,
          "train_metric": 0.0003511871909722686,
          "validation_metric": 0.00030691211577504873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.352785685001436
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.004809458099771291,
          "train_loss": 0.004018434323370457,
          "validation_loss": 0.003810635767877102,
          "train_metric": 0.00040472199907526374,
          "validation_metric": 0.0003837932599708438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3646967160002532
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0037614972279698122,
          "train_loss": 0.0029776038136333227,
          "validation_loss": 0.002448973013088107,
          "train_metric": 0.0002998933778144419,
          "validation_metric": 0.00024665164528414607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.464806578000207
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0048097101171151735,
          "train_loss": 0.004213848151266575,
          "validation_loss": 0.004110181238502264,
          "train_metric": 0.000424403406213969,
          "validation_metric": 0.0004139624070376158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4777172030007932
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.004848333752306644,
          "train_loss": 0.0036085802130401134,
          "validation_loss": 0.0024455194361507893,
          "train_metric": 0.0003634429886005819,
          "validation_metric": 0.00024630376719869673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4908822380002675
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.0052853366414638,
          "train_loss": 0.008047995157539845,
          "validation_loss": 0.007545045576989651,
          "train_metric": 0.0008105646120384336,
          "validation_metric": 0.0007599094533361495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5035131390013703
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.007464153402906959,
          "train_loss": 0.007227043155580759,
          "validation_loss": 0.005922267213463783,
          "train_metric": 0.0007278811535798013,
          "validation_metric": 0.000596469035372138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5162747430003947
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.007580421995953657,
          "train_loss": 0.010719379410147667,
          "validation_loss": 0.008915221318602562,
          "train_metric": 0.0010796166025102139,
          "validation_metric": 0.0008979085250757635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5286717910003063
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005770513624156592,
          "train_loss": 0.003950092475861311,
          "validation_loss": 0.004198051057755947,
          "train_metric": 0.0003978388267569244,
          "validation_metric": 0.0004228123580105603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5415819420013577
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0034987453000212554,
          "train_loss": 0.002795945154502988,
          "validation_loss": 0.0021445685997605324,
          "train_metric": 0.00028159740031696856,
          "validation_metric": 0.0002159931173082441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5541778010010603
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.003033986593436566,
          "train_loss": 0.003139025066047907,
          "validation_loss": 0.002784611191600561,
          "train_metric": 0.0003161511558573693,
          "validation_metric": 0.00028045588987879455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5663371810005629
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0036182252515573055,
          "train_loss": 0.004737726878374815,
          "validation_loss": 0.00505541916936636,
          "train_metric": 0.00047716652625240386,
          "validation_metric": 0.0005091633065603673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5788782400013588
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0077437510844902135,
          "train_loss": 0.011963716708123684,
          "validation_loss": 0.011556046083569527,
          "train_metric": 0.0012049416545778513,
          "validation_metric": 0.0011638826690614223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5909920049998618
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0061263949355634395,
          "train_loss": 0.0026774860452860594,
          "validation_loss": 0.002582505578175187,
          "train_metric": 0.00026966657605953515,
          "validation_metric": 0.00026010049623437226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6073811000005662
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0031712435302324593,
          "train_loss": 0.003161311149597168,
          "validation_loss": 0.003632894018664956,
          "train_metric": 0.0003183957014698535,
          "validation_metric": 0.00036589178489521146,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6195683170008124
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.003476606110780267,
          "train_loss": 0.002484841737896204,
          "validation_loss": 0.0019491533748805523,
          "train_metric": 0.00025026415823958814,
          "validation_metric": 0.00019631159375421703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6325719570013462
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.003992638492491096,
          "train_loss": 0.006922051310539246,
          "validation_loss": 0.0043421718291938305,
          "train_metric": 0.0006971635739319026,
          "validation_metric": 0.00043732766062021255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.645303318000515
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.010024498609709553,
          "train_loss": 0.009600440040230751,
          "validation_loss": 0.008177990093827248,
          "train_metric": 0.0009669212158769369,
          "validation_metric": 0.000823657086584717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6575103550003405
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.007117582790669985,
          "train_loss": 0.002912071067839861,
          "validation_loss": 0.002165331970900297,
          "train_metric": 0.0002932931820396334,
          "validation_metric": 0.00021808428573422134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6702282460009883
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.005201683001359925,
          "train_loss": 0.007934290915727615,
          "validation_loss": 0.008846268057823181,
          "train_metric": 0.0007991126622073352,
          "validation_metric": 0.0008909638854674995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6823553170015657
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.003758432183531113,
          "train_loss": 0.0026037783827632666,
          "validation_loss": 0.0026130646001547575,
          "train_metric": 0.00026224309112876654,
          "validation_metric": 0.00026317828451283276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6943437210011325
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.003455892983765807,
          "train_loss": 0.003139335894957185,
          "validation_loss": 0.002716542687267065,
          "train_metric": 0.00031618241337127984,
          "validation_metric": 0.0002736002497840673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7071715130005032
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0038186703895917162,
          "train_loss": 0.0028373398818075657,
          "validation_loss": 0.0016404918860644102,
          "train_metric": 0.00028576646582223475,
          "validation_metric": 0.00016522433725185692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.723054210000555
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.002580409069196321,
          "train_loss": 0.002384669380262494,
          "validation_loss": 0.0024622250348329544,
          "train_metric": 0.00024017518444452435,
          "validation_metric": 0.0002479862596374005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7378018939998583
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.002946120763226645,
          "train_loss": 0.00438419496640563,
          "validation_loss": 0.0034322754945605993,
          "train_metric": 0.00044156002695672214,
          "validation_metric": 0.00034568621776998043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7501367370005028
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.003275718307122588,
          "train_loss": 0.0026692224200814962,
          "validation_loss": 0.0027403293643146753,
          "train_metric": 0.00026883429381996393,
          "validation_metric": 0.0002759959897957742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7620158420013468
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0029393051663646474,
          "train_loss": 0.002616056939586997,
          "validation_loss": 0.001844989019446075,
          "train_metric": 0.00026347965467721224,
          "validation_metric": 0.000185820521437563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7746933060006995
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.003027440556252259,
          "train_loss": 0.007916397415101528,
          "validation_loss": 0.00437186611816287,
          "train_metric": 0.0007973104948177934,
          "validation_metric": 0.0004403183120302856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7867917950006813
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.006687912144116126,
          "train_loss": 0.003483147593215108,
          "validation_loss": 0.002071235328912735,
          "train_metric": 0.00035080991801805794,
          "validation_metric": 0.00020860726363025606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7987820540001849
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.005686274922481971,
          "train_loss": 0.011589550413191319,
          "validation_loss": 0.009877367876470089,
          "train_metric": 0.0011672570835798979,
          "validation_metric": 0.0009948123479261994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8109944490006455
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.005210378189076437,
          "train_loss": 0.0035799299366772175,
          "validation_loss": 0.0023319467436522245,
          "train_metric": 0.0003605574311222881,
          "validation_metric": 0.0002348651032662019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8229456370008847
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.004197766429570038,
          "train_loss": 0.004317245911806822,
          "validation_loss": 0.00274511962197721,
          "train_metric": 0.0004348172224126756,
          "validation_metric": 0.0002764783857855946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8356718220002222
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.00357499254096183,
          "train_loss": 0.0025259973481297493,
          "validation_loss": 0.0019720508717000484,
          "train_metric": 0.0002544092130847275,
          "validation_metric": 0.00019861775217577815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8475494450012775
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.002768678157735849,
          "train_loss": 0.002665579319000244,
          "validation_loss": 0.0022789076901972294,
          "train_metric": 0.0002684674400370568,
          "validation_metric": 0.00022952321160119027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8593423180009268
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0026120798247575294,
          "train_loss": 0.00595118897035718,
          "validation_loss": 0.004451894201338291,
          "train_metric": 0.0005993819795548916,
          "validation_metric": 0.00044837850145995617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8712939440010814
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.004647571637178771,
          "train_loss": 0.009551094844937325,
          "validation_loss": 0.006961069069802761,
          "train_metric": 0.0009619513875804842,
          "validation_metric": 0.0007010934641584754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8832409360002202
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.005447974821436219,
          "train_loss": 0.0027361735701560974,
          "validation_loss": 0.0023015120532363653,
          "train_metric": 0.000275577389402315,
          "validation_metric": 0.00023179984418675303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8953775160007353
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.01009372684347909,
          "train_loss": 0.0031190747395157814,
          "validation_loss": 0.002818513661623001,
          "train_metric": 0.00031414179829880595,
          "validation_metric": 0.00028387035126797855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9079208160001144
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.004368423396954313,
          "train_loss": 0.0032084789127111435,
          "validation_loss": 0.0024119210429489613,
          "train_metric": 0.00032314626150764525,
          "validation_metric": 0.00024291985027957708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.920601401001477
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0032967130182441906,
          "train_loss": 0.0021714456379413605,
          "validation_loss": 0.0015058484859764576,
          "train_metric": 0.00021870005002710968,
          "validation_metric": 0.00015166355296969414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9335789470005693
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0031260423857020214,
          "train_loss": 0.0025659019593149424,
          "validation_loss": 0.0022219920065253973,
          "train_metric": 0.00025842824834398925,
          "validation_metric": 0.00022379089205060154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9462686290007696
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.009384249504364561,
          "train_loss": 0.016376744955778122,
          "validation_loss": 0.01678539253771305,
          "train_metric": 0.0016494058072566986,
          "validation_metric": 0.0016905635129660368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.958256120000442
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.008284934156108648,
          "train_loss": 0.0058981915935873985,
          "validation_loss": 0.0034257464576512575,
          "train_metric": 0.0005940442788414657,
          "validation_metric": 0.00034502861672081053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9707850480008346
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0057917209196602926,
          "train_loss": 0.004453660454601049,
          "validation_loss": 0.004415683448314667,
          "train_metric": 0.0004485563840717077,
          "validation_metric": 0.00044473150046542287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9834723000003578
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0030781236455368344,
          "train_loss": 0.0031056732404977083,
          "validation_loss": 0.002945525338873267,
          "train_metric": 0.00031279202084988356,
          "validation_metric": 0.00029666253249160945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.995833720000519
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0031512973182543647,
          "train_loss": 0.004320158623158932,
          "validation_loss": 0.0023350268602371216,
          "train_metric": 0.0004351105308160186,
          "validation_metric": 0.00023517533554695547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0084615760006272
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.002579562278697267,
          "train_loss": 0.0035990234464406967,
          "validation_loss": 0.003105646464973688,
          "train_metric": 0.0003624804667197168,
          "validation_metric": 0.0003127893141936511,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.020536564001304
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0025704377148940694,
          "train_loss": 0.001744432607665658,
          "validation_loss": 0.0013013780117034912,
          "train_metric": 0.0001756928686518222,
          "validation_metric": 0.00013107003178447485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.033148352000353
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0022302033812593436,
          "train_loss": 0.0020634885877370834,
          "validation_loss": 0.002004141453653574,
          "train_metric": 0.00020782700448762625,
          "validation_metric": 0.0002018497762037441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0461524260008446
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0023723116428300273,
          "train_loss": 0.0016037577297538519,
          "validation_loss": 0.0012292723404243588,
          "train_metric": 0.00016152457101270556,
          "validation_metric": 0.00012380778207443655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.059366999001213
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0056439107793266885,
          "train_loss": 0.002998527605086565,
          "validation_loss": 0.002283751033246517,
          "train_metric": 0.0003020006697624922,
          "validation_metric": 0.0002300110354553908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.072549180000351
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.002975222083478002,
          "train_loss": 0.0041765072382986546,
          "validation_loss": 0.00305955670773983,
          "train_metric": 0.0004206425219308585,
          "validation_metric": 0.0003081473405472934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.084909981000237
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.00392988315797993,
          "train_loss": 0.00201786239631474,
          "validation_loss": 0.0023296461440622807,
          "train_metric": 0.00020323174248915166,
          "validation_metric": 0.0002346334367757663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.097641428001225
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.002441780219669454,
          "train_loss": 0.0031751031056046486,
          "validation_loss": 0.0026954610366374254,
          "train_metric": 0.00031978473998606205,
          "validation_metric": 0.00027147697983309627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1100961310003186
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0031104303016036283,
          "train_loss": 0.0022488466929644346,
          "validation_loss": 0.002741734031587839,
          "train_metric": 0.0002264956128783524,
          "validation_metric": 0.00027613743441179395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1226148210007523
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0031105684029171243,
          "train_loss": 0.001962154172360897,
          "validation_loss": 0.0022357667330652475,
          "train_metric": 0.00019762097508646548,
          "validation_metric": 0.00022517822799272835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.135195251001278
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0024889320320653496,
          "train_loss": 0.00282339658588171,
          "validation_loss": 0.0032464037649333477,
          "train_metric": 0.00028436220600269735,
          "validation_metric": 0.00032696593552827835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1475716780005314
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.004280040582671063,
          "train_loss": 0.0032508892472833395,
          "validation_loss": 0.003358421614393592,
          "train_metric": 0.0003274176560807973,
          "validation_metric": 0.00033824791898950934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.159686775001319
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.008145271625835449,
          "train_loss": 0.004487836733460426,
          "validation_loss": 0.0020363759249448776,
          "train_metric": 0.00045199846499599516,
          "validation_metric": 0.000205096323043108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1724506610007666
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.003325648480313248,
          "train_loss": 0.0018331578467041254,
          "validation_loss": 0.001261439174413681,
          "train_metric": 0.00018462896696291864,
          "validation_metric": 0.00012704751861747354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.184432536001623
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0020706877894554054,
          "train_loss": 0.0020606631878763437,
          "validation_loss": 0.001721513457596302,
          "train_metric": 0.00020754245633725077,
          "validation_metric": 0.00017338454199489206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.19632335300048
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.00208579691752675,
          "train_loss": 0.0018750185845419765,
          "validation_loss": 0.002153074601665139,
          "train_metric": 0.00018884499149862677,
          "validation_metric": 0.00021684981766156852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.208576962000734
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0021063219792267773,
          "train_loss": 0.004167519975453615,
          "validation_loss": 0.002370566828176379,
          "train_metric": 0.0004197373636998236,
          "validation_metric": 0.00023875481565482914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.220640830000775
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.003150990254653152,
          "train_loss": 0.0017905916320160031,
          "validation_loss": 0.001522790640592575,
          "train_metric": 0.00018034185632131994,
          "validation_metric": 0.0001533698959974572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2327341780001007
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.002409156513749622,
          "train_loss": 0.003990856930613518,
          "validation_loss": 0.0028454759158194065,
          "train_metric": 0.0004019444459117949,
          "validation_metric": 0.0002865858841687441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.244735625999965
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0040443452598992735,
          "train_loss": 0.0018562175100669265,
          "validation_loss": 0.00200914079323411,
          "train_metric": 0.000186951452633366,
          "validation_metric": 0.00020235325791873038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2569024140011607
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.002495573200576473,
          "train_loss": 0.002307482762262225,
          "validation_loss": 0.0032223775051534176,
          "train_metric": 0.00023240121663548052,
          "validation_metric": 0.00032454606844112277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2704966540004534
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0016662903526594164,
          "train_loss": 0.001617777394130826,
          "validation_loss": 0.0013279950944706798,
          "train_metric": 0.00016293660155497491,
          "validation_metric": 0.00013375081471167505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.283066691001295
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0024897904841054697,
          "train_loss": 0.005747993942350149,
          "validation_loss": 0.00596633180975914,
          "train_metric": 0.0005789169226773083,
          "validation_metric": 0.0006009070784784853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.295454539000275
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.005138890559464926,
          "train_loss": 0.009778994135558605,
          "validation_loss": 0.007887748070061207,
          "train_metric": 0.0009849044727161527,
          "validation_metric": 0.000794425024650991,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3076206740006455
        }
      ],
      "best": {
        "epoch": 130,
        "updates": 2080
      },
      "best_validation_metric": 0.00012380778207443655,
      "config": {
        "label": "ReLU",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.011126894538658868
    },
    "d7d20cfbbc18_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.8044150937348604,
          "train_loss": 0.2429167479276657,
          "validation_loss": 0.19579942524433136,
          "train_metric": 0.024465685710310936,
          "validation_metric": 0.019720202311873436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.022065019000365282
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.1870921808294952,
          "train_loss": 0.12636354565620422,
          "validation_loss": 0.1034063845872879,
          "train_metric": 0.01272687315940857,
          "validation_metric": 0.010414712131023407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.040496854000593885
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.11879568663425744,
          "train_loss": 0.09372429549694061,
          "validation_loss": 0.07738872617483139,
          "train_metric": 0.009439567103981972,
          "validation_metric": 0.007794309873133898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05856158700044034
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.0851041772402823,
          "train_loss": 0.07420113682746887,
          "validation_loss": 0.0606280192732811,
          "train_metric": 0.0074732666835188866,
          "validation_metric": 0.006106232292950153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07689357900017058
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.06589304620865732,
          "train_loss": 0.06057261675596237,
          "validation_loss": 0.04898910969495773,
          "train_metric": 0.00610065134242177,
          "validation_metric": 0.004934003576636314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10969533900060924
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.056766400462947786,
          "train_loss": 0.05399569496512413,
          "validation_loss": 0.04458855465054512,
          "train_metric": 0.005438248626887798,
          "validation_metric": 0.004490796010941267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16045713700077613
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.05240736552514136,
          "train_loss": 0.04771651327610016,
          "validation_loss": 0.03921524062752724,
          "train_metric": 0.004805833101272583,
          "validation_metric": 0.003949615638703108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2030299180005386
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.04689522925764322,
          "train_loss": 0.044059935957193375,
          "validation_loss": 0.03677556291222572,
          "train_metric": 0.004437556024640799,
          "validation_metric": 0.0037038999143987894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2541262230006396
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.0440852134488523,
          "train_loss": 0.04834325611591339,
          "validation_loss": 0.04004870727658272,
          "train_metric": 0.004868955351412296,
          "validation_metric": 0.004033559001982212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2972714770003222
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.0471903458237648,
          "train_loss": 0.041890617460012436,
          "validation_loss": 0.03535372018814087,
          "train_metric": 0.004219069145619869,
          "validation_metric": 0.003560697427019477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3454432180005824
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04023603833047673,
          "train_loss": 0.03696395829319954,
          "validation_loss": 0.03132256120443344,
          "train_metric": 0.003722874214872718,
          "validation_metric": 0.003154693404212594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39346548299909045
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.03660678002052009,
          "train_loss": 0.033676471561193466,
          "validation_loss": 0.02870030514895916,
          "train_metric": 0.0033917708788067102,
          "validation_metric": 0.0028905898798257113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44141082399983134
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.034812972764484584,
          "train_loss": 0.037384506314992905,
          "validation_loss": 0.0302804596722126,
          "train_metric": 0.00376523076556623,
          "validation_metric": 0.0030497368425130844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48158596400026
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03341828694101423,
          "train_loss": 0.03033176064491272,
          "validation_loss": 0.025632550939917564,
          "train_metric": 0.003054903820157051,
          "validation_metric": 0.002581616397947073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5977571519997582
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.032035356445703655,
          "train_loss": 0.029622448608279228,
          "validation_loss": 0.02499735727906227,
          "train_metric": 0.002983464626595378,
          "validation_metric": 0.002517642453312874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6161685850001959
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.028347863408271223,
          "train_loss": 0.029569003731012344,
          "validation_loss": 0.02469450980424881,
          "train_metric": 0.002978081814944744,
          "validation_metric": 0.0024871404748409986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6343634059994656
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.026479578984435648,
          "train_loss": 0.026618938893079758,
          "validation_loss": 0.022611606866121292,
          "train_metric": 0.0026809622067958117,
          "validation_metric": 0.0022773579694330692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6522715289993357
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.025765705038793385,
          "train_loss": 0.02319283038377762,
          "validation_loss": 0.01962474174797535,
          "train_metric": 0.002335896948352456,
          "validation_metric": 0.001976531930267811,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6704606740004238
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.023702766571659595,
          "train_loss": 0.02468171902000904,
          "validation_loss": 0.020523663610219955,
          "train_metric": 0.002485852222889662,
          "validation_metric": 0.002067068126052618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6873878049991617
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.027035322273150086,
          "train_loss": 0.023569690063595772,
          "validation_loss": 0.019256599247455597,
          "train_metric": 0.00237385299988091,
          "validation_metric": 0.0019394539995118976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7051626369993755
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.024231369083281606,
          "train_loss": 0.02096431888639927,
          "validation_loss": 0.018189555034041405,
          "train_metric": 0.0021114496048539877,
          "validation_metric": 0.0018319853115826845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7228604920001089
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.028583730629179627,
          "train_loss": 0.02846529707312584,
          "validation_loss": 0.0228520929813385,
          "train_metric": 0.002866920083761215,
          "validation_metric": 0.0023015791084617376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7580411510007252
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.02236662874929607,
          "train_loss": 0.01823396049439907,
          "validation_loss": 0.015749260783195496,
          "train_metric": 0.0018364576390013099,
          "validation_metric": 0.0015862080035731196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8086641859990777
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019407302752370015,
          "train_loss": 0.017195433378219604,
          "validation_loss": 0.014424235559999943,
          "train_metric": 0.0017318609170615673,
          "validation_metric": 0.0014527563471347094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8592245540003205
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.02198730458621867,
          "train_loss": 0.016548244282603264,
          "validation_loss": 0.013975901529192924,
          "train_metric": 0.0016666786978021264,
          "validation_metric": 0.0014076016377657652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8772115859992482
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.019766455719945952,
          "train_loss": 0.016055293381214142,
          "validation_loss": 0.01397432666271925,
          "train_metric": 0.0016170304734259844,
          "validation_metric": 0.0014074431965127587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8958889940004156
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.018066079472191632,
          "train_loss": 0.015439325012266636,
          "validation_loss": 0.013047521002590656,
          "train_metric": 0.0015549923991784453,
          "validation_metric": 0.001314098364673555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.913688283000738
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.015645972365746275,
          "train_loss": 0.014692176133394241,
          "validation_loss": 0.012932147830724716,
          "train_metric": 0.001479742000810802,
          "validation_metric": 0.001302478602156043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9326705219991709
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.018018380331341177,
          "train_loss": 0.02725718356668949,
          "validation_loss": 0.02171926572918892,
          "train_metric": 0.0027452437207102776,
          "validation_metric": 0.0021874848753213882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9503656979995867
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01754122640704736,
          "train_loss": 0.016439730301499367,
          "validation_loss": 0.01306527853012085,
          "train_metric": 0.0016557492781430483,
          "validation_metric": 0.0013158870860934258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9676730989995121
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.017435521876905113,
          "train_loss": 0.02235007844865322,
          "validation_loss": 0.01915544830262661,
          "train_metric": 0.002251017838716507,
          "validation_metric": 0.0019292667275294662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9883589469991421
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.02170187936280854,
          "train_loss": 0.018943287432193756,
          "validation_loss": 0.015097816474735737,
          "train_metric": 0.0019078985787928104,
          "validation_metric": 0.0015205965610221028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.02773494900066
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.01810578120057471,
          "train_loss": 0.013324378058314323,
          "validation_loss": 0.011162937618792057,
          "train_metric": 0.0013419826282188296,
          "validation_metric": 0.0011242901673540473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0500501389997225
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01584213098976761,
          "train_loss": 0.027000177651643753,
          "validation_loss": 0.022715752944350243,
          "train_metric": 0.0027193594723939896,
          "validation_metric": 0.0022878474555909634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0670815899993613
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.018394911021459848,
          "train_loss": 0.013409027829766273,
          "validation_loss": 0.012042668648064137,
          "train_metric": 0.001350508420728147,
          "validation_metric": 0.001212893519550562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0840908890004357
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.014787977343075909,
          "train_loss": 0.013721950352191925,
          "validation_loss": 0.011373530142009258,
          "train_metric": 0.0013820247258991003,
          "validation_metric": 0.0011455003404989839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.101512601999275
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.015369479107903317,
          "train_loss": 0.014852654188871384,
          "validation_loss": 0.011957141570746899,
          "train_metric": 0.0014959049876779318,
          "validation_metric": 0.0012042793678119779,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.118843915999605
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.012873228028183803,
          "train_loss": 0.012814885005354881,
          "validation_loss": 0.011936694383621216,
          "train_metric": 0.0012906682677567005,
          "validation_metric": 0.001202220213599503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1369145450007636
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.013424148259218782,
          "train_loss": 0.01583930104970932,
          "validation_loss": 0.011695927008986473,
          "train_metric": 0.0015952764078974724,
          "validation_metric": 0.0011779710184782743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.154380215000856
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.013970585816423409,
          "train_loss": 0.013790346682071686,
          "validation_loss": 0.012640640139579773,
          "train_metric": 0.0013889133697375655,
          "validation_metric": 0.0012731190072372556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1721917790000589
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.013271831121528521,
          "train_loss": 0.01119884941726923,
          "validation_loss": 0.009249459020793438,
          "train_metric": 0.0011279070749878883,
          "validation_metric": 0.0009315717034041882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.189636270999472
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.01159201210248284,
          "train_loss": 0.01050405390560627,
          "validation_loss": 0.009355021640658379,
          "train_metric": 0.0010579299414530396,
          "validation_metric": 0.0009422034490853548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2071994439993432
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.011418641050113365,
          "train_loss": 0.01422963198274374,
          "validation_loss": 0.010923593305051327,
          "train_metric": 0.001433156430721283,
          "validation_metric": 0.0011001843959093094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2245868509999127
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.014601574803236872,
          "train_loss": 0.012003142386674881,
          "validation_loss": 0.009487897157669067,
          "train_metric": 0.0012089126976206899,
          "validation_metric": 0.0009555862634442747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2419345199996314
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013204803093685769,
          "train_loss": 0.012686978094279766,
          "validation_loss": 0.01206391304731369,
          "train_metric": 0.0012777859810739756,
          "validation_metric": 0.001215033233165741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2582880600002682
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.012819723138818517,
          "train_loss": 0.012320191599428654,
          "validation_loss": 0.011551903560757637,
          "train_metric": 0.0012408447219058871,
          "validation_metric": 0.0011634655529633164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2746714330005489
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.012924227950861678,
          "train_loss": 0.016175657510757446,
          "validation_loss": 0.012536747381091118,
          "train_metric": 0.001629152917303145,
          "validation_metric": 0.001262655365280807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.291472987999441
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.012322669994318858,
          "train_loss": 0.009928863495588303,
          "validation_loss": 0.008589740842580795,
          "train_metric": 0.000999998883344233,
          "validation_metric": 0.0008651271928101778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3089977700001327
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.010198263975325972,
          "train_loss": 0.009641319513320923,
          "validation_loss": 0.00889771431684494,
          "train_metric": 0.0009710384765639901,
          "validation_metric": 0.00089614512398839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3260232800002996
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.010246199526591226,
          "train_loss": 0.009774981066584587,
          "validation_loss": 0.008472250774502754,
          "train_metric": 0.0009845003951340914,
          "validation_metric": 0.0008532939828000963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3445956649993605
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.011311871960060671,
          "train_loss": 0.010349616408348083,
          "validation_loss": 0.00842268206179142,
          "train_metric": 0.0010423754574730992,
          "validation_metric": 0.0008483016863465309,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.363193911000053
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.01465673885832075,
          "train_loss": 0.01081780344247818,
          "validation_loss": 0.009592870250344276,
          "train_metric": 0.0010895294835790992,
          "validation_metric": 0.0009661587537266314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3826322320001054
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.010308697412256151,
          "train_loss": 0.009508473798632622,
          "validation_loss": 0.008145428262650967,
          "train_metric": 0.0009576584561727941,
          "validation_metric": 0.0008203777251765132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4006024520003848
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009352348715765402,
          "train_loss": 0.009950908832252026,
          "validation_loss": 0.008984668180346489,
          "train_metric": 0.0010022191563621163,
          "validation_metric": 0.0009049028158187866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4180173719996674
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.010348400566726923,
          "train_loss": 0.007953369989991188,
          "validation_loss": 0.006937202997505665,
          "train_metric": 0.0008010342717170715,
          "validation_metric": 0.0006986897205933928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4439408620000904
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.012355427505099215,
          "train_loss": 0.008674860931932926,
          "validation_loss": 0.006868367549031973,
          "train_metric": 0.0008737003663554788,
          "validation_metric": 0.0006917567225173116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.475086390000797
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.00928299137740396,
          "train_loss": 0.00779365561902523,
          "validation_loss": 0.007054467685520649,
          "train_metric": 0.0007849485264159739,
          "validation_metric": 0.0007105001714080572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4966784019998158
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.00963281060103327,
          "train_loss": 0.007974233478307724,
          "validation_loss": 0.007237949408590794,
          "train_metric": 0.0008031356846913695,
          "validation_metric": 0.0007289796485565603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.526242509999065
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.008164931408828124,
          "train_loss": 0.007786103058606386,
          "validation_loss": 0.006780148018151522,
          "train_metric": 0.0007841877522878349,
          "validation_metric": 0.0006828716141171753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5491135970005416
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007935682544484735,
          "train_loss": 0.0073556117713451385,
          "validation_loss": 0.006855516228824854,
          "train_metric": 0.0007408303790725768,
          "validation_metric": 0.0006904624169692397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5687750130000495
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.007902754878159612,
          "train_loss": 0.00722744083032012,
          "validation_loss": 0.005976418033242226,
          "train_metric": 0.0007279214332811534,
          "validation_metric": 0.0006019229185767472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5852818099992874
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.007543633662862703,
          "train_loss": 0.006611998192965984,
          "validation_loss": 0.005726413335651159,
          "train_metric": 0.0006659362697973847,
          "validation_metric": 0.0005767433904111385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5993488339991018
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.009666586731327698,
          "train_loss": 0.014753266237676144,
          "validation_loss": 0.009954906068742275,
          "train_metric": 0.0014858947833999991,
          "validation_metric": 0.001002621604129672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.61297057199954
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.011440971153206192,
          "train_loss": 0.008051286451518536,
          "validation_loss": 0.007899723947048187,
          "train_metric": 0.0008108961046673357,
          "validation_metric": 0.0007956312038004398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6263972060005472
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.007984957497683354,
          "train_loss": 0.006501683499664068,
          "validation_loss": 0.0055784243158996105,
          "train_metric": 0.0006548257661052048,
          "validation_metric": 0.0005618385039269924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6398282430000108
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.006897279352415353,
          "train_loss": 0.0067572398111224174,
          "validation_loss": 0.005641228519380093,
          "train_metric": 0.0006805644952692091,
          "validation_metric": 0.000568163872230798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.652943408000283
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0083939760515932,
          "train_loss": 0.009118188172578812,
          "validation_loss": 0.008781718090176582,
          "train_metric": 0.000918350531719625,
          "validation_metric": 0.0008844624971970916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.665305876000275
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.008580772526329383,
          "train_loss": 0.00807133037596941,
          "validation_loss": 0.006589137017726898,
          "train_metric": 0.0008129148627631366,
          "validation_metric": 0.0006636337493546307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6776825230008399
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.011109586237580515,
          "train_loss": 0.006373533047735691,
          "validation_loss": 0.004987717606127262,
          "train_metric": 0.0006419188575819135,
          "validation_metric": 0.000502344686537981,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6905764959992666
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.008426171145401895,
          "train_loss": 0.00838069710880518,
          "validation_loss": 0.00793797243386507,
          "train_metric": 0.0008440731908194721,
          "validation_metric": 0.0007994835614226758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7029145159995096
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.007210739488073159,
          "train_loss": 0.006572912912815809,
          "validation_loss": 0.0050555323250591755,
          "train_metric": 0.0006619998603127897,
          "validation_metric": 0.0005091746570542455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7152388590002374
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0065463825449114665,
          "train_loss": 0.006071347743272781,
          "validation_loss": 0.004915774799883366,
          "train_metric": 0.0006114839925430715,
          "validation_metric": 0.0004950988222844899,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.728535632000785
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.006305865521426313,
          "train_loss": 0.006719997618347406,
          "validation_loss": 0.0060499063692986965,
          "train_metric": 0.0006768135935999453,
          "validation_metric": 0.00060932437190786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.741033198999503
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.006790031067794189,
          "train_loss": 0.006161171477288008,
          "validation_loss": 0.00486963102594018,
          "train_metric": 0.0006205306854099035,
          "validation_metric": 0.0004904514062218368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7541558360007912
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.007946100202389061,
          "train_loss": 0.00534816924482584,
          "validation_loss": 0.004658317659050226,
          "train_metric": 0.0005386479897424579,
          "validation_metric": 0.00046916870633140206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7676014579992625
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.006416511096176691,
          "train_loss": 0.0061532314866781235,
          "validation_loss": 0.004654313903301954,
          "train_metric": 0.0006197310285642743,
          "validation_metric": 0.0004687655018642545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7808009479995235
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.0072897224890766665,
          "train_loss": 0.00768575444817543,
          "validation_loss": 0.007230259478092194,
          "train_metric": 0.0007740810397081077,
          "validation_metric": 0.0007282051956281066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7931809879992215
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006018827516527381,
          "train_loss": 0.005309337750077248,
          "validation_loss": 0.004323083441704512,
          "train_metric": 0.0005347371334210038,
          "validation_metric": 0.00043540517799556255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8103670029995556
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006542081144289114,
          "train_loss": 0.005113048013299704,
          "validation_loss": 0.0045873685739934444,
          "train_metric": 0.0005149674834683537,
          "validation_metric": 0.0004620229301508516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.824184619999869
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.005390375095885247,
          "train_loss": 0.0050711557269096375,
          "validation_loss": 0.0040695928037166595,
          "train_metric": 0.0005107481847517192,
          "validation_metric": 0.0004098744539078325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8509939169998688
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0074567216070136055,
          "train_loss": 0.01155670266598463,
          "validation_loss": 0.009457407519221306,
          "train_metric": 0.0011639486765488982,
          "validation_metric": 0.0009525154600851238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8743488309992244
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.010841197348781861,
          "train_loss": 0.007164538837969303,
          "validation_loss": 0.0064179301261901855,
          "train_metric": 0.0007215860532596707,
          "validation_metric": 0.000646390428300947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8963243439993676
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.008425459556747228,
          "train_loss": 0.005060063675045967,
          "validation_loss": 0.004097746219485998,
          "train_metric": 0.0005096311797387898,
          "validation_metric": 0.00041270998190157115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9266090790006274
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.005369617305404972,
          "train_loss": 0.005505145527422428,
          "validation_loss": 0.004116369411349297,
          "train_metric": 0.0005544581217691302,
          "validation_metric": 0.00041458557825535536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.026727555999969
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.007623823155881837,
          "train_loss": 0.01393161341547966,
          "validation_loss": 0.011591765098273754,
          "train_metric": 0.0014031410682946444,
          "validation_metric": 0.0011674801353365183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0389633939994383
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.01111048727761954,
          "train_loss": 0.015113571658730507,
          "validation_loss": 0.009592129848897457,
          "train_metric": 0.0015221837675198913,
          "validation_metric": 0.000966084364335984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0510993890002283
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.011885778949363157,
          "train_loss": 0.006833478342741728,
          "validation_loss": 0.00606330344453454,
          "train_metric": 0.0006882428424432874,
          "validation_metric": 0.000610673741903156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.062807934000375
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.006130210691480897,
          "train_loss": 0.005364680662751198,
          "validation_loss": 0.004389415495097637,
          "train_metric": 0.0005403109826147556,
          "validation_metric": 0.00044208590406924486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.075430168000821
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.005021722739911638,
          "train_loss": 0.00440845126286149,
          "validation_loss": 0.003584228688850999,
          "train_metric": 0.00044400308979675174,
          "validation_metric": 0.0003609903797041625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0879703589998826
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004931285810016561,
          "train_loss": 0.004349071532487869,
          "validation_loss": 0.0036433821078389883,
          "train_metric": 0.00043802251457236707,
          "validation_metric": 0.00036694807931780815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1001300780008023
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.006381518498528749,
          "train_loss": 0.004895043093711138,
          "validation_loss": 0.003925356548279524,
          "train_metric": 0.0004930108552798629,
          "validation_metric": 0.00039534750976599753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1121188599991
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0055897237471072,
          "train_loss": 0.004494658671319485,
          "validation_loss": 0.003998330794274807,
          "train_metric": 0.00045268560643307865,
          "validation_metric": 0.00040269727469421923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1244630990004225
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.005735230268328451,
          "train_loss": 0.00911650899797678,
          "validation_loss": 0.006201489828526974,
          "train_metric": 0.0009181814384646714,
          "validation_metric": 0.000624591251835227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1396823699997185
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.005074139160569757,
          "train_loss": 0.009522434324026108,
          "validation_loss": 0.007935792207717896,
          "train_metric": 0.0009590646950528026,
          "validation_metric": 0.0007992639439180493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1520376349999424
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.004155101029027719,
          "train_loss": 0.005300888791680336,
          "validation_loss": 0.004852017387747765,
          "train_metric": 0.0005338860210031271,
          "validation_metric": 0.000488677469547838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1654865100008465
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.005414259940152988,
          "train_loss": 0.003995438572019339,
          "validation_loss": 0.003463434986770153,
          "train_metric": 0.0004024059744551778,
          "validation_metric": 0.00034882448380813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1789023790006468
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.006546056698425673,
          "train_loss": 0.006884937174618244,
          "validation_loss": 0.004892976954579353,
          "train_metric": 0.0006934256525710225,
          "validation_metric": 0.0004928027046844363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1913183220003702
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.006845640134997666,
          "train_loss": 0.004689313471317291,
          "validation_loss": 0.004034270066767931,
          "train_metric": 0.00047229044139385223,
          "validation_metric": 0.00040631694719195366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.20410179099963
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.005325268677552231,
          "train_loss": 0.005320175085216761,
          "validation_loss": 0.004474917892366648,
          "train_metric": 0.0005358285270631313,
          "validation_metric": 0.0004506973491515964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2166534429998137
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.006616686179768294,
          "train_loss": 0.005308895371854305,
          "validation_loss": 0.004988147411495447,
          "train_metric": 0.0005346925463527441,
          "validation_metric": 0.0005023878766223788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2284166749996075
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.006149091088445857,
          "train_loss": 0.004543749149888754,
          "validation_loss": 0.0037650507874786854,
          "train_metric": 0.000457629794254899,
          "validation_metric": 0.00037920213071629405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2407281510004395
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.007483936242351774,
          "train_loss": 0.00539410300552845,
          "validation_loss": 0.004432002082467079,
          "train_metric": 0.0005432742764241993,
          "validation_metric": 0.00044637496466748416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2545456389998435
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.009308796637924388,
          "train_loss": 0.006445842329412699,
          "validation_loss": 0.005544780753552914,
          "train_metric": 0.0006492015672847629,
          "validation_metric": 0.0005584500031545758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.266616225000689
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.005561418525758199,
          "train_loss": 0.004003529436886311,
          "validation_loss": 0.0029813116416335106,
          "train_metric": 0.0004032208235003054,
          "validation_metric": 0.00030026675085537136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.282869254999241
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.005496915015100967,
          "train_loss": 0.005087155383080244,
          "validation_loss": 0.0037439640145748854,
          "train_metric": 0.0005123596056364477,
          "validation_metric": 0.0003770783368963748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.298629823999363
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.005087374702270608,
          "train_loss": 0.004476629197597504,
          "validation_loss": 0.00332259782589972,
          "train_metric": 0.0004508697020355612,
          "validation_metric": 0.0003346399462316185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.315563198999371
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.00430886081448989,
          "train_loss": 0.010418150573968887,
          "validation_loss": 0.008648399263620377,
          "train_metric": 0.0010492779547348619,
          "validation_metric": 0.0008710350375622511,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.332637629000601
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.005104605974338483,
          "train_loss": 0.0061800843104720116,
          "validation_loss": 0.00405865116044879,
          "train_metric": 0.0006224354729056358,
          "validation_metric": 0.0004087724955752492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.350434738000331
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.00556728859373834,
          "train_loss": 0.0032953694462776184,
          "validation_loss": 0.0028615829069167376,
          "train_metric": 0.00033189752139151096,
          "validation_metric": 0.00028820818988606334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.364345616999344
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.005813666695758002,
          "train_loss": 0.004429598338901997,
          "validation_loss": 0.0033305350225418806,
          "train_metric": 0.0004461329081095755,
          "validation_metric": 0.00033543925383128226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3765667410007154
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.005742874549468979,
          "train_loss": 0.0053133233450353146,
          "validation_loss": 0.005243764724582434,
          "train_metric": 0.0005351384170353413,
          "validation_metric": 0.00052813277579844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.390204748000542
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.0051096877905365545,
          "train_loss": 0.0041269478388130665,
          "validation_loss": 0.003674024250358343,
          "train_metric": 0.00041565109859220684,
          "validation_metric": 0.00037003427860327065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.402200410999285
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.004020453670818824,
          "train_loss": 0.0035053500905632973,
          "validation_loss": 0.0032802806235849857,
          "train_metric": 0.0003530460235197097,
          "validation_metric": 0.00033037783578038216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4140388849991723
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0034824878748622723,
          "train_loss": 0.004781292285770178,
          "validation_loss": 0.00325078796595335,
          "train_metric": 0.00048155427793972194,
          "validation_metric": 0.00032740741153247654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4260291259997757
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0049214100145036355,
          "train_loss": 0.005687706172466278,
          "validation_loss": 0.004938316065818071,
          "train_metric": 0.00057284499052912,
          "validation_metric": 0.0004973690956830978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.43733499299924
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0035912905950681306,
          "train_loss": 0.00435817614197731,
          "validation_loss": 0.0033550672233104706,
          "train_metric": 0.0004389395471662283,
          "validation_metric": 0.0003379100817255676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4489006889998564
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.004135905794100836,
          "train_loss": 0.0045005339197814465,
          "validation_loss": 0.0028847043868154287,
          "train_metric": 0.0004532773164100945,
          "validation_metric": 0.00029053687467239797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4606962020006904
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003810544076259248,
          "train_loss": 0.003223689738661051,
          "validation_loss": 0.0028640874661505222,
          "train_metric": 0.0003246782289352268,
          "validation_metric": 0.0002884604036808014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.472376876999988
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0034818675921997055,
          "train_loss": 0.00292445276863873,
          "validation_loss": 0.002340627135708928,
          "train_metric": 0.00029454019386321306,
          "validation_metric": 0.0002357393823331222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4846819340000366
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.003432213983614929,
          "train_loss": 0.0030202437192201614,
          "validation_loss": 0.0024558864533901215,
          "train_metric": 0.0003041879099328071,
          "validation_metric": 0.00024734792532399297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.496561052999823
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.005426803865702823,
          "train_loss": 0.003961409442126751,
          "validation_loss": 0.0029158785473555326,
          "train_metric": 0.00039897862006910145,
          "validation_metric": 0.00029367656679823995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5087691829994583
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.004040402942337096,
          "train_loss": 0.0033713877201080322,
          "validation_loss": 0.002954941475763917,
          "train_metric": 0.000339553807862103,
          "validation_metric": 0.00029761085170321167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5210463130006247
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.003376275926711969,
          "train_loss": 0.0031763529404997826,
          "validation_loss": 0.002466425998136401,
          "train_metric": 0.00031991058494895697,
          "validation_metric": 0.00024840940022841096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.533600536000449
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.003329936975205783,
          "train_loss": 0.0034438087604939938,
          "validation_loss": 0.0023803075309842825,
          "train_metric": 0.0003468478098511696,
          "validation_metric": 0.00023973584757186472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.548615281999446
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0034420964439050294,
          "train_loss": 0.0031074550934135914,
          "validation_loss": 0.002880032639950514,
          "train_metric": 0.00031297150417231023,
          "validation_metric": 0.00029006635304540396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.561604905000422
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.003631671075709164,
          "train_loss": 0.003757245372980833,
          "validation_loss": 0.0027421261183917522,
          "train_metric": 0.00037841597804799676,
          "validation_metric": 0.00027617692830972373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.573604642000646
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0033073878003051504,
          "train_loss": 0.0028035673312842846,
          "validation_loss": 0.0022357089910656214,
          "train_metric": 0.0002823650138452649,
          "validation_metric": 0.000225172407226637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5861736639999435
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.004888009898422752,
          "train_loss": 0.003749913303181529,
          "validation_loss": 0.0027790367603302,
          "train_metric": 0.00037767758476547897,
          "validation_metric": 0.00027989441878162324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5983721339998738
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.009450155532249482,
          "train_loss": 0.007103859446942806,
          "validation_loss": 0.0060860710218548775,
          "train_metric": 0.000715474656317383,
          "validation_metric": 0.0006129668327048421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6108578570001555
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.00971792425843887,
          "train_loss": 0.005452544428408146,
          "validation_loss": 0.0036226161755621433,
          "train_metric": 0.0005491602933034301,
          "validation_metric": 0.00036485667806118727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.622541550999813
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0059479376941453665,
          "train_loss": 0.0077172331511974335,
          "validation_loss": 0.006815851200371981,
          "train_metric": 0.0007772514363750815,
          "validation_metric": 0.0006864675087854266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.63467417100037
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.008447614236501977,
          "train_loss": 0.0151943638920784,
          "validation_loss": 0.0098756467923522,
          "train_metric": 0.001530320500023663,
          "validation_metric": 0.0009946390055119991,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.646649570999216
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.008603556256275624,
          "train_loss": 0.004318565130233765,
          "validation_loss": 0.004299588035792112,
          "train_metric": 0.00043495005229488015,
          "validation_metric": 0.00043303874554112554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.658466934999524
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.003695645667903591,
          "train_loss": 0.002824033610522747,
          "validation_loss": 0.0023344741202890873,
          "train_metric": 0.000284426350845024,
          "validation_metric": 0.00023511967447120696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6702997259999393
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0030381248107005376,
          "train_loss": 0.00551297701895237,
          "validation_loss": 0.0051863850094377995,
          "train_metric": 0.0005552469519898295,
          "validation_metric": 0.0005223536863923073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6819548349994875
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.003563573358405847,
          "train_loss": 0.0028364018071442842,
          "validation_loss": 0.002195230685174465,
          "train_metric": 0.00028567196568474174,
          "validation_metric": 0.0002210956154158339,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.694535447000817
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.003492096751870122,
          "train_loss": 0.003653447376564145,
          "validation_loss": 0.0025937564205378294,
          "train_metric": 0.00036796185304410756,
          "validation_metric": 0.0002612336538732052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7067412409996905
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.004275370192772243,
          "train_loss": 0.0028869558591395617,
          "validation_loss": 0.002550545148551464,
          "train_metric": 0.0002907636226154864,
          "validation_metric": 0.0002568815543781966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.718664992000413
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.003292162189609371,
          "train_loss": 0.005038602277636528,
          "validation_loss": 0.003641009796410799,
          "train_metric": 0.0005074695800431073,
          "validation_metric": 0.0003667091950774193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7308577790008712
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.005875593767996179,
          "train_loss": 0.010535136796534061,
          "validation_loss": 0.00946925301104784,
          "train_metric": 0.0010610604658722878,
          "validation_metric": 0.0009537084843032062,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.742882381000527
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.010587691256660037,
          "train_loss": 0.008004630915820599,
          "validation_loss": 0.004766499157994986,
          "train_metric": 0.0008061970002017915,
          "validation_metric": 0.00048006430733948946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.757477437999114
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.014069110475247726,
          "train_loss": 0.005116206593811512,
          "validation_loss": 0.004384080413728952,
          "train_metric": 0.0005152855301275849,
          "validation_metric": 0.0004415484727360308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.769244150000304
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.005136335705174133,
          "train_loss": 0.0037536765448749065,
          "validation_loss": 0.0030796751379966736,
          "train_metric": 0.0003780565457418561,
          "validation_metric": 0.00031017360743135214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.781588544999977
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0034505486582929734,
          "train_loss": 0.003820228623226285,
          "validation_loss": 0.0034956736490130424,
          "train_metric": 0.00038475944893434644,
          "validation_metric": 0.00035207142354920506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7934616549991915
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.004803555391845293,
          "train_loss": 0.00815190002322197,
          "validation_loss": 0.0057352506555616856,
          "train_metric": 0.0008210294763557613,
          "validation_metric": 0.0005776335019618273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.805198202999236
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0037221284765109885,
          "train_loss": 0.0025462103076279163,
          "validation_loss": 0.002132151974365115,
          "train_metric": 0.00025644496781751513,
          "validation_metric": 0.00021474255481734872,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8178766229993926
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0033347385251545347,
          "train_loss": 0.0026467570569366217,
          "validation_loss": 0.0022005285136401653,
          "train_metric": 0.0002665716747287661,
          "validation_metric": 0.00022162919049151242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8298916690000624
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0030366467690328136,
          "train_loss": 0.002766222693026066,
          "validation_loss": 0.002563424175605178,
          "train_metric": 0.00027860378031618893,
          "validation_metric": 0.0002581786538939923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.844622859000083
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.002808665914926678,
          "train_loss": 0.0025274986401200294,
          "validation_loss": 0.0022198250517249107,
          "train_metric": 0.0002545603783801198,
          "validation_metric": 0.00022357267152983695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8563777300005313
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0024492241427651607,
          "train_loss": 0.0031271164771169424,
          "validation_loss": 0.002282728673890233,
          "train_metric": 0.0003149517287965864,
          "validation_metric": 0.0002299080661032349,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.877929360000053
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.002676710584637476,
          "train_loss": 0.0025372339878231287,
          "validation_loss": 0.0018870929488912225,
          "train_metric": 0.0002555409155320376,
          "validation_metric": 0.00019006106595043093,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8967657560006046
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0027484017773531377,
          "train_loss": 0.0027663311921060085,
          "validation_loss": 0.0019293756922706962,
          "train_metric": 0.0002786147524602711,
          "validation_metric": 0.0001943196402862668,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9142321399995126
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.005018782874685712,
          "train_loss": 0.002890325151383877,
          "validation_loss": 0.002406979212537408,
          "train_metric": 0.00029110300238244236,
          "validation_metric": 0.00024242213112302125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9309330950000003
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.002745165540545713,
          "train_loss": 0.0025180880911648273,
          "validation_loss": 0.0018320484086871147,
          "train_metric": 0.0002536125830374658,
          "validation_metric": 0.00018451722280588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.949061974999495
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0027294727624394,
          "train_loss": 0.0031284773722290993,
          "validation_loss": 0.0027955682016909122,
          "train_metric": 0.0003150887496303767,
          "validation_metric": 0.00028155939071439207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9661065809996217
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0035642268994706683,
          "train_loss": 0.0029080347158014774,
          "validation_loss": 0.0023589287884533405,
          "train_metric": 0.0002928866306319833,
          "validation_metric": 0.00023758268798701465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9830184719994577
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0032276790880132467,
          "train_loss": 0.0025511470157653093,
          "validation_loss": 0.00238602701574564,
          "train_metric": 0.000256942177657038,
          "validation_metric": 0.00024031191424001008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0006619830001
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.002501054070307873,
          "train_loss": 0.002175412606447935,
          "validation_loss": 0.001731468248181045,
          "train_metric": 0.00021909958741161972,
          "validation_metric": 0.00017438713985029608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.018973050000568
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.00261098282499006,
          "train_loss": 0.0025316718965768814,
          "validation_loss": 0.001854279194958508,
          "train_metric": 0.00025498075410723686,
          "validation_metric": 0.00018675622413866222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.036269718000767
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.002599723873572657,
          "train_loss": 0.0032288029324263334,
          "validation_loss": 0.0030935814138501883,
          "train_metric": 0.00032519325031898916,
          "validation_metric": 0.000311574200168252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0548936430004687
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.002916885599915986,
          "train_loss": 0.0027302587404847145,
          "validation_loss": 0.0017612174851819873,
          "train_metric": 0.0002749816922005266,
          "validation_metric": 0.0001773833646439016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.071847130999231
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.003347485340782441,
          "train_loss": 0.0024598382879048586,
          "validation_loss": 0.0020294091664254665,
          "train_metric": 0.00024774589110165834,
          "validation_metric": 0.00020439464424271137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0888427959998808
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0032778353488538414,
          "train_loss": 0.0019986198749393225,
          "validation_loss": 0.0016924797091633081,
          "train_metric": 0.00020129370386712253,
          "validation_metric": 0.0001704603637335822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1067214510003396
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.002771033614408225,
          "train_loss": 0.002330189570784569,
          "validation_loss": 0.0016826222417876124,
          "train_metric": 0.0002346881665289402,
          "validation_metric": 0.0001694675302132964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1284121089993278
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0027795352252724115,
          "train_loss": 0.008106683380901814,
          "validation_loss": 0.004623236134648323,
          "train_metric": 0.0008164755417965353,
          "validation_metric": 0.0004656354431062937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1466225500007567
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0036036399760632776,
          "train_loss": 0.004375492222607136,
          "validation_loss": 0.003155309706926346,
          "train_metric": 0.0004406835651025176,
          "validation_metric": 0.0003177912440150976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.164240862999577
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.003415670529648196,
          "train_loss": 0.0025362123269587755,
          "validation_loss": 0.0022041385527700186,
          "train_metric": 0.00025543803349137306,
          "validation_metric": 0.00022199272643774748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.181526983000367
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0029446225889842026,
          "train_loss": 0.004302089102566242,
          "validation_loss": 0.002998379524797201,
          "train_metric": 0.0004332906100898981,
          "validation_metric": 0.0003019857977051288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.199354369999128
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0036602668769774027,
          "train_loss": 0.0028483488131314516,
          "validation_loss": 0.0029154664371162653,
          "train_metric": 0.00028687529265880585,
          "validation_metric": 0.0002936350938398391,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.224751149999065
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0032369010586990044,
          "train_loss": 0.002119882730767131,
          "validation_loss": 0.0020749515388160944,
          "train_metric": 0.00021350682072807103,
          "validation_metric": 0.00020898152433801442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2645021169992106
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0027358451043255627,
          "train_loss": 0.0027251210995018482,
          "validation_loss": 0.0018270218279212713,
          "train_metric": 0.00027446422609500587,
          "validation_metric": 0.00018401093257125467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2940062830002717
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0028254585558897816,
          "train_loss": 0.0035051119048148394,
          "validation_loss": 0.0027384525164961815,
          "train_metric": 0.0003530220128595829,
          "validation_metric": 0.0002758069313131273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.325472381000509
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0025928760951501317,
          "train_loss": 0.0018982780165970325,
          "validation_loss": 0.0014237778959795833,
          "train_metric": 0.00019118760246783495,
          "validation_metric": 0.00014339770132210106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3556977739990543
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.002282688512423192,
          "train_loss": 0.004309812560677528,
          "validation_loss": 0.0027881504502147436,
          "train_metric": 0.0004340686136856675,
          "validation_metric": 0.00028081226628273726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3776550229995337
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0030784733971813694,
          "train_loss": 0.0021796857472509146,
          "validation_loss": 0.0015737006906419992,
          "train_metric": 0.0002195299748564139,
          "validation_metric": 0.00015849732153583318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3989129639994644
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0018557227413111832,
          "train_loss": 0.0022411663085222244,
          "validation_loss": 0.0017254184931516647,
          "train_metric": 0.00022572206216864288,
          "validation_metric": 0.0001737778220558539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.419508458000564
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0030362235447682906,
          "train_loss": 0.002662638435140252,
          "validation_loss": 0.0021319547668099403,
          "train_metric": 0.0002681711339391768,
          "validation_metric": 0.00021472270600497723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4434900140004174
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0033315425826003775,
          "train_loss": 0.0028539381455630064,
          "validation_loss": 0.001949957339093089,
          "train_metric": 0.0002874382189475,
          "validation_metric": 0.00019639256061054766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4644889079991117
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.004300391607102938,
          "train_loss": 0.002659587422385812,
          "validation_loss": 0.0019698257092386484,
          "train_metric": 0.000267863943008706,
          "validation_metric": 0.0001983936526812613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.486213948999648
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.003752019452804234,
          "train_loss": 0.002355888020247221,
          "validation_loss": 0.0017798489425331354,
          "train_metric": 0.00023727642837911844,
          "validation_metric": 0.0001792598923202604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5062952780008345
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0024064749050012324,
          "train_loss": 0.001987146446481347,
          "validation_loss": 0.0014740756014361978,
          "train_metric": 0.00020013810717500746,
          "validation_metric": 0.0001484634994994849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.526641018999726
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0027345131729816785,
          "train_loss": 0.0039122095331549644,
          "validation_loss": 0.002709099790081382,
          "train_metric": 0.00039402334368787706,
          "validation_metric": 0.0002728505933191627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.547058431999176
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0029758994860458188,
          "train_loss": 0.0030743100214749575,
          "validation_loss": 0.002259613946080208,
          "train_metric": 0.00030963323661126196,
          "validation_metric": 0.00022758005070500076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.567137809999622
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.002925732835137751,
          "train_loss": 0.0016903537325561047,
          "validation_loss": 0.0013502520741894841,
          "train_metric": 0.000170246246852912,
          "validation_metric": 0.0001359924499411136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5883068589992035
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0023343241737165954,
          "train_loss": 0.0018844432197511196,
          "validation_loss": 0.0013953085290268064,
          "train_metric": 0.00018979422748088837,
          "validation_metric": 0.00014053034828975797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6093989919991145
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.002020268642809242,
          "train_loss": 0.004372765775769949,
          "validation_loss": 0.0036174289416521788,
          "train_metric": 0.0004404089704621583,
          "validation_metric": 0.00036433417699299753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6301356719995965
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.00214692694862606,
          "train_loss": 0.0029882811941206455,
          "validation_loss": 0.0019067004323005676,
          "train_metric": 0.0003009687352459878,
          "validation_metric": 0.00019203587726224214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.651099945000169
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0025626441747590434,
          "train_loss": 0.001501294318586588,
          "validation_loss": 0.0011990993516519666,
          "train_metric": 0.00015120487660169601,
          "validation_metric": 0.00012076888378942385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.772736630999134
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.002209004498581635,
          "train_loss": 0.004260517191141844,
          "validation_loss": 0.0025506955571472645,
          "train_metric": 0.00042910370393656194,
          "validation_metric": 0.0002568966883700341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7921911619996536
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.005988674571199226,
          "train_loss": 0.0060157557018101215,
          "validation_loss": 0.00571016501635313,
          "train_metric": 0.0006058848230168223,
          "validation_metric": 0.0005751068238168955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8107951780002622
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.005274490249576047,
          "train_loss": 0.002284142654389143,
          "validation_loss": 0.0020611470099538565,
          "train_metric": 0.00023005045659374446,
          "validation_metric": 0.00020759119070135057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8296896739993826
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.002513638239179272,
          "train_loss": 0.0022847638465464115,
          "validation_loss": 0.0017948850290849805,
          "train_metric": 0.00023011302982922643,
          "validation_metric": 0.0001807742373785004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8473671979991195
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.007478674418962328,
          "train_loss": 0.008171825669705868,
          "validation_loss": 0.0071977064944803715,
          "train_metric": 0.0008230363600887358,
          "validation_metric": 0.0007249265909194946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8658521420002216
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.009501460226601921,
          "train_loss": 0.0062537482008337975,
          "validation_loss": 0.004075979348272085,
          "train_metric": 0.0006298546213656664,
          "validation_metric": 0.0004105177358724177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.884021464000398
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.007782308115565684,
          "train_loss": 0.007451843470335007,
          "validation_loss": 0.0059558129869401455,
          "train_metric": 0.0007505223620682955,
          "validation_metric": 0.0005998475826345384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.901994358000593
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.005119529119838262,
          "train_loss": 0.003148523159325123,
          "validation_loss": 0.0024387664161622524,
          "train_metric": 0.00031710771145299077,
          "validation_metric": 0.0002456236397847533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.920253041000251
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0027364241395844147,
          "train_loss": 0.002015328034758568,
          "validation_loss": 0.0017455490306019783,
          "train_metric": 0.0002029764436883852,
          "validation_metric": 0.00017580529674887657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.938385539000592
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.003680453541164752,
          "train_loss": 0.0034533392172306776,
          "validation_loss": 0.003417222062125802,
          "train_metric": 0.0003478076832834631,
          "validation_metric": 0.00034417008282616735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.956306214999131
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0062311565488926135,
          "train_loss": 0.0065756626427173615,
          "validation_loss": 0.004375837743282318,
          "train_metric": 0.0006622766377404332,
          "validation_metric": 0.0004407182859722525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9738404510007967
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.003419980417675106,
          "train_loss": 0.0019040961051359773,
          "validation_loss": 0.0014347813557833433,
          "train_metric": 0.0001917735644383356,
          "validation_metric": 0.0001445059315301478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9919236130008358
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0028708166355499998,
          "train_loss": 0.0037422121968120337,
          "validation_loss": 0.0035124318674206734,
          "train_metric": 0.0003769019094761461,
          "validation_metric": 0.0003537592710927129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.009140743999524
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.002715341726798215,
          "train_loss": 0.0037257822696119547,
          "validation_loss": 0.0023720404133200645,
          "train_metric": 0.0003752471529878676,
          "validation_metric": 0.00023890325974207371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.026438426000823
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.002663046463567298,
          "train_loss": 0.0017837713239714503,
          "validation_loss": 0.0013913578586652875,
          "train_metric": 0.00017965491861104965,
          "validation_metric": 0.0001401324407197535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.044089630000599
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.002976154712087009,
          "train_loss": 0.0016319905407726765,
          "validation_loss": 0.001318478723987937,
          "train_metric": 0.00016436811711173505,
          "validation_metric": 0.00013279233826324344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.062008017999688
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.001999223000893835,
          "train_loss": 0.0015547755174338818,
          "validation_loss": 0.0013206800213083625,
          "train_metric": 0.00015659129712730646,
          "validation_metric": 0.00013301405124366283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.078987229999257
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0017819538297771942,
          "train_loss": 0.0026959984097629786,
          "validation_loss": 0.0018031668150797486,
          "train_metric": 0.00027153108385391533,
          "validation_metric": 0.0001816083531593904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.096545578000587
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0023591756507812534,
          "train_loss": 0.002032922813668847,
          "validation_loss": 0.0017351761925965548,
          "train_metric": 0.00020474856137298048,
          "validation_metric": 0.0001747606002027169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.113857854999878
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0017535538572701626,
          "train_loss": 0.0013710932107642293,
          "validation_loss": 0.0011836981866508722,
          "train_metric": 0.00013809149095322937,
          "validation_metric": 0.00011921774421352893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.131550178999532
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0014361769226525212,
          "train_loss": 0.0014021523529663682,
          "validation_loss": 0.0011854288168251514,
          "train_metric": 0.00014121962885838002,
          "validation_metric": 0.00011939204705413431,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.148646294999708
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.0017079649405786768,
          "train_loss": 0.0013635038631036878,
          "validation_loss": 0.0010821669129654765,
          "train_metric": 0.00013732712250202894,
          "validation_metric": 0.00010899188055191189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.166970897000283
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0019848945103149163,
          "train_loss": 0.003305947408080101,
          "validation_loss": 0.0021190582774579525,
          "train_metric": 0.0003329628671053797,
          "validation_metric": 0.00021342378749977797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.184483357999852
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.002114820686983876,
          "train_loss": 0.0030460129491984844,
          "validation_loss": 0.0022529743146151304,
          "train_metric": 0.0003067832440137863,
          "validation_metric": 0.00022691131744068116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.201850895000462
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0018231498397653922,
          "train_loss": 0.001605037017725408,
          "validation_loss": 0.0013227806193754077,
          "train_metric": 0.00016165342822205275,
          "validation_metric": 0.00013322560698725283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.219244183999763
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.0015460985741810873,
          "train_loss": 0.0016193068586289883,
          "validation_loss": 0.0014360194327309728,
          "train_metric": 0.00016309066268149763,
          "validation_metric": 0.00014463062689173967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.236300313999891
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0019181947463948745,
          "train_loss": 0.0013767688069492579,
          "validation_loss": 0.0011024002451449633,
          "train_metric": 0.00013866310473531485,
          "validation_metric": 0.00011102970893261954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.254213132999212
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0019171326639479958,
          "train_loss": 0.003916984423995018,
          "validation_loss": 0.0035163129214197397,
          "train_metric": 0.00039450431359000504,
          "validation_metric": 0.0003541501355357468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.271894064000662
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.002931443559646141,
          "train_loss": 0.001525631989352405,
          "validation_loss": 0.0012209132546558976,
          "train_metric": 0.0001536560448585078,
          "validation_metric": 0.00012296591012272984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.289643349000471
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0026003494749602396,
          "train_loss": 0.0017142976867035031,
          "validation_loss": 0.0014738781610503793,
          "train_metric": 0.0001726578047964722,
          "validation_metric": 0.00014844360703136772,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.306894179000665
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.006158553238492459,
          "train_loss": 0.006011241115629673,
          "validation_loss": 0.003922694828361273,
          "train_metric": 0.0006054301629774272,
          "validation_metric": 0.0003950794052798301,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.324673028999314
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.012073924590367824,
          "train_loss": 0.005788904614746571,
          "validation_loss": 0.004197000525891781,
          "train_metric": 0.0005830372683703899,
          "validation_metric": 0.0004227065364830196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.341876534999756
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.004506126129854238,
          "train_loss": 0.0025661473628133535,
          "validation_loss": 0.002496721688657999,
          "train_metric": 0.000258452957496047,
          "validation_metric": 0.0002514606458134949,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.359189463999428
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.002448389095661696,
          "train_loss": 0.0015583847416564822,
          "validation_loss": 0.0013361925957724452,
          "train_metric": 0.00015695478941779584,
          "validation_metric": 0.00013457641762215644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.378107088999968
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.0015233335107041057,
          "train_loss": 0.002870924072340131,
          "validation_loss": 0.0023541091941297054,
          "train_metric": 0.00028914891299791634,
          "validation_metric": 0.00023709725064691156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.398064552999131
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0035466974404698703,
          "train_loss": 0.004315533675253391,
          "validation_loss": 0.0027261481154710054,
          "train_metric": 0.00043464478221721947,
          "validation_metric": 0.00027456763200461864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.415159259000575
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0022041316333343275,
          "train_loss": 0.00150584289804101,
          "validation_loss": 0.001246781088411808,
          "train_metric": 0.000151662970893085,
          "validation_metric": 0.00012557122681755573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.432279196000309
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0026300782828911906,
          "train_loss": 0.003133349120616913,
          "validation_loss": 0.0026328484527766705,
          "train_metric": 0.0003155794402118772,
          "validation_metric": 0.00026517087826505303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.449683749999167
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0036281935681472532,
          "train_loss": 0.0027331605087965727,
          "validation_loss": 0.002026243368163705,
          "train_metric": 0.00027527392376214266,
          "validation_metric": 0.0002040757826762274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.466621040999598
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.004021229848149233,
          "train_loss": 0.00384915666654706,
          "validation_loss": 0.0025892090052366257,
          "train_metric": 0.0003876730042975396,
          "validation_metric": 0.00026077564689330757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.48494691400083
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.0021741960663348436,
          "train_loss": 0.0015667431289330125,
          "validation_loss": 0.0014555738307535648,
          "train_metric": 0.0001577966322656721,
          "validation_metric": 0.0001466000685468316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.502409134000118
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0018464066943124635,
          "train_loss": 0.007024601101875305,
          "validation_loss": 0.005076666362583637,
          "train_metric": 0.000707492115907371,
          "validation_metric": 0.000511303311213851,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.520446403999813
        }
      ],
      "best": {
        "epoch": 210,
        "updates": 3360
      },
      "best_validation_metric": 0.00010899188055191189,
      "config": {
        "label": "Concatenated skip",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": true,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2243,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010439917650628855
    },
    "be44f6a2021f_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.27039185469038785,
          "train_loss": 0.10831719636917114,
          "validation_loss": 0.09477120637893677,
          "train_metric": 0.010909311473369598,
          "validation_metric": 0.009545008651912212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02215230399997381
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.08604311244562268,
          "train_loss": 0.08449298143386841,
          "validation_loss": 0.08675695210695267,
          "train_metric": 0.008509824052453041,
          "validation_metric": 0.008737842552363873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04397402900031011
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.060514786979183555,
          "train_loss": 0.047580424696207047,
          "validation_loss": 0.03620985895395279,
          "train_metric": 0.00479212636128068,
          "validation_metric": 0.0036469243932515383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06593956299911952
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.03882622008677572,
          "train_loss": 0.03523484990000725,
          "validation_loss": 0.036904942244291306,
          "train_metric": 0.003548725275322795,
          "validation_metric": 0.0037169307470321655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08675005800068902
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.03007221716688946,
          "train_loss": 0.02236410602927208,
          "validation_loss": 0.01739375852048397,
          "train_metric": 0.0022524306550621986,
          "validation_metric": 0.0017518356908112764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10811500400086516
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.03099765721708536,
          "train_loss": 0.024014728143811226,
          "validation_loss": 0.01601189374923706,
          "train_metric": 0.00241867545992136,
          "validation_metric": 0.0016126594273373485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13024445600058243
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.027296714659314603,
          "train_loss": 0.02134745381772518,
          "validation_loss": 0.012599899433553219,
          "train_metric": 0.0021500373259186745,
          "validation_metric": 0.0012690158328041434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1527221720007219
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.021192819200223312,
          "train_loss": 0.02039089985191822,
          "validation_loss": 0.012964699417352676,
          "train_metric": 0.0020536966621875763,
          "validation_metric": 0.0013057570904493332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17419403699932445
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.030007397668669,
          "train_loss": 0.03706315532326698,
          "validation_loss": 0.03235498070716858,
          "train_metric": 0.0037328654434531927,
          "validation_metric": 0.0032586753368377686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19459128999915265
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.030091354914475232,
          "train_loss": 0.02330632321536541,
          "validation_loss": 0.016952436417341232,
          "train_metric": 0.0023473273031413555,
          "validation_metric": 0.0017073872731998563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21454212500066205
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.03209615338710137,
          "train_loss": 0.024077557027339935,
          "validation_loss": 0.0241977758705616,
          "train_metric": 0.0024250035639852285,
          "validation_metric": 0.002437111223116517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23455574500076182
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.024692562234122306,
          "train_loss": 0.019976388663053513,
          "validation_loss": 0.015801269561052322,
          "train_metric": 0.0020119487307965755,
          "validation_metric": 0.0015914459945634007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2549180950009031
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.016155207675183192,
          "train_loss": 0.011977332644164562,
          "validation_loss": 0.009538279846310616,
          "train_metric": 0.0012063130270689726,
          "validation_metric": 0.0009606605744920671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27547513200079266
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.014945351402275264,
          "train_loss": 0.009783338755369186,
          "validation_loss": 0.007489394396543503,
          "train_metric": 0.0009853420779109001,
          "validation_metric": 0.0007543043466284871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29685694399995555
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.02490110631333664,
          "train_loss": 0.039069853723049164,
          "validation_loss": 0.03355080634355545,
          "train_metric": 0.003934972919523716,
          "validation_metric": 0.003379114205017686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31822607300091477
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.02574498814647086,
          "train_loss": 0.027779800817370415,
          "validation_loss": 0.023996975272893906,
          "train_metric": 0.002797879744321108,
          "validation_metric": 0.002416887553408742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3397131300007459
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.014243885641917586,
          "train_loss": 0.011660939082503319,
          "validation_loss": 0.007793628145009279,
          "train_metric": 0.001174447126686573,
          "validation_metric": 0.0007849456742405891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3600908150001487
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.014597671921364963,
          "train_loss": 0.012175031937658787,
          "validation_loss": 0.009340544231235981,
          "train_metric": 0.0012262248201295733,
          "validation_metric": 0.0009407453471794724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3802256420003687
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.011154008767334744,
          "train_loss": 0.00809113122522831,
          "validation_loss": 0.006103809457272291,
          "train_metric": 0.0008149091736413538,
          "validation_metric": 0.0006147532840259373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4013120399995387
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.01245289073267486,
          "train_loss": 0.01601196639239788,
          "validation_loss": 0.015226091258227825,
          "train_metric": 0.0016126667615026236,
          "validation_metric": 0.0015335161006078124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42193807500007097
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.014955991646274924,
          "train_loss": 0.009401122108101845,
          "validation_loss": 0.006404159590601921,
          "train_metric": 0.0009468466741964221,
          "validation_metric": 0.0006450035143643618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44210579200080247
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.016770219488535076,
          "train_loss": 0.022339433431625366,
          "validation_loss": 0.02121872827410698,
          "train_metric": 0.0022499458864331245,
          "validation_metric": 0.0021370723843574524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.462134628000058
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.018702914705500007,
          "train_loss": 0.008285162039101124,
          "validation_loss": 0.006565423216670752,
          "train_metric": 0.0008344511734321713,
          "validation_metric": 0.0006612454308196902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48285551899971324
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.010724682273576036,
          "train_loss": 0.006374263670295477,
          "validation_loss": 0.004598214756697416,
          "train_metric": 0.0006419925484806299,
          "validation_metric": 0.00046311537153087556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5047671459997218
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.006870915822219104,
          "train_loss": 0.007550710812211037,
          "validation_loss": 0.005002892576158047,
          "train_metric": 0.0007604800048284233,
          "validation_metric": 0.000503873045090586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5251960580008017
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.006254155043279752,
          "train_loss": 0.00838354229927063,
          "validation_loss": 0.006725983694195747,
          "train_metric": 0.0008443596307188272,
          "validation_metric": 0.000677416508551687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5455151070000284
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01153010233247187,
          "train_loss": 0.012380441650748253,
          "validation_loss": 0.00788271427154541,
          "train_metric": 0.0012469131033867598,
          "validation_metric": 0.0007939182105474174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5657501709993085
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01210236651240848,
          "train_loss": 0.012877147644758224,
          "validation_loss": 0.010447257198393345,
          "train_metric": 0.0012969390954822302,
          "validation_metric": 0.001052209408953786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5859068390000175
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.00964202600880526,
          "train_loss": 0.007029639557003975,
          "validation_loss": 0.006290952675044537,
          "train_metric": 0.0007079995120875537,
          "validation_metric": 0.0006336017395369709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6064083100009157
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.005755072925239801,
          "train_loss": 0.007404249627143145,
          "validation_loss": 0.006765091326087713,
          "train_metric": 0.0007457288447767496,
          "validation_metric": 0.0006813552463427186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6264343219991133
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.0072822551446734,
          "train_loss": 0.0044168513268232346,
          "validation_loss": 0.004020411055535078,
          "train_metric": 0.00044484907994046807,
          "validation_metric": 0.0004049211274832487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6471056669997779
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.0059974856703775,
          "train_loss": 0.003360067494213581,
          "validation_loss": 0.0027285295072942972,
          "train_metric": 0.00033841366530396044,
          "validation_metric": 0.0002748075348790735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.667931526000757
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.0036758054775418714,
          "train_loss": 0.007153711281716824,
          "validation_loss": 0.004597029648721218,
          "train_metric": 0.0007204957073554397,
          "validation_metric": 0.00046299598761834204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6884858770008577
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.014916526299202815,
          "train_loss": 0.019613247364759445,
          "validation_loss": 0.017338203266263008,
          "train_metric": 0.0019753742963075638,
          "validation_metric": 0.0017462404211983085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7086607180008286
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014810927532380447,
          "train_loss": 0.006818210706114769,
          "validation_loss": 0.004920503124594688,
          "train_metric": 0.0006867052288725972,
          "validation_metric": 0.000495575019158423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7287517620006838
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.0063115113152889535,
          "train_loss": 0.0066928197629749775,
          "validation_loss": 0.004379418212920427,
          "train_metric": 0.0006740762619301677,
          "validation_metric": 0.0004410789697431028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7490352650002023
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.00683075851702597,
          "train_loss": 0.003941405098885298,
          "validation_loss": 0.002759720664471388,
          "train_metric": 0.0003969638200942427,
          "validation_metric": 0.0002779489441309124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.769118570999126
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.010884807867114432,
          "train_loss": 0.0069356816820800304,
          "validation_loss": 0.005292749498039484,
          "train_metric": 0.0006985364598222077,
          "validation_metric": 0.0005330663407221437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7898523330004537
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.014509179018205032,
          "train_loss": 0.013422197662293911,
          "validation_loss": 0.011587749235332012,
          "train_metric": 0.0013518345076590776,
          "validation_metric": 0.0011670757085084915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8105226660009066
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012970162235433236,
          "train_loss": 0.005950959399342537,
          "validation_loss": 0.0036894117947667837,
          "train_metric": 0.0005993589293211699,
          "validation_metric": 0.00037158402847126126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8307322079999722
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.008675849130668212,
          "train_loss": 0.011716296896338463,
          "validation_loss": 0.011606852523982525,
          "train_metric": 0.0011800226056948304,
          "validation_metric": 0.0011689997045323253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8510210690001259
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.008912016564863734,
          "train_loss": 0.011386807076632977,
          "validation_loss": 0.011526749469339848,
          "train_metric": 0.001146837486885488,
          "validation_metric": 0.0011609321227297187,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.871349900999121
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.008542801624571439,
          "train_loss": 0.012400511652231216,
          "validation_loss": 0.00755611527711153,
          "train_metric": 0.0012489340733736753,
          "validation_metric": 0.0007610243046656251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8919061689994123
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.008397302095545456,
          "train_loss": 0.004891679622232914,
          "validation_loss": 0.004657783079892397,
          "train_metric": 0.0004926720866933465,
          "validation_metric": 0.00046911483514122665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9124848709998332
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.013263224878755864,
          "train_loss": 0.021950790658593178,
          "validation_loss": 0.01871773973107338,
          "train_metric": 0.0022108033299446106,
          "validation_metric": 0.0018851823406293988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9318301099992823
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.023905816895421594,
          "train_loss": 0.011688205413520336,
          "validation_loss": 0.008635766804218292,
          "train_metric": 0.0011771932477131486,
          "validation_metric": 0.0008697628509253263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9515056460004416
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.008953980068326928,
          "train_loss": 0.0068208216689527035,
          "validation_loss": 0.0037311483174562454,
          "train_metric": 0.0006869682110846043,
          "validation_metric": 0.00037578761111944914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9715799400000833
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0050441122220945545,
          "train_loss": 0.0040923915803432465,
          "validation_loss": 0.0031706776935607195,
          "train_metric": 0.0004121706879232079,
          "validation_metric": 0.0003193390730302781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9914335899993603
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.0062146626733010635,
          "train_loss": 0.005316478665918112,
          "validation_loss": 0.004243744071573019,
          "train_metric": 0.0005354562308639288,
          "validation_metric": 0.0004274144012015313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.011681750000207
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.00785657007509144,
          "train_loss": 0.018319016322493553,
          "validation_loss": 0.012462455779314041,
          "train_metric": 0.0018450241768732667,
          "validation_metric": 0.0012551731197163463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.032455292999657
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.01976904197363183,
          "train_loss": 0.014160430990159512,
          "validation_loss": 0.012054923921823502,
          "train_metric": 0.0014261867618188262,
          "validation_metric": 0.001214127754792571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0531284919998143
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008223216631449759,
          "train_loss": 0.004758309572935104,
          "validation_loss": 0.0031408758368343115,
          "train_metric": 0.00047923956299200654,
          "validation_metric": 0.0003163375367876142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0732590099996742
        }
      ],
      "best": {
        "epoch": 32,
        "updates": 512
      },
      "best_validation_metric": 0.0002748075348790735,
      "config": {
        "label": "LayerNorm + tanh + constant LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01657731989433375
    },
    "ff9112ae3885_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3594899591989815,
          "train_loss": 0.10689476132392883,
          "validation_loss": 0.10651864856481552,
          "train_metric": 0.010766048915684223,
          "validation_metric": 0.010728168301284313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05636693699852913
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.08005768503062427,
          "train_loss": 0.05876293405890465,
          "validation_loss": 0.06888415664434433,
          "train_metric": 0.0059183877892792225,
          "validation_metric": 0.006937759928405285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09247872000014468
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.047159415553323925,
          "train_loss": 0.04439746215939522,
          "validation_loss": 0.04207828640937805,
          "train_metric": 0.004471549764275551,
          "validation_metric": 0.004237971268594265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12653412199870218
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.03627265279646963,
          "train_loss": 0.036621857434511185,
          "validation_loss": 0.0362851545214653,
          "train_metric": 0.0036884187720716,
          "validation_metric": 0.003654508152976632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15868030099954922
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.02609884802950546,
          "train_loss": 0.02378969080746174,
          "validation_loss": 0.019558057188987732,
          "train_metric": 0.0023960103280842304,
          "validation_metric": 0.0019698156975209713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19348510899908433
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.020956436870619655,
          "train_loss": 0.027749165892601013,
          "validation_loss": 0.02930276282131672,
          "train_metric": 0.002794794039800763,
          "validation_metric": 0.0029512671753764153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22367208099967684
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.017646923603024334,
          "train_loss": 0.0202290341258049,
          "validation_loss": 0.020458824932575226,
          "train_metric": 0.0020373943261802197,
          "validation_metric": 0.002060537924990058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2531994279997889
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.013338577758986503,
          "train_loss": 0.02697916328907013,
          "validation_loss": 0.02338114194571972,
          "train_metric": 0.0027172425761818886,
          "validation_metric": 0.0023548630997538567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2835775969997485
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.03040843659255188,
          "train_loss": 0.02058366686105728,
          "validation_loss": 0.021947549656033516,
          "train_metric": 0.002073111478239298,
          "validation_metric": 0.002210476668551564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31236049799917964
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.02013958990573883,
          "train_loss": 0.014335322193801403,
          "validation_loss": 0.011812934651970863,
          "train_metric": 0.0014438010985031724,
          "validation_metric": 0.0011897553922608495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.326241435999691
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.02076608722563833,
          "train_loss": 0.014632223173975945,
          "validation_loss": 0.018382254987955093,
          "train_metric": 0.0014737040037289262,
          "validation_metric": 0.0018513933755457401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3395075270000234
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.01418373743945267,
          "train_loss": 0.00860548485070467,
          "validation_loss": 0.010486365295946598,
          "train_metric": 0.0008667129441164434,
          "validation_metric": 0.0010561482049524784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35374193899951933
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.0173728795198258,
          "train_loss": 0.014441790990531445,
          "validation_loss": 0.011986266821622849,
          "train_metric": 0.0014545242302119732,
          "validation_metric": 0.0012072130339220166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36677902299925336
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.014862617274047807,
          "train_loss": 0.0154472840949893,
          "validation_loss": 0.010748198255896568,
          "train_metric": 0.001555793802253902,
          "validation_metric": 0.0010825189528986812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3800760679987434
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.025403889041626826,
          "train_loss": 0.041615959256887436,
          "validation_loss": 0.03319608047604561,
          "train_metric": 0.00419140700250864,
          "validation_metric": 0.003343387506902218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3931334729986702
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.021664694766514003,
          "train_loss": 0.01344259362667799,
          "validation_loss": 0.01134704053401947,
          "train_metric": 0.001353888656012714,
          "validation_metric": 0.0011428323341533542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40599716499855276
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.014402267828700133,
          "train_loss": 0.01665353961288929,
          "validation_loss": 0.022806476801633835,
          "train_metric": 0.0016772834351286292,
          "validation_metric": 0.0022969848942011595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4193448719997832
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.013661834585946053,
          "train_loss": 0.009685010649263859,
          "validation_loss": 0.011342911049723625,
          "train_metric": 0.0009754388011060655,
          "validation_metric": 0.0011424164986237884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43260765299964987
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.015104164893273264,
          "train_loss": 0.010758163407444954,
          "validation_loss": 0.009069183841347694,
          "train_metric": 0.0010835228022187948,
          "validation_metric": 0.0009134150459431112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44678310899871576
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.009322157100541517,
          "train_loss": 0.009018831886351109,
          "validation_loss": 0.0059044333174824715,
          "train_metric": 0.0009083438199013472,
          "validation_metric": 0.0005946729215793312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46115633899898967
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.006748501502443105,
          "train_loss": 0.005163723137229681,
          "validation_loss": 0.012859508395195007,
          "train_metric": 0.0005200713058002293,
          "validation_metric": 0.0012951624812558293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47469627399914316
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.012611139762157109,
          "train_loss": 0.007669901009649038,
          "validation_loss": 0.007750188000500202,
          "train_metric": 0.0007724842871539295,
          "validation_metric": 0.0007805704954080284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.487670149999758
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01289312489097938,
          "train_loss": 0.010503493249416351,
          "validation_loss": 0.02323693037033081,
          "train_metric": 0.00105787324719131,
          "validation_metric": 0.0023403384257107973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5009213519988407
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019172809479641728,
          "train_loss": 0.012283912859857082,
          "validation_loss": 0.01608487218618393,
          "train_metric": 0.0012371905613690615,
          "validation_metric": 0.0016200094250962138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5145333719992777
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.010826966870808974,
          "train_loss": 0.009194883517920971,
          "validation_loss": 0.009451521560549736,
          "train_metric": 0.0009260750375688076,
          "validation_metric": 0.0009519225568510592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5279700489991228
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.008604193280916661,
          "train_loss": 0.005289078690111637,
          "validation_loss": 0.007678186986595392,
          "train_metric": 0.0005326967220753431,
          "validation_metric": 0.0007733188685961068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5413076029999502
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01894094643648714,
          "train_loss": 0.02910364232957363,
          "validation_loss": 0.04844817891716957,
          "train_metric": 0.0029312120750546455,
          "validation_metric": 0.004879523068666458,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5544418749996112
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.020172985445242375,
          "train_loss": 0.013426442630589008,
          "validation_loss": 0.014586572535336018,
          "train_metric": 0.001352262101136148,
          "validation_metric": 0.0014691061805933714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5674463569994259
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.012159680161857978,
          "train_loss": 0.010088502429425716,
          "validation_loss": 0.016597436740994453,
          "train_metric": 0.001016077003441751,
          "validation_metric": 0.0016716329846531153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5809858949996851
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.008545430930098519,
          "train_loss": 0.005742490291595459,
          "validation_loss": 0.012234696187078953,
          "train_metric": 0.0005783626111224294,
          "validation_metric": 0.0012322337133809924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5978097389997856
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.01068879006197676,
          "train_loss": 0.030793415382504463,
          "validation_loss": 0.04476664587855339,
          "train_metric": 0.003101400099694729,
          "validation_metric": 0.004508732818067074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6155524919995514
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.022684087161906064,
          "train_loss": 0.014037562534213066,
          "validation_loss": 0.021958045661449432,
          "train_metric": 0.001413811813108623,
          "validation_metric": 0.0022115339525043964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6320775879994471
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.014062464353628457,
          "train_loss": 0.008616777136921883,
          "validation_loss": 0.017803922295570374,
          "train_metric": 0.0008678502053953707,
          "validation_metric": 0.001793145900592208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6526733239988971
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.010633023019181564,
          "train_loss": 0.01081909891217947,
          "validation_loss": 0.013694113120436668,
          "train_metric": 0.0010896598687395453,
          "validation_metric": 0.0013792209792882204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6668453089987452
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.00751630833838135,
          "train_loss": 0.0042108274064958096,
          "validation_loss": 0.013436044566333294,
          "train_metric": 0.0004240990965627134,
          "validation_metric": 0.0013532292796298862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.687920596999902
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.005518843223399017,
          "train_loss": 0.003819165751338005,
          "validation_loss": 0.011026397347450256,
          "train_metric": 0.00038465243414975703,
          "validation_metric": 0.0011105382582172751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7053457450001588
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.005732314566557761,
          "train_loss": 0.007129563018679619,
          "validation_loss": 0.023137005046010017,
          "train_metric": 0.0007180635002441704,
          "validation_metric": 0.002330274321138859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7232967689997167
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.006866877149150241,
          "train_loss": 0.004204190336167812,
          "validation_loss": 0.01876065507531166,
          "train_metric": 0.0004234306688886136,
          "validation_metric": 0.0018895044922828674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7399766649996309
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01180263006244786,
          "train_loss": 0.01606961339712143,
          "validation_loss": 0.0300159715116024,
          "train_metric": 0.0016184727428480983,
          "validation_metric": 0.0030230984557420015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7576011279998056
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.01294154638890177,
          "train_loss": 0.007650719955563545,
          "validation_loss": 0.016415277495980263,
          "train_metric": 0.0007705523748882115,
          "validation_metric": 0.0016532866284251213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7759719850000693
        }
      ],
      "best": {
        "epoch": 20,
        "updates": 320
      },
      "best_validation_metric": 0.0005946729215793312,
      "config": {
        "label": "LayerNorm + relu + constant LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.024385916459697208
    },
    "9d4c0f0efefb_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7050002012401819,
          "train_loss": 0.28691911697387695,
          "validation_loss": 0.20147113502025604,
          "train_metric": 0.028897441923618317,
          "validation_metric": 0.020291438326239586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02108714099995268
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.17274193931370974,
          "train_loss": 0.10724043846130371,
          "validation_loss": 0.08937232196331024,
          "train_metric": 0.010800864547491074,
          "validation_metric": 0.009001253172755241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05127141800039681
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.09413822274655104,
          "train_loss": 0.07160875201225281,
          "validation_loss": 0.07328219711780548,
          "train_metric": 0.007212171331048012,
          "validation_metric": 0.007380714640021324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07333429900063493
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.06729799672029912,
          "train_loss": 0.055052559822797775,
          "validation_loss": 0.04895579814910889,
          "train_metric": 0.005544692277908325,
          "validation_metric": 0.0049306489527225494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09219797699915944
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.04980971175245941,
          "train_loss": 0.04863223060965538,
          "validation_loss": 0.041030172258615494,
          "train_metric": 0.004898060578852892,
          "validation_metric": 0.004132408648729324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10455328799980634
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.04394618363585323,
          "train_loss": 0.04424542933702469,
          "validation_loss": 0.039304472506046295,
          "train_metric": 0.0044562374241650105,
          "validation_metric": 0.003958602901548147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11717182500069612
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.041344128898344934,
          "train_loss": 0.03573659062385559,
          "validation_loss": 0.0313391275703907,
          "train_metric": 0.003599258605390787,
          "validation_metric": 0.0031563621014356613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1295972959997016
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.03503185708541423,
          "train_loss": 0.03175372630357742,
          "validation_loss": 0.02884584479033947,
          "train_metric": 0.0031981193460524082,
          "validation_metric": 0.0029052477329969406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14187053800014837
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.031045667768921703,
          "train_loss": 0.02844640426337719,
          "validation_loss": 0.024947280064225197,
          "train_metric": 0.002865017391741276,
          "validation_metric": 0.0025125986430794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15401680300055887
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.02896959410281852,
          "train_loss": 0.028151480481028557,
          "validation_loss": 0.02492579258978367,
          "train_metric": 0.0028353140223771334,
          "validation_metric": 0.002510434715077281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16619175500090932
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.027469585184007883,
          "train_loss": 0.02537483721971512,
          "validation_loss": 0.02314770594239235,
          "train_metric": 0.0025556606706231833,
          "validation_metric": 0.002331352327018976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17845626800044556
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.02965931163635105,
          "train_loss": 0.02589038759469986,
          "validation_loss": 0.02060387097299099,
          "train_metric": 0.002607585396617651,
          "validation_metric": 0.002075146185234189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1907631499998388
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.02585070999339223,
          "train_loss": 0.019568506628274918,
          "validation_loss": 0.01680118776857853,
          "train_metric": 0.001970868557691574,
          "validation_metric": 0.0016921540955081582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2031447000008484
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.022595795046072453,
          "train_loss": 0.023672930896282196,
          "validation_loss": 0.02443583309650421,
          "train_metric": 0.002384250983595848,
          "validation_metric": 0.0024610876571387053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21516167300069355
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.018558361451141536,
          "train_loss": 0.019495747983455658,
          "validation_loss": 0.01587093062698841,
          "train_metric": 0.0019635402131825686,
          "validation_metric": 0.0015984619967639446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3039982730006159
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.016856092668604106,
          "train_loss": 0.014491734094917774,
          "validation_loss": 0.013611108995974064,
          "train_metric": 0.001459554536268115,
          "validation_metric": 0.0013708610786125064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31699061899962544
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.015511883742874488,
          "train_loss": 0.013932151719927788,
          "validation_loss": 0.013291517272591591,
          "train_metric": 0.0014031953178346157,
          "validation_metric": 0.0013386729406192899,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.329601122000895
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.01509113804786466,
          "train_loss": 0.013327684253454208,
          "validation_loss": 0.010525483638048172,
          "train_metric": 0.0013423156924545765,
          "validation_metric": 0.001060088281519711,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34222168999986025
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.014127399030257948,
          "train_loss": 0.013853229582309723,
          "validation_loss": 0.011781918816268444,
          "train_metric": 0.0013952463632449508,
          "validation_metric": 0.0011866316199302673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35395124000024225
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.016405421716626734,
          "train_loss": 0.02321820706129074,
          "validation_loss": 0.018766550347208977,
          "train_metric": 0.0023384527303278446,
          "validation_metric": 0.0018900982104241848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3657658209995134
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.012881455113529228,
          "train_loss": 0.010585556738078594,
          "validation_loss": 0.010381308384239674,
          "train_metric": 0.0010661386186257005,
          "validation_metric": 0.001045567449182272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37788990499939246
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.013161487702745944,
          "train_loss": 0.015021375380456448,
          "validation_loss": 0.012616414576768875,
          "train_metric": 0.001512897782959044,
          "validation_metric": 0.0012706790585070848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38965459400060354
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.024356837471714243,
          "train_loss": 0.019724711775779724,
          "validation_loss": 0.019498197361826897,
          "train_metric": 0.0019866006914526224,
          "validation_metric": 0.0019637872464954853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40136010400055966
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.015044581901747733,
          "train_loss": 0.010036997497081757,
          "validation_loss": 0.008834431879222393,
          "train_metric": 0.0010108896531164646,
          "validation_metric": 0.0008897717343643308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41365130499980296
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.010508958119316958,
          "train_loss": 0.011870963498950005,
          "validation_loss": 0.011397434398531914,
          "train_metric": 0.001195599907077849,
          "validation_metric": 0.0011479078093543649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42526886199993896
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.010162755992496386,
          "train_loss": 0.00944861676543951,
          "validation_loss": 0.007191171403974295,
          "train_metric": 0.0009516301215626299,
          "validation_metric": 0.000724268495105207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4372638240001834
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.009371989115606993,
          "train_loss": 0.008101182989776134,
          "validation_loss": 0.007279512472450733,
          "train_metric": 0.0008159214630723,
          "validation_metric": 0.0007331657689064741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44872750700051256
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.009293428069213405,
          "train_loss": 0.0072727813385427,
          "validation_loss": 0.00683234166353941,
          "train_metric": 0.000732487766072154,
          "validation_metric": 0.0006881284061819315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46075396000014734
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.009531655348837376,
          "train_loss": 0.008531746454536915,
          "validation_loss": 0.008733045309782028,
          "train_metric": 0.000859286286868155,
          "validation_metric": 0.000879560480825603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4724740500005282
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.012777429656125605,
          "train_loss": 0.007119226269423962,
          "validation_loss": 0.0069771516136825085,
          "train_metric": 0.0007170223980210721,
          "validation_metric": 0.0007027131505310535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4840672469999845
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008610082179075107,
          "train_loss": 0.017902640625834465,
          "validation_loss": 0.01745501719415188,
          "train_metric": 0.001803088467568159,
          "validation_metric": 0.0017580052372068167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49523113100076444
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01600136494380422,
          "train_loss": 0.006772779859602451,
          "validation_loss": 0.006944308988749981,
          "train_metric": 0.0006821295828558505,
          "validation_metric": 0.0006994053837843239,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5071128820000013
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.010719839134253561,
          "train_loss": 0.013310972601175308,
          "validation_loss": 0.010746842250227928,
          "train_metric": 0.0013406324433162808,
          "validation_metric": 0.0010823825141415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5188429209993046
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.010145125153940171,
          "train_loss": 0.006453029811382294,
          "validation_loss": 0.006798273883759975,
          "train_metric": 0.0006499254959635437,
          "validation_metric": 0.0006846972974017262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.531295840999519
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.007759873849863652,
          "train_loss": 0.007150338962674141,
          "validation_loss": 0.0076555730774998665,
          "train_metric": 0.0007201560074463487,
          "validation_metric": 0.0007710413774475455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5435970960006671
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.00830216116446536,
          "train_loss": 0.006321385968476534,
          "validation_loss": 0.007175783161073923,
          "train_metric": 0.0006366668967530131,
          "validation_metric": 0.0007227185415104032,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.555643866000537
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.006399199482984841,
          "train_loss": 0.006602111272513866,
          "validation_loss": 0.005680766887962818,
          "train_metric": 0.000664940511342138,
          "validation_metric": 0.000572146032936871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5685257589993853
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.006437242642277852,
          "train_loss": 0.005138068925589323,
          "validation_loss": 0.005342321004718542,
          "train_metric": 0.0005174875259399414,
          "validation_metric": 0.0005380589864216745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5810002060006809
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.007077939502778463,
          "train_loss": 0.010447354055941105,
          "validation_loss": 0.007250035181641579,
          "train_metric": 0.0010522193042561412,
          "validation_metric": 0.0007301969453692436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5928524390001257
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.008693394789588638,
          "train_loss": 0.008410956710577011,
          "validation_loss": 0.007602078374475241,
          "train_metric": 0.0008471207693219185,
          "validation_metric": 0.0007656534435227513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6044808450005803
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.006143342819996178,
          "train_loss": 0.004438408184796572,
          "validation_loss": 0.004541188478469849,
          "train_metric": 0.00044702019658870995,
          "validation_metric": 0.000457371847005561,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6168646950009133
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.00467075064079836,
          "train_loss": 0.004335374571382999,
          "validation_loss": 0.004371497314423323,
          "train_metric": 0.0004366431094240397,
          "validation_metric": 0.0004402811755426228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6293752799992944
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.006131597689090995,
          "train_loss": 0.006984661798924208,
          "validation_loss": 0.006742445752024651,
          "train_metric": 0.000703469617292285,
          "validation_metric": 0.0006790744373574853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6410655079998833
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.006669067879556678,
          "train_loss": 0.006646012421697378,
          "validation_loss": 0.006550450809299946,
          "train_metric": 0.0006693620816804469,
          "validation_metric": 0.0006597373867407441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.653043099999195
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.005331256470526569,
          "train_loss": 0.0061996327713131905,
          "validation_loss": 0.004703868180513382,
          "train_metric": 0.000624404230620712,
          "validation_metric": 0.00047375637223012745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6675931310001033
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.005578597789281048,
          "train_loss": 0.0043454356491565704,
          "validation_loss": 0.004700619727373123,
          "train_metric": 0.0004376563592813909,
          "validation_metric": 0.00047342918696813285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6806231579994346
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.005856705502083059,
          "train_loss": 0.005570260342210531,
          "validation_loss": 0.005530715920031071,
          "train_metric": 0.0005610162625089288,
          "validation_metric": 0.0005570334033109248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6969411410009343
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0055267462084884755,
          "train_loss": 0.012103386223316193,
          "validation_loss": 0.011215858161449432,
          "train_metric": 0.0012190088164061308,
          "validation_metric": 0.0011296201264485717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7128798579997238
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.006211136584170163,
          "train_loss": 0.004434387199580669,
          "validation_loss": 0.003139832755550742,
          "train_metric": 0.0004466152749955654,
          "validation_metric": 0.0003162324719596654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7296177690004697
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.004359201273473445,
          "train_loss": 0.003302095690742135,
          "validation_loss": 0.0028553460724651814,
          "train_metric": 0.0003325750003568828,
          "validation_metric": 0.0002875800128094852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7461572659994999
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.005176383794605499,
          "train_loss": 0.009496084414422512,
          "validation_loss": 0.00835470948368311,
          "train_metric": 0.0009564108913764358,
          "validation_metric": 0.0008414557087235153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8886641810004221
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.013195205814554356,
          "train_loss": 0.008537136018276215,
          "validation_loss": 0.009820844978094101,
          "train_metric": 0.000859829131513834,
          "validation_metric": 0.0009891196386888623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9009760150001966
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.009484574678936042,
          "train_loss": 0.009770087897777557,
          "validation_loss": 0.009624110534787178,
          "train_metric": 0.0009840074926614761,
          "validation_metric": 0.0009693052852526307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9128965769996285
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.006253362771531101,
          "train_loss": 0.003888901323080063,
          "validation_loss": 0.004088472109287977,
          "train_metric": 0.0003916759160347283,
          "validation_metric": 0.0004117759526707232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9248396610000782
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.0072447815400664695,
          "train_loss": 0.012647396884858608,
          "validation_loss": 0.011404111050069332,
          "train_metric": 0.0012737995712086558,
          "validation_metric": 0.001148580340668559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9365120839993324
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.006679106616502395,
          "train_loss": 0.004418913275003433,
          "validation_loss": 0.0045089502818882465,
          "train_metric": 0.0004450567066669464,
          "validation_metric": 0.0004541249363683164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9485951149999892
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.0038562213449040428,
          "train_loss": 0.003231059992685914,
          "validation_loss": 0.003004261292517185,
          "train_metric": 0.0003254205221310258,
          "validation_metric": 0.0003025782061740756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9603677420000167
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0075220788894512225,
          "train_loss": 0.015141400508582592,
          "validation_loss": 0.01278905663639307,
          "train_metric": 0.0015249863499775529,
          "validation_metric": 0.0012880669673904777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9722975050008245
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.014357350315549411,
          "train_loss": 0.011311506852507591,
          "validation_loss": 0.010252506472170353,
          "train_metric": 0.0011392533779144287,
          "validation_metric": 0.0010325948242098093,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9841736180005682
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.008004285096831154,
          "train_loss": 0.003624373348429799,
          "validation_loss": 0.002938582096248865,
          "train_metric": 0.0003650336293503642,
          "validation_metric": 0.00029596322565339506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9959830269999657
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0047952501408872195,
          "train_loss": 0.003691833931952715,
          "validation_loss": 0.0035481785889714956,
          "train_metric": 0.0003718279767781496,
          "validation_metric": 0.0003573595604393631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0094304209997063
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.003888095072397846,
          "train_loss": 0.0029377213213592768,
          "validation_loss": 0.002188287675380707,
          "train_metric": 0.000295876496238634,
          "validation_metric": 0.00022039632312953472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0262665869995544
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006986497985053575,
          "train_loss": 0.016044432297348976,
          "validation_loss": 0.012086225673556328,
          "train_metric": 0.0016159364022314548,
          "validation_metric": 0.0012172802817076445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0445803759994305
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.009794601559406146,
          "train_loss": 0.004587536212056875,
          "validation_loss": 0.0032667950727045536,
          "train_metric": 0.0004620398976840079,
          "validation_metric": 0.00032901964732445776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0619932370009337
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.004315413945732871,
          "train_loss": 0.004174700006842613,
          "validation_loss": 0.004023138899356127,
          "train_metric": 0.00042046053567901254,
          "validation_metric": 0.00040519580943509936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0780512119999912
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.00364719620483811,
          "train_loss": 0.0032703187316656113,
          "validation_loss": 0.003099875058978796,
          "train_metric": 0.00032937456853687763,
          "validation_metric": 0.0003122081106994301,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0969954840002174
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.003620768344262615,
          "train_loss": 0.004391487222164869,
          "validation_loss": 0.0035311742685735226,
          "train_metric": 0.00044229452032595873,
          "validation_metric": 0.0003556468873284757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1129134029997658
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.004531054339167895,
          "train_loss": 0.0042821685783565044,
          "validation_loss": 0.0036448666360229254,
          "train_metric": 0.0004312843084335327,
          "validation_metric": 0.0003670975856948644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1383593030004704
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.005244120802672114,
          "train_loss": 0.008316396735608578,
          "validation_loss": 0.007668662816286087,
          "train_metric": 0.0008375971228815615,
          "validation_metric": 0.0007723596063442528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.156162048000624
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.004875374241237296,
          "train_loss": 0.0029918800573796034,
          "validation_loss": 0.0025843351613730192,
          "train_metric": 0.00030133119435049593,
          "validation_metric": 0.0002602847816888243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1741243169999507
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.00293776031685411,
          "train_loss": 0.0026360079646110535,
          "validation_loss": 0.002063687890768051,
          "train_metric": 0.00026548904133960605,
          "validation_metric": 0.0002078470861306414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1922175100007735
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0033687393224681728,
          "train_loss": 0.0024132265243679285,
          "validation_loss": 0.0019690815825015306,
          "train_metric": 0.00024305135593749583,
          "validation_metric": 0.00019831868121400476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2101750420006283
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.003837940843368415,
          "train_loss": 0.003160844324156642,
          "validation_loss": 0.0030723069794476032,
          "train_metric": 0.0003183486114721745,
          "validation_metric": 0.00030943151796236634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2280066280000028
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.003138846592264599,
          "train_loss": 0.0027508512139320374,
          "validation_loss": 0.002467536134645343,
          "train_metric": 0.000277055602055043,
          "validation_metric": 0.0002485211589373648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2571034320008039
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.003621138501330279,
          "train_loss": 0.007537864148616791,
          "validation_loss": 0.00548481335863471,
          "train_metric": 0.0007591861067339778,
          "validation_metric": 0.0005524103180505335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2731927780005208
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.009200153588608373,
          "train_loss": 0.011608482338488102,
          "validation_loss": 0.00968504510819912,
          "train_metric": 0.0011691637337207794,
          "validation_metric": 0.0009754422353580594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2894379900008062
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.008320151231600903,
          "train_loss": 0.003419185522943735,
          "validation_loss": 0.0030820276588201523,
          "train_metric": 0.00034436784335412085,
          "validation_metric": 0.00031041057081893086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3022588820003875
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.003635357614257373,
          "train_loss": 0.0028572920709848404,
          "validation_loss": 0.0026446899864822626,
          "train_metric": 0.00028777599800378084,
          "validation_metric": 0.00026636349502950907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.32768673999999
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.006579027132829651,
          "train_loss": 0.005779228638857603,
          "validation_loss": 0.0050468286499381065,
          "train_metric": 0.0005820626975037158,
          "validation_metric": 0.0005082980496808887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.35436862700044
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.003526122134644538,
          "train_loss": 0.002264071721583605,
          "validation_loss": 0.0018541770987212658,
          "train_metric": 0.0002280289918417111,
          "validation_metric": 0.0001867459068307653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3863669189995562
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.003170347372360993,
          "train_loss": 0.004942181520164013,
          "validation_loss": 0.0042056068778038025,
          "train_metric": 0.000497758446726948,
          "validation_metric": 0.00042357330676168203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4220088170004601
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.00799715134780854,
          "train_loss": 0.0168495811522007,
          "validation_loss": 0.015240857377648354,
          "train_metric": 0.0016970282886177301,
          "validation_metric": 0.0015350033063441515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4598941750009544
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.014071476840399555,
          "train_loss": 0.004382811486721039,
          "validation_loss": 0.0031371060758829117,
          "train_metric": 0.00044142076512798667,
          "validation_metric": 0.0003159578482154757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.557122435000565
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.009563856317981845,
          "train_loss": 0.010939707048237324,
          "validation_loss": 0.008279618807137012,
          "train_metric": 0.0011018074583262205,
          "validation_metric": 0.0008338928455486894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.680010804999256
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.009972337531507947,
          "train_loss": 0.011797497980296612,
          "validation_loss": 0.011797357350587845,
          "train_metric": 0.0011882007820531726,
          "validation_metric": 0.0011881865793839097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7013533099998313
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.01011471898527816,
          "train_loss": 0.0035584564320743084,
          "validation_loss": 0.0022907943930476904,
          "train_metric": 0.00035839469637721777,
          "validation_metric": 0.00023072041221894324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7226359380001668
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.004846902258577757,
          "train_loss": 0.0026799356564879417,
          "validation_loss": 0.002335650846362114,
          "train_metric": 0.00026991331833414733,
          "validation_metric": 0.000235238199820742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8634324340000603
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.004376559018055559,
          "train_loss": 0.0026580209378153086,
          "validation_loss": 0.002035677433013916,
          "train_metric": 0.0002677061129361391,
          "validation_metric": 0.00020502597908489406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8870307790002698
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0027774416194006335,
          "train_loss": 0.0021549833472818136,
          "validation_loss": 0.0018619997426867485,
          "train_metric": 0.00021704203390982002,
          "validation_metric": 0.00018753382028080523,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9072756770001433
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0028143735144112725,
          "train_loss": 0.0024574536364525557,
          "validation_loss": 0.0019086049869656563,
          "train_metric": 0.0002475057262927294,
          "validation_metric": 0.00019222772971261293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.925600239999767
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.00273829284014937,
          "train_loss": 0.0020827604457736015,
          "validation_loss": 0.0015389835461974144,
          "train_metric": 0.00020976799714844674,
          "validation_metric": 0.00015500077279284596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9432295730002807
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0020525133704722975,
          "train_loss": 0.001966248033568263,
          "validation_loss": 0.0016453972784802318,
          "train_metric": 0.00019803331815637648,
          "validation_metric": 0.00016571836022194475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9611790369999653
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.002772823609120678,
          "train_loss": 0.0027843182906508446,
          "validation_loss": 0.0023238176945596933,
          "train_metric": 0.00028042634949088097,
          "validation_metric": 0.00023404639796353877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.979077772999517
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0023520003887824714,
          "train_loss": 0.002927845809608698,
          "validation_loss": 0.002791647333651781,
          "train_metric": 0.0002948818728327751,
          "validation_metric": 0.0002811645099427551,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9975357439998334
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.002655118893017061,
          "train_loss": 0.0028100123163312674,
          "validation_loss": 0.0019661064725369215,
          "train_metric": 0.0002830141456797719,
          "validation_metric": 0.00019801905727945268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.015712492999228
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0022596421695197932,
          "train_loss": 0.003142349421977997,
          "validation_loss": 0.0026928347069770098,
          "train_metric": 0.0003164859372191131,
          "validation_metric": 0.00027121242601424456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0349540860006528
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0035817066454910673,
          "train_loss": 0.0018582985503599048,
          "validation_loss": 0.0017364188097417355,
          "train_metric": 0.00018716100021265447,
          "validation_metric": 0.00017488576122559607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.055447336000725
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.002570564545749221,
          "train_loss": 0.0018190513364970684,
          "validation_loss": 0.0014340305933728814,
          "train_metric": 0.0001832082198234275,
          "validation_metric": 0.00014443030522670597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.074353334999614
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.00413549212316866,
          "train_loss": 0.0051988279446959496,
          "validation_loss": 0.004981496371328831,
          "train_metric": 0.0005236069555394351,
          "validation_metric": 0.0005017180810682476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.092268950000289
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.004806055116205243,
          "train_loss": 0.004731020890176296,
          "validation_loss": 0.004327301401644945,
          "train_metric": 0.0004764910554513335,
          "validation_metric": 0.00043582997750490904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1147442309993494
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0023940183000377147,
          "train_loss": 0.0020271185785531998,
          "validation_loss": 0.0017923389095813036,
          "train_metric": 0.00020416395273059607,
          "validation_metric": 0.00018051781808026135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1394474589997117
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.001879721521618194,
          "train_loss": 0.0030142818577587605,
          "validation_loss": 0.0026476734783500433,
          "train_metric": 0.00030358743970282376,
          "validation_metric": 0.00026666399207897484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.181069211999784
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0023856415937189013,
          "train_loss": 0.002041477710008621,
          "validation_loss": 0.0018513654358685017,
          "train_metric": 0.00020561016572173685,
          "validation_metric": 0.00018646274111233652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2273599679992913
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.001754217281813908,
          "train_loss": 0.0021812543272972107,
          "validation_loss": 0.0015532353427261114,
          "train_metric": 0.00021968795044813305,
          "validation_metric": 0.0001564361882628873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.27526720700007
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.00433738558422192,
          "train_loss": 0.008574957959353924,
          "validation_loss": 0.007257168181240559,
          "train_metric": 0.0008636383572593331,
          "validation_metric": 0.0007309154025278986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2979128070001025
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.005999747118039522,
          "train_loss": 0.00329317944124341,
          "validation_loss": 0.0026212146040052176,
          "train_metric": 0.0003316769434604794,
          "validation_metric": 0.000263999099843204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3161509790006676
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0031622368696844205,
          "train_loss": 0.00437642727047205,
          "validation_loss": 0.00346594606526196,
          "train_metric": 0.0004407778033055365,
          "validation_metric": 0.00034907739609479904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3353132930005813
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0024877173964341637,
          "train_loss": 0.0016995479818433523,
          "validation_loss": 0.001450573792681098,
          "train_metric": 0.00017117225797846913,
          "validation_metric": 0.00014609648496843874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3538332139996783
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0016295983878080733,
          "train_loss": 0.001925049815326929,
          "validation_loss": 0.001882924698293209,
          "train_metric": 0.00019388401415199041,
          "validation_metric": 0.00018964125774800777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3725484570004483
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0018218346704088617,
          "train_loss": 0.0018765453714877367,
          "validation_loss": 0.00150972418487072,
          "train_metric": 0.00018899879069067538,
          "validation_metric": 0.0001520539226476103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3921611059995485
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.001899085043987725,
          "train_loss": 0.0017630977090448141,
          "validation_loss": 0.0013492224970832467,
          "train_metric": 0.00017757274326868355,
          "validation_metric": 0.00013588875299319625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.411080445999687
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.0020710459521069424,
          "train_loss": 0.0029842774383723736,
          "validation_loss": 0.0023707463406026363,
          "train_metric": 0.00030056550167500973,
          "validation_metric": 0.00023877287458162755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4289437780007574
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.003280071332483203,
          "train_loss": 0.0018610230181366205,
          "validation_loss": 0.0017006039852276444,
          "train_metric": 0.00018743543478194624,
          "validation_metric": 0.00017127860337495804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.44737586500014
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0028016531050525373,
          "train_loss": 0.0016006631776690483,
          "validation_loss": 0.001229457906447351,
          "train_metric": 0.00016121292719617486,
          "validation_metric": 0.00012382649583742023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4670786610004143
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0034923766761494335,
          "train_loss": 0.002619336359202862,
          "validation_loss": 0.00255479640327394,
          "train_metric": 0.00026380998315289617,
          "validation_metric": 0.00025730972993187606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4857923409999785
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.004609447758411989,
          "train_loss": 0.0017133104847744107,
          "validation_loss": 0.0016148354625329375,
          "train_metric": 0.00017255837155971676,
          "validation_metric": 0.00016264031000901014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.50487844700001
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.002180600495194085,
          "train_loss": 0.0016441468615084887,
          "validation_loss": 0.0012680244399234653,
          "train_metric": 0.0001655924424994737,
          "validation_metric": 0.00012771078036166728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5231818619995465
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.001749370904690295,
          "train_loss": 0.0016652756603434682,
          "validation_loss": 0.0013629624154418707,
          "train_metric": 0.00016772045637480915,
          "validation_metric": 0.00013727258192375302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5409142059997976
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0014212412825145293,
          "train_loss": 0.001287868362851441,
          "validation_loss": 0.0012057229178026319,
          "train_metric": 0.0001297093986067921,
          "validation_metric": 0.0001214360017911531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5599740449997626
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0014129820956441108,
          "train_loss": 0.00273360894061625,
          "validation_loss": 0.0024631742853671312,
          "train_metric": 0.00027531912201084197,
          "validation_metric": 0.00024808189482428133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5790394480009127
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0017202673689098447,
          "train_loss": 0.001275935908779502,
          "validation_loss": 0.0011139403795823455,
          "train_metric": 0.0001285076141357422,
          "validation_metric": 0.00011219199222978204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.596973268000511
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.002182695769079146,
          "train_loss": 0.001565875019878149,
          "validation_loss": 0.0017155869863927364,
          "train_metric": 0.00015770918980706483,
          "validation_metric": 0.00017278762243222445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6159625820000656
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0018418122053844854,
          "train_loss": 0.0020127161405980587,
          "validation_loss": 0.002201995113864541,
          "train_metric": 0.000202713388716802,
          "validation_metric": 0.00022177686332724988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7357256049999705
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0017088500353565905,
          "train_loss": 0.0018964973278343678,
          "validation_loss": 0.001574307563714683,
          "train_metric": 0.00019100826466456056,
          "validation_metric": 0.00015855849778745323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7559520560007513
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0014578693317162106,
          "train_loss": 0.002385388594120741,
          "validation_loss": 0.0021893430966883898,
          "train_metric": 0.00024024758022278547,
          "validation_metric": 0.00022050258121453226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7746944360005728
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.002566412636952009,
          "train_loss": 0.005915559828281403,
          "validation_loss": 0.004562043119221926,
          "train_metric": 0.0005957934772595763,
          "validation_metric": 0.000459472241345793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7917326200004027
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0032027412289608037,
          "train_loss": 0.010913432575762272,
          "validation_loss": 0.008652782998979092,
          "train_metric": 0.001099160872399807,
          "validation_metric": 0.0008714766008779407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.808210789000441
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.02447867696173489,
          "train_loss": 0.017465919256210327,
          "validation_loss": 0.013125299476087093,
          "train_metric": 0.0017591032665222883,
          "validation_metric": 0.0013219323009252548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.825133262000236
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.011300355603452772,
          "train_loss": 0.0068909963592886925,
          "validation_loss": 0.006294074468314648,
          "train_metric": 0.0006940358434803784,
          "validation_metric": 0.0006339161773212254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.842010714000935
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.006982446255278774,
          "train_loss": 0.0040626828558743,
          "validation_loss": 0.004192215856164694,
          "train_metric": 0.00040917861042544246,
          "validation_metric": 0.000422224635258317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.858235247000266
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.004824272506084526,
          "train_loss": 0.0029823530931025743,
          "validation_loss": 0.002031839918345213,
          "train_metric": 0.0003003716701641679,
          "validation_metric": 0.00020463948021642864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.874675412000215
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0025836627173703164,
          "train_loss": 0.005504194181412458,
          "validation_loss": 0.0036804487463086843,
          "train_metric": 0.0005543622537516057,
          "validation_metric": 0.0003706812858581543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.890993769000488
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0038910616021894384,
          "train_loss": 0.0026339245960116386,
          "validation_loss": 0.0017489294987171888,
          "train_metric": 0.0002652792609296739,
          "validation_metric": 0.0001761457824613899,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.907355450999603
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0035942450595030095,
          "train_loss": 0.0020491841714829206,
          "validation_loss": 0.0017727033700793982,
          "train_metric": 0.00020638633577618748,
          "validation_metric": 0.00017854018369689584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.923686137999539
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.004442299985385034,
          "train_loss": 0.010838889516890049,
          "validation_loss": 0.009693920612335205,
          "train_metric": 0.0010916531318798661,
          "validation_metric": 0.0009763361304067075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.93998973700036
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.00376727672846755,
          "train_loss": 0.001998581225052476,
          "validation_loss": 0.0014798587653785944,
          "train_metric": 0.0002012897894019261,
          "validation_metric": 0.00014904596901033074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9564401070001622
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.005119179897519643,
          "train_loss": 0.0031030368991196156,
          "validation_loss": 0.0029733593109995127,
          "train_metric": 0.00031252653570845723,
          "validation_metric": 0.0002994659007526934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.974923007999678
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0033261494336329633,
          "train_loss": 0.0021288362331688404,
          "validation_loss": 0.0015360371908172965,
          "train_metric": 0.00021440861746668816,
          "validation_metric": 0.0001547040301375091,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.992350102000273
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.002251818545119022,
          "train_loss": 0.004207191057503223,
          "validation_loss": 0.003195105353370309,
          "train_metric": 0.00042373285396024585,
          "validation_metric": 0.00032179930713027716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.010410070000944
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.004258597120497143,
          "train_loss": 0.004527298733592033,
          "validation_loss": 0.0048430124297738075,
          "train_metric": 0.0004559729131869972,
          "validation_metric": 0.0004877704777754843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0402198859992495
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.005712863072403707,
          "train_loss": 0.003031670581549406,
          "validation_loss": 0.003113386919721961,
          "train_metric": 0.0003053387627005577,
          "validation_metric": 0.00031356894760392606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.085403115999725
        }
      ],
      "best": {
        "epoch": 121,
        "updates": 1936
      },
      "best_validation_metric": 0.00011219199222978204,
      "config": {
        "label": "ReLU",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010592072140510658
    },
    "a7bb18d2c173_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.583145534619689,
          "train_loss": 0.23918689787387848,
          "validation_loss": 0.1588190495967865,
          "train_metric": 0.024090027436614037,
          "validation_metric": 0.015995673835277557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.013954624999314547
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.21619082754477859,
          "train_loss": 0.15001872181892395,
          "validation_loss": 0.12154883146286011,
          "train_metric": 0.015109335072338581,
          "validation_metric": 0.012241953052580357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.028711767999993754
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.14033567206934094,
          "train_loss": 0.11539312452077866,
          "validation_loss": 0.09925363957881927,
          "train_metric": 0.011621971614658833,
          "validation_metric": 0.009996462613344193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04145810100089875
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.12017507012933493,
          "train_loss": 0.09647403657436371,
          "validation_loss": 0.08120380342006683,
          "train_metric": 0.009716511704027653,
          "validation_metric": 0.008178549818694592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05485182200027339
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.10254045785404742,
          "train_loss": 0.08029206097126007,
          "validation_loss": 0.06496105343103409,
          "train_metric": 0.00808672234416008,
          "validation_metric": 0.006542638875544071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14564567200068268
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.08901933906599879,
          "train_loss": 0.06823525577783585,
          "validation_loss": 0.055345870554447174,
          "train_metric": 0.006872405298054218,
          "validation_metric": 0.005574233829975128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15889567300109775
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.0808242631610483,
          "train_loss": 0.06248796358704567,
          "validation_loss": 0.050468526780605316,
          "train_metric": 0.006293558981269598,
          "validation_metric": 0.005083005875349045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17831483500049217
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.07092826254665852,
          "train_loss": 0.05796564742922783,
          "validation_loss": 0.04778416082262993,
          "train_metric": 0.005838087759912014,
          "validation_metric": 0.004812645725905895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2175905590011098
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.06893360381945968,
          "train_loss": 0.056175123900175095,
          "validation_loss": 0.04852449893951416,
          "train_metric": 0.005657752975821495,
          "validation_metric": 0.004887210205197334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2646419100001367
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.06747353286482394,
          "train_loss": 0.05182138830423355,
          "validation_loss": 0.04270927608013153,
          "train_metric": 0.005219260696321726,
          "validation_metric": 0.004301522392779589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28534347599998
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.06099209189414978,
          "train_loss": 0.048597075045108795,
          "validation_loss": 0.04020797088742256,
          "train_metric": 0.004894518759101629,
          "validation_metric": 0.004049600102007389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34776477400009753
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.05993118160404265,
          "train_loss": 0.04428774490952492,
          "validation_loss": 0.03736518323421478,
          "train_metric": 0.004460499621927738,
          "validation_metric": 0.003763284534215927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3689198699994449
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.055952056078240275,
          "train_loss": 0.04572293907403946,
          "validation_loss": 0.039240289479494095,
          "train_metric": 0.0046050469391047955,
          "validation_metric": 0.003952138125896454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40227911900001345
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.05123729654587805,
          "train_loss": 0.040401849895715714,
          "validation_loss": 0.0353718176484108,
          "train_metric": 0.004069126211106777,
          "validation_metric": 0.0035625193268060684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43668522000007215
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.05247232725378126,
          "train_loss": 0.03608043119311333,
          "validation_loss": 0.030558360740542412,
          "train_metric": 0.003633888904005289,
          "validation_metric": 0.0030777263455092907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45644438700037426
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.049872543197125196,
          "train_loss": 0.03516532480716705,
          "validation_loss": 0.03041493147611618,
          "train_metric": 0.0035417224280536175,
          "validation_metric": 0.0030632803682237864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.475094782999804
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.05155374598689377,
          "train_loss": 0.03128402680158615,
          "validation_loss": 0.026211518794298172,
          "train_metric": 0.0031508125830441713,
          "validation_metric": 0.00263992790132761,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49398511400067946
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.046620334847830236,
          "train_loss": 0.03074883297085762,
          "validation_loss": 0.02369905449450016,
          "train_metric": 0.0030969101935625076,
          "validation_metric": 0.0023868815042078495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5131106049993832
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.0443161993753165,
          "train_loss": 0.032443102449178696,
          "validation_loss": 0.027127718552947044,
          "train_metric": 0.003267550142481923,
          "validation_metric": 0.0027322042733430862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5315593459999945
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.03877060819650069,
          "train_loss": 0.025923391804099083,
          "validation_loss": 0.02276069112122059,
          "train_metric": 0.0026109092868864536,
          "validation_metric": 0.002292373450472951,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5514828510004008
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.04536195588298142,
          "train_loss": 0.02591816335916519,
          "validation_loss": 0.02195199392735958,
          "train_metric": 0.002610382391139865,
          "validation_metric": 0.0022109246347099543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5713596810001036
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.036643749102950096,
          "train_loss": 0.025550682097673416,
          "validation_loss": 0.022591307759284973,
          "train_metric": 0.0025733711663633585,
          "validation_metric": 0.0022753137163817883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5903014630002872
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.036209376878105104,
          "train_loss": 0.021915432065725327,
          "validation_loss": 0.018473803997039795,
          "train_metric": 0.002207241952419281,
          "validation_metric": 0.0018606138182803988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6096817350007768
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.036819413769990206,
          "train_loss": 0.026123736053705215,
          "validation_loss": 0.022974416613578796,
          "train_metric": 0.002631087088957429,
          "validation_metric": 0.0023138991091400385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6278907430005347
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.04139749030582607,
          "train_loss": 0.020830700173974037,
          "validation_loss": 0.017089635133743286,
          "train_metric": 0.0020979917608201504,
          "validation_metric": 0.0017212056554853916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6475414369997452
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.03845822799485177,
          "train_loss": 0.01982652023434639,
          "validation_loss": 0.015363338403403759,
          "train_metric": 0.001996854320168495,
          "validation_metric": 0.0015473393723368645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6673763270009658
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.033596856403164566,
          "train_loss": 0.019700221717357635,
          "validation_loss": 0.0158709567040205,
          "train_metric": 0.001984134316444397,
          "validation_metric": 0.0015984646743163466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6856708400009666
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.03133945353329182,
          "train_loss": 0.018460363149642944,
          "validation_loss": 0.014883436262607574,
          "train_metric": 0.0018592601409181952,
          "validation_metric": 0.0014990052441135049,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7041672020004626
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.03546860767528415,
          "train_loss": 0.018470128998160362,
          "validation_loss": 0.015985602512955666,
          "train_metric": 0.0018602438503876328,
          "validation_metric": 0.0016100113280117512,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7227105519996257
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.03289047162979841,
          "train_loss": 0.025497883558273315,
          "validation_loss": 0.02120673656463623,
          "train_metric": 0.0025680535472929478,
          "validation_metric": 0.002135864458978176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7411373909999384
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.03823907161131501,
          "train_loss": 0.02366740256547928,
          "validation_loss": 0.017989380285143852,
          "train_metric": 0.002383694052696228,
          "validation_metric": 0.001811824506148696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7592021670006943
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.03255100181559101,
          "train_loss": 0.016902409493923187,
          "validation_loss": 0.013053094036877155,
          "train_metric": 0.0017023488180711865,
          "validation_metric": 0.0013146597193554044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7775904900008754
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.031171356444247067,
          "train_loss": 0.016454454511404037,
          "validation_loss": 0.012593857944011688,
          "train_metric": 0.0016572322929278016,
          "validation_metric": 0.001268407329916954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7967156600007002
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.030643924605101347,
          "train_loss": 0.01702081970870495,
          "validation_loss": 0.014147600159049034,
          "train_metric": 0.0017142745200544596,
          "validation_metric": 0.0014248943189159036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8147964899999351
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.029848567326553166,
          "train_loss": 0.017430180683732033,
          "validation_loss": 0.012624048627912998,
          "train_metric": 0.0017555039376020432,
          "validation_metric": 0.00127144786529243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8328831550006726
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.027194515627343208,
          "train_loss": 0.013445649296045303,
          "validation_loss": 0.01169041357934475,
          "train_metric": 0.0013541965745389462,
          "validation_metric": 0.0011774157173931599,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8516243569993094
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.024144760565832257,
          "train_loss": 0.014662371948361397,
          "validation_loss": 0.012624217197299004,
          "train_metric": 0.0014767402317374945,
          "validation_metric": 0.0012714650947600603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8699567990006472
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.02960393321700394,
          "train_loss": 0.014535856433212757,
          "validation_loss": 0.010947571136057377,
          "train_metric": 0.0014639983419328928,
          "validation_metric": 0.0011025994317606091,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8895407660002093
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.028562736231833696,
          "train_loss": 0.012849782593548298,
          "validation_loss": 0.01077282801270485,
          "train_metric": 0.0012941830791532993,
          "validation_metric": 0.0010849997634068131,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.909096295999916
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.025025217793881893,
          "train_loss": 0.013503411784768105,
          "validation_loss": 0.009816810488700867,
          "train_metric": 0.00136001396458596,
          "validation_metric": 0.0009887132328003645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9284375910010567
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.03105824935482815,
          "train_loss": 0.015044720843434334,
          "validation_loss": 0.011305529624223709,
          "train_metric": 0.0015152491396293044,
          "validation_metric": 0.0011386515107005835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9470092579995253
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.025593703670892864,
          "train_loss": 0.01842604950070381,
          "validation_loss": 0.0122760608792305,
          "train_metric": 0.0018558043520897627,
          "validation_metric": 0.0012364001013338566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9651464489998034
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.02602449746336788,
          "train_loss": 0.014046718366444111,
          "validation_loss": 0.012496375478804111,
          "train_metric": 0.0014147341717034578,
          "validation_metric": 0.0012585893273353577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9834382099998038
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.024754128884524107,
          "train_loss": 0.012098665349185467,
          "validation_loss": 0.009162173606455326,
          "train_metric": 0.0012185333762317896,
          "validation_metric": 0.0009227804839611053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0024862900008884
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.024605963204521686,
          "train_loss": 0.012127991765737534,
          "validation_loss": 0.009645994752645493,
          "train_metric": 0.001221486832946539,
          "validation_metric": 0.0009715090855024755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0211734879994765
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.02541693253442645,
          "train_loss": 0.011188000440597534,
          "validation_loss": 0.009703710675239563,
          "train_metric": 0.001126814284361899,
          "validation_metric": 0.0009773223428055644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0397122300000774
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.023123145452700555,
          "train_loss": 0.010296531952917576,
          "validation_loss": 0.008516470901668072,
          "train_metric": 0.0010370289674028754,
          "validation_metric": 0.0008577477419748902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0592707119994884
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.023128371918573976,
          "train_loss": 0.012265319935977459,
          "validation_loss": 0.01024721935391426,
          "train_metric": 0.0012353180209174752,
          "validation_metric": 0.0010320624569430947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.077897209999719
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.024912253487855196,
          "train_loss": 0.00975191593170166,
          "validation_loss": 0.00805110763758421,
          "train_metric": 0.000982177327387035,
          "validation_metric": 0.0008108781185001135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0977164230007475
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.021633679571095854,
          "train_loss": 0.009421179071068764,
          "validation_loss": 0.0073700337670743465,
          "train_metric": 0.0009488666546531022,
          "validation_metric": 0.0007422828348353505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1172858250010904
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.020809931156691164,
          "train_loss": 0.008879273198544979,
          "validation_loss": 0.007537367753684521,
          "train_metric": 0.000894287833943963,
          "validation_metric": 0.0007591360481455922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1357413179994182
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.024204535642638803,
          "train_loss": 0.009218823164701462,
          "validation_loss": 0.008263557218015194,
          "train_metric": 0.0009284861152991652,
          "validation_metric": 0.0008322752546519041,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1542642639997212
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.023089694615919143,
          "train_loss": 0.009133871644735336,
          "validation_loss": 0.007681307382881641,
          "train_metric": 0.0009199301712214947,
          "validation_metric": 0.0007736330735497177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1723195850008779
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.021587878407444805,
          "train_loss": 0.008793112821877003,
          "validation_loss": 0.00699585722759366,
          "train_metric": 0.0008856100612320006,
          "validation_metric": 0.0007045970996841788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1919472139998106
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.02073128125630319,
          "train_loss": 0.010980757884681225,
          "validation_loss": 0.007897197268903255,
          "train_metric": 0.0011059417156502604,
          "validation_metric": 0.0007953767199069262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.211001137000494
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.019971489207819104,
          "train_loss": 0.0096426447853446,
          "validation_loss": 0.007423128932714462,
          "train_metric": 0.0009711718303151429,
          "validation_metric": 0.0007476303726434708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.229089524000301
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.021085271320771426,
          "train_loss": 0.009779761545360088,
          "validation_loss": 0.007093274500221014,
          "train_metric": 0.000984981656074524,
          "validation_metric": 0.0007144086412154138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2470167869996658
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.018940960289910436,
          "train_loss": 0.00891863089054823,
          "validation_loss": 0.006573219317942858,
          "train_metric": 0.0008982518338598311,
          "validation_metric": 0.0006620305357500911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2656934360002197
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.021462410571984947,
          "train_loss": 0.008511198684573174,
          "validation_loss": 0.006389295682311058,
          "train_metric": 0.0008572165970690548,
          "validation_metric": 0.0006435065297409892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.284939392000524
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.019898533762898296,
          "train_loss": 0.011134332045912743,
          "validation_loss": 0.008830059319734573,
          "train_metric": 0.0011214092373847961,
          "validation_metric": 0.0008893311605788767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3035264200007077
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.02153321827063337,
          "train_loss": 0.010999622754752636,
          "validation_loss": 0.0076390099711716175,
          "train_metric": 0.0011078417301177979,
          "validation_metric": 0.0007693731458857656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3223523010001372
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.02004678745288402,
          "train_loss": 0.009530022740364075,
          "validation_loss": 0.007875119335949421,
          "train_metric": 0.0009598288452252746,
          "validation_metric": 0.0007931531872600317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3413527600005182
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.020416632760316133,
          "train_loss": 0.00872294045984745,
          "validation_loss": 0.00697387894615531,
          "train_metric": 0.0008785425452515483,
          "validation_metric": 0.0007023835787549615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3609671230005915
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.018564796540886164,
          "train_loss": 0.008890722878277302,
          "validation_loss": 0.006263816729187965,
          "train_metric": 0.0008954409859143198,
          "validation_metric": 0.0006308686570264399,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.381230852999579
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.022153248020913452,
          "train_loss": 0.009469646029174328,
          "validation_loss": 0.007023356389254332,
          "train_metric": 0.0009537481237202883,
          "validation_metric": 0.0007073668530210853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.400629026000388
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.01995422347681597,
          "train_loss": 0.007725271862000227,
          "validation_loss": 0.005780433304607868,
          "train_metric": 0.0007780611049383879,
          "validation_metric": 0.0005821841768920422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4206387089998316
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.019805845629889518,
          "train_loss": 0.009953683242201805,
          "validation_loss": 0.008227323181927204,
          "train_metric": 0.001002498553134501,
          "validation_metric": 0.0008286259253509343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4394652379996842
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.022364881122484803,
          "train_loss": 0.010926750488579273,
          "validation_loss": 0.008509433828294277,
          "train_metric": 0.0011005024425685406,
          "validation_metric": 0.0008570390637032688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4579739700002392
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.020397144195158035,
          "train_loss": 0.007965474389493465,
          "validation_loss": 0.0073965503834187984,
          "train_metric": 0.0008022533729672432,
          "validation_metric": 0.0007449535187333822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4764309260008304
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.017519804125186056,
          "train_loss": 0.007023228332400322,
          "validation_loss": 0.006384979467839003,
          "train_metric": 0.0007073538145050406,
          "validation_metric": 0.0006430717185139656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4951363310010493
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.01796493277652189,
          "train_loss": 0.007754918187856674,
          "validation_loss": 0.00590237881988287,
          "train_metric": 0.0007810470415279269,
          "validation_metric": 0.0005944659933447838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5134539150003548
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.019344422849826515,
          "train_loss": 0.00723143108189106,
          "validation_loss": 0.005709747318178415,
          "train_metric": 0.0007283232989721,
          "validation_metric": 0.0005750648560933769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.532787759000712
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.01949010050157085,
          "train_loss": 0.006534535903483629,
          "validation_loss": 0.005009078420698643,
          "train_metric": 0.00065813452238217,
          "validation_metric": 0.0005044960416853428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5527884910006833
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.020220360078383237,
          "train_loss": 0.007325645070523024,
          "validation_loss": 0.004796404391527176,
          "train_metric": 0.0007378121372312307,
          "validation_metric": 0.0004830762918572873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5730162500003644
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.018277368100825697,
          "train_loss": 0.009103439748287201,
          "validation_loss": 0.006984527688473463,
          "train_metric": 0.0009168651886284351,
          "validation_metric": 0.0007034559966996312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.697398341000735
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.018829548615030944,
          "train_loss": 0.008593647740781307,
          "validation_loss": 0.006106977351009846,
          "train_metric": 0.0008655207930132747,
          "validation_metric": 0.0006150724366307259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7167043259996717
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.018984148395247757,
          "train_loss": 0.006331007461994886,
          "validation_loss": 0.005222175270318985,
          "train_metric": 0.0006376358214765787,
          "validation_metric": 0.0005259583704173565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7358507129993086
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.017842875968199223,
          "train_loss": 0.00789596326649189,
          "validation_loss": 0.00568165211006999,
          "train_metric": 0.0007952525047585368,
          "validation_metric": 0.0005722351488657296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7551121910000802
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.017470664344727993,
          "train_loss": 0.01552698016166687,
          "validation_loss": 0.011428066529333591,
          "train_metric": 0.0015638206386938691,
          "validation_metric": 0.0011509931646287441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7744275179993565
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.019846891111228615,
          "train_loss": 0.009262616746127605,
          "validation_loss": 0.006086915265768766,
          "train_metric": 0.000932896975427866,
          "validation_metric": 0.0006130518158897758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7925114080007916
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.021518419089261442,
          "train_loss": 0.00785287469625473,
          "validation_loss": 0.0047897109761834145,
          "train_metric": 0.0007909128325991333,
          "validation_metric": 0.00048240215983241796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8117832019997877
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.018924006493762136,
          "train_loss": 0.006804631091654301,
          "validation_loss": 0.00503664743155241,
          "train_metric": 0.0006853376398794353,
          "validation_metric": 0.000507272663526237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8345511139996233
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.018907354853581637,
          "train_loss": 0.006405750755220652,
          "validation_loss": 0.004428555257618427,
          "train_metric": 0.0006451637018471956,
          "validation_metric": 0.0004460277850739658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8544502400000056
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.01722480240277946,
          "train_loss": 0.007460963912308216,
          "validation_loss": 0.006361308507621288,
          "train_metric": 0.0007514409371651709,
          "validation_metric": 0.0006406877073459327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8729445980006858
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.017365486593917012,
          "train_loss": 0.0053807892836630344,
          "validation_loss": 0.004090455826371908,
          "train_metric": 0.0005419334047473967,
          "validation_metric": 0.0004119757213629782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8921493819998432
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.020795975753571838,
          "train_loss": 0.007963091135025024,
          "validation_loss": 0.005869241431355476,
          "train_metric": 0.000802013382781297,
          "validation_metric": 0.0005911285406909883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9104743780007993
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.019364648673217744,
          "train_loss": 0.006685366854071617,
          "validation_loss": 0.004605391528457403,
          "train_metric": 0.0006733257905580103,
          "validation_metric": 0.00046383822336792946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9286347890010802
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.015265618392731994,
          "train_loss": 0.005820934660732746,
          "validation_loss": 0.0042185112833976746,
          "train_metric": 0.0005862632533535361,
          "validation_metric": 0.0004248730547260493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9475855590007995
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.017723861790727824,
          "train_loss": 0.007101224735379219,
          "validation_loss": 0.006440713070333004,
          "train_metric": 0.0007152094622142613,
          "validation_metric": 0.0006486850907094777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9664293360001466
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.019524429633747786,
          "train_loss": 0.00924504641443491,
          "validation_loss": 0.007068206090480089,
          "train_metric": 0.0009311272297054529,
          "validation_metric": 0.0007118838839232922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9850645120004629
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.019296037906315178,
          "train_loss": 0.007251558359712362,
          "validation_loss": 0.005511088762432337,
          "train_metric": 0.0007303504389710724,
          "validation_metric": 0.0005550566711463034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0041374409993296
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.01788614405086264,
          "train_loss": 0.007384337019175291,
          "validation_loss": 0.004865778610110283,
          "train_metric": 0.000743723358027637,
          "validation_metric": 0.0004900634521618485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0233313660010026
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.018322292889934033,
          "train_loss": 0.004990776535123587,
          "validation_loss": 0.00401315139606595,
          "train_metric": 0.000502652779687196,
          "validation_metric": 0.00040418992284685373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0433486870006163
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.017734187189489603,
          "train_loss": 0.008819016627967358,
          "validation_loss": 0.0065688518807291985,
          "train_metric": 0.0008882189868018031,
          "validation_metric": 0.0006615907186642289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0617791330005275
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.018715678248554468,
          "train_loss": 0.008496689610183239,
          "validation_loss": 0.005275501403957605,
          "train_metric": 0.0008557554101571441,
          "validation_metric": 0.0005313291912898421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.079413556999498
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.016607561439741403,
          "train_loss": 0.006312018260359764,
          "validation_loss": 0.004349726717919111,
          "train_metric": 0.0006357232923619449,
          "validation_metric": 0.00043808852205984294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0974818329996197
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.015707819955423474,
          "train_loss": 0.005158456973731518,
          "validation_loss": 0.004023711662739515,
          "train_metric": 0.0005195409175939858,
          "validation_metric": 0.0004052535514347255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1151707049994
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.01645484659820795,
          "train_loss": 0.004884906113147736,
          "validation_loss": 0.0038570377510041,
          "train_metric": 0.0004919898929074407,
          "validation_metric": 0.0003884666948579252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.133341709000888
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.016547104518394917,
          "train_loss": 0.009299798868596554,
          "validation_loss": 0.00725331949070096,
          "train_metric": 0.000936641707085073,
          "validation_metric": 0.0007305277395062149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1518010439995123
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.016702577471733093,
          "train_loss": 0.012863186188042164,
          "validation_loss": 0.008407799527049065,
          "train_metric": 0.0012955330312252045,
          "validation_metric": 0.0008468027808703482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.170162996999352
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.019518282962962985,
          "train_loss": 0.006602454464882612,
          "validation_loss": 0.004160347394645214,
          "train_metric": 0.0006649750284850597,
          "validation_metric": 0.00041901497752405703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1880975770000077
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.016329762001987547,
          "train_loss": 0.006056387908756733,
          "validation_loss": 0.005156324245035648,
          "train_metric": 0.0006099772290326655,
          "validation_metric": 0.0005193260731175542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.206410090000645
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0160474261501804,
          "train_loss": 0.006563578732311726,
          "validation_loss": 0.006185609847307205,
          "train_metric": 0.0006610596901737154,
          "validation_metric": 0.0006229920545592904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2249171149996982
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.015809388365596533,
          "train_loss": 0.008158598095178604,
          "validation_loss": 0.0064994595013558865,
          "train_metric": 0.0008217041613534093,
          "validation_metric": 0.0006546017248183489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.243091739999727
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.014150380447972566,
          "train_loss": 0.0045024859718978405,
          "validation_loss": 0.003493482945486903,
          "train_metric": 0.00045347397099249065,
          "validation_metric": 0.0003518507583066821,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.262722104000204
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.014863786345813423,
          "train_loss": 0.0064453016966581345,
          "validation_loss": 0.0050814407877624035,
          "train_metric": 0.0006491472595371306,
          "validation_metric": 0.0005117840482853353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2813249099999666
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.016050366626586765,
          "train_loss": 0.005050681531429291,
          "validation_loss": 0.0036165751516819,
          "train_metric": 0.0005086861783638597,
          "validation_metric": 0.0003642481751739979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.299668333000227
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.015906329615972936,
          "train_loss": 0.004570840857923031,
          "validation_loss": 0.003379126777872443,
          "train_metric": 0.00046035830746404827,
          "validation_metric": 0.00034033323754556477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.319230556000548
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.016339238442014903,
          "train_loss": 0.004799462389200926,
          "validation_loss": 0.003311796812340617,
          "train_metric": 0.00048338426859118044,
          "validation_metric": 0.0003335519868414849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.338460066999687
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.01554487890098244,
          "train_loss": 0.011468933895230293,
          "validation_loss": 0.007348158862441778,
          "train_metric": 0.0011551090283319354,
          "validation_metric": 0.0007400795584544539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.357683034000729
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.01886324695078656,
          "train_loss": 0.006430373527109623,
          "validation_loss": 0.004062050022184849,
          "train_metric": 0.0006476436974480748,
          "validation_metric": 0.00040911478572525084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.376359363001029
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.01472752052359283,
          "train_loss": 0.004769343417137861,
          "validation_loss": 0.003549217013642192,
          "train_metric": 0.00048035080544650555,
          "validation_metric": 0.00035746413050219417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.394701626000824
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.01436617597937584,
          "train_loss": 0.005831939168274403,
          "validation_loss": 0.004365586210042238,
          "train_metric": 0.0005873715272173285,
          "validation_metric": 0.0004396858566906303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4135809740000695
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.015203275135718286,
          "train_loss": 0.006044384092092514,
          "validation_loss": 0.005489745177328587,
          "train_metric": 0.0006087681977078319,
          "validation_metric": 0.0005529070622287691,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4315060900007666
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.014911962673068047,
          "train_loss": 0.00418620090931654,
          "validation_loss": 0.003256772644817829,
          "train_metric": 0.0004216188390273601,
          "validation_metric": 0.00032801018096506596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.450533410999924
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.014822183700744063,
          "train_loss": 0.004003551788628101,
          "validation_loss": 0.0036102524027228355,
          "train_metric": 0.0004032230644952506,
          "validation_metric": 0.00036361144157126546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4692155570010073
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.01554122380912304,
          "train_loss": 0.004455531947314739,
          "validation_loss": 0.003624026896432042,
          "train_metric": 0.00044874486047774553,
          "validation_metric": 0.0003649987338576466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4927972219993535
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.013570793031249195,
          "train_loss": 0.004392179660499096,
          "validation_loss": 0.0041083721444010735,
          "train_metric": 0.00044236425310373306,
          "validation_metric": 0.00041378018795512617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5117848120007693
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.014021792158018798,
          "train_loss": 0.004463333170861006,
          "validation_loss": 0.0030872118659317493,
          "train_metric": 0.00044953057658858597,
          "validation_metric": 0.0003109326644334942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5310222360003536
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.01504700374789536,
          "train_loss": 0.00486497487872839,
          "validation_loss": 0.003278169082477689,
          "train_metric": 0.000489982427097857,
          "validation_metric": 0.00033016520319506526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5494911480000155
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.01478029030840844,
          "train_loss": 0.005267373286187649,
          "validation_loss": 0.003932798281311989,
          "train_metric": 0.000530510616954416,
          "validation_metric": 0.0003960970207117498,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5681812830007402
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.01653344085207209,
          "train_loss": 0.006532173603773117,
          "validation_loss": 0.005000823177397251,
          "train_metric": 0.0006578966276720166,
          "validation_metric": 0.0005036646034568548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.586411598000268
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.015848466951865703,
          "train_loss": 0.005199444945901632,
          "validation_loss": 0.004964206833392382,
          "train_metric": 0.0005236691213212907,
          "validation_metric": 0.0004999767406843603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.604161051000119
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.014319182315375656,
          "train_loss": 0.004615378566086292,
          "validation_loss": 0.0030909599736332893,
          "train_metric": 0.0004648440226446837,
          "validation_metric": 0.0003113101702183485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.622239921000073
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.01358671736670658,
          "train_loss": 0.004653864540159702,
          "validation_loss": 0.003004670375958085,
          "train_metric": 0.00046872018720023334,
          "validation_metric": 0.0003026193880941719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.641363103999538
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.015166949131526053,
          "train_loss": 0.004717507399618626,
          "validation_loss": 0.0032523947302252054,
          "train_metric": 0.0004751300730276853,
          "validation_metric": 0.0003275692870374769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.659837040000639
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.015379663469502702,
          "train_loss": 0.0039151436649262905,
          "validation_loss": 0.0027773103211075068,
          "train_metric": 0.0003943188930861652,
          "validation_metric": 0.0002797205524984747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6788748010003474
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.014406244968995452,
          "train_loss": 0.006932790391147137,
          "validation_loss": 0.004732001107186079,
          "train_metric": 0.0006982451886869967,
          "validation_metric": 0.00047658986295573413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.697369489000266
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.015166704601142555,
          "train_loss": 0.005133121274411678,
          "validation_loss": 0.00370183028280735,
          "train_metric": 0.0005169891519472003,
          "validation_metric": 0.0003728347655851394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7157006680008635
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.01820658316137269,
          "train_loss": 0.005512767471373081,
          "validation_loss": 0.004939240403473377,
          "train_metric": 0.000555225764401257,
          "validation_metric": 0.0004974622279405594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.734417976000259
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.015013813623227179,
          "train_loss": 0.00451656011864543,
          "validation_loss": 0.0037642414681613445,
          "train_metric": 0.00045489135663956404,
          "validation_metric": 0.0003791205526795238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7527976460005448
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.016328913799952716,
          "train_loss": 0.010108955204486847,
          "validation_loss": 0.00879727117717266,
          "train_metric": 0.0010181369725614786,
          "validation_metric": 0.0008860289817675948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7711802590001753
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.015173602441791445,
          "train_loss": 0.003738661063835025,
          "validation_loss": 0.003129070159047842,
          "train_metric": 0.00037654422339983284,
          "validation_metric": 0.0003151484706904739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7895906300000206
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.014281875686720014,
          "train_loss": 0.004130019806325436,
          "validation_loss": 0.003755978075787425,
          "train_metric": 0.00041596044320613146,
          "validation_metric": 0.0003782883286476135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8080724510000437
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.013381469005253166,
          "train_loss": 0.003421222325414419,
          "validation_loss": 0.0028652334585785866,
          "train_metric": 0.0003445730544626713,
          "validation_metric": 0.0002885757712647319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8273761109994666
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.017847806331701577,
          "train_loss": 0.015742478892207146,
          "validation_loss": 0.011049389839172363,
          "train_metric": 0.0015855248784646392,
          "validation_metric": 0.0011128541082143784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.845607815999756
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.018332528008613735,
          "train_loss": 0.004827683791518211,
          "validation_loss": 0.0030395176727324724,
          "train_metric": 0.0004862266650889069,
          "validation_metric": 0.00030612910632044077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.863708275001045
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.016283005941659212,
          "train_loss": 0.005018835887312889,
          "validation_loss": 0.005080660805106163,
          "train_metric": 0.0005054787034168839,
          "validation_metric": 0.000511705526150763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8814986439992936
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.015069816348841414,
          "train_loss": 0.008064697496592999,
          "validation_loss": 0.007236929144710302,
          "train_metric": 0.0008122467552311718,
          "validation_metric": 0.0007288770284503698,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8990843650008173
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.01559811970219016,
          "train_loss": 0.004605747293680906,
          "validation_loss": 0.004106815438717604,
          "train_metric": 0.00046387402107939124,
          "validation_metric": 0.0004136233765166253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.917228251999404
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.014190227433573455,
          "train_loss": 0.005646538455039263,
          "validation_loss": 0.004139223136007786,
          "train_metric": 0.0005686987424269319,
          "validation_metric": 0.0004168873419985175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9354402129993105
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.013470391335431486,
          "train_loss": 0.0033964747563004494,
          "validation_loss": 0.0029904996044933796,
          "train_metric": 0.0003420804569032043,
          "validation_metric": 0.0003011921653524041,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9533585579993087
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.012760994664859027,
          "train_loss": 0.006217855494469404,
          "validation_loss": 0.004171121399849653,
          "train_metric": 0.0006262396345846355,
          "validation_metric": 0.00042010011384263635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9716402850008308
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0141978845349513,
          "train_loss": 0.0056684925220906734,
          "validation_loss": 0.003397771157324314,
          "train_metric": 0.0005709098186343908,
          "validation_metric": 0.00034221107489429414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.99027159400066
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.01403747097356245,
          "train_loss": 0.003662599716335535,
          "validation_loss": 0.002740745199844241,
          "train_metric": 0.0003688835713546723,
          "validation_metric": 0.000276037841103971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.010627026000293
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.016365124145522714,
          "train_loss": 0.006897726096212864,
          "validation_loss": 0.005028407089412212,
          "train_metric": 0.0006947136716917157,
          "validation_metric": 0.0005064427969045937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.029624082999362
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.017011267016641796,
          "train_loss": 0.005068371072411537,
          "validation_loss": 0.002482428215444088,
          "train_metric": 0.0005104677402414382,
          "validation_metric": 0.00025002111215144396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0482512489998044
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.01617591019021347,
          "train_loss": 0.00468714814633131,
          "validation_loss": 0.003523468039929867,
          "train_metric": 0.0004720723954960704,
          "validation_metric": 0.0003548708336893469,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0664464740002586
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.014859966526273638,
          "train_loss": 0.009026868268847466,
          "validation_loss": 0.007507195696234703,
          "train_metric": 0.0009091530810110271,
          "validation_metric": 0.0007560972589999437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0851132629995845
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.01728635816834867,
          "train_loss": 0.008197237737476826,
          "validation_loss": 0.007255083881318569,
          "train_metric": 0.0008255958091467619,
          "validation_metric": 0.0007307055639103055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.103836976000821
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.01678591110976413,
          "train_loss": 0.007208440452814102,
          "validation_loss": 0.007604413200169802,
          "train_metric": 0.0007260077400133014,
          "validation_metric": 0.0007658886606805027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.122070982999503
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.01657929306384176,
          "train_loss": 0.004792307503521442,
          "validation_loss": 0.0034175398759543896,
          "train_metric": 0.0004826636577490717,
          "validation_metric": 0.00034420209703966975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.140468879000764
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.013522500958060846,
          "train_loss": 0.003583407262340188,
          "validation_loss": 0.002724085468798876,
          "train_metric": 0.0003609076957218349,
          "validation_metric": 0.00027435991796664894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1818576519999624
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.013969776860903949,
          "train_loss": 0.008781859651207924,
          "validation_loss": 0.006276392377912998,
          "train_metric": 0.0008844767580740154,
          "validation_metric": 0.0006321353721432388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.198518340999726
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.016274560359306633,
          "train_loss": 0.007317096460610628,
          "validation_loss": 0.0050555868074297905,
          "train_metric": 0.0007369511295109987,
          "validation_metric": 0.0005091801867820323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.212746237000829
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.014667844807263464,
          "train_loss": 0.006760227493941784,
          "validation_loss": 0.003893602406606078,
          "train_metric": 0.0006808653706684709,
          "validation_metric": 0.00039214937714859843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2258319230004417
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.015286057139746845,
          "train_loss": 0.005308606196194887,
          "validation_loss": 0.003120443318039179,
          "train_metric": 0.0005346633843146265,
          "validation_metric": 0.00031427963403984904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2392186090000905
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.015347961394581944,
          "train_loss": 0.003661593422293663,
          "validation_loss": 0.0024587465450167656,
          "train_metric": 0.0003687822609208524,
          "validation_metric": 0.0002476359368301928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2553191099996184
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.015216314524877816,
          "train_loss": 0.009320829063653946,
          "validation_loss": 0.00503311725333333,
          "train_metric": 0.0009387597674503922,
          "validation_metric": 0.0005069171893410385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2730837399994925
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.01685221679508686,
          "train_loss": 0.005401005037128925,
          "validation_loss": 0.004387157037854195,
          "train_metric": 0.0005439694505184889,
          "validation_metric": 0.00044185842853039503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2954803600005107
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.017023240856360644,
          "train_loss": 0.005767563823610544,
          "validation_loss": 0.004724212922155857,
          "train_metric": 0.0005808878922834992,
          "validation_metric": 0.0004758054274134338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3154023839997535
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.013986320060212165,
          "train_loss": 0.003507035318762064,
          "validation_loss": 0.0028365531470626593,
          "train_metric": 0.0003532157570589334,
          "validation_metric": 0.000285687274299562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3354220260007423
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.013713097956497222,
          "train_loss": 0.006116082426160574,
          "validation_loss": 0.005880878306925297,
          "train_metric": 0.0006159894983284175,
          "validation_metric": 0.000592300551943481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.354176537000967
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.013763209950411692,
          "train_loss": 0.006554191000759602,
          "validation_loss": 0.003948063123971224,
          "train_metric": 0.0006601141067221761,
          "validation_metric": 0.0003976344596594572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.373950249999325
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.014187205088092014,
          "train_loss": 0.004005005583167076,
          "validation_loss": 0.0035545083228498697,
          "train_metric": 0.00040336951497010887,
          "validation_metric": 0.0003579970798455179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.406951786000718
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.013178720138967037,
          "train_loss": 0.0033005434088408947,
          "validation_loss": 0.0023744727950543165,
          "train_metric": 0.00033241865457966924,
          "validation_metric": 0.00023914818302728236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.427335382999445
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.013080474338494241,
          "train_loss": 0.005015597678720951,
          "validation_loss": 0.004321936517953873,
          "train_metric": 0.0005051526823081076,
          "validation_metric": 0.00043528960668481886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4468467320002674
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.019564329471904784,
          "train_loss": 0.004260550253093243,
          "validation_loss": 0.0032932888716459274,
          "train_metric": 0.000429107021773234,
          "validation_metric": 0.0003316879738122225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4662703510002757
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.014721063955221325,
          "train_loss": 0.0070428973995149136,
          "validation_loss": 0.004805033095180988,
          "train_metric": 0.0007093349704518914,
          "validation_metric": 0.0004839453031308949,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4860899049999716
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.012813681998522952,
          "train_loss": 0.00403105141595006,
          "validation_loss": 0.002723976271227002,
          "train_metric": 0.00040599267231300473,
          "validation_metric": 0.00027434894582256675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.505585818000327
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.013679171534022316,
          "train_loss": 0.005523900501430035,
          "validation_loss": 0.0029866397380828857,
          "train_metric": 0.0005563470767810941,
          "validation_metric": 0.0003008034254889935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.52554124100061
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.017269466654397547,
          "train_loss": 0.004482210613787174,
          "validation_loss": 0.003582220058888197,
          "train_metric": 0.00045143187162466347,
          "validation_metric": 0.0003607880789786577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5442668739997316
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.017262926092371345,
          "train_loss": 0.007442153058946133,
          "validation_loss": 0.005790430121123791,
          "train_metric": 0.0007495463942177594,
          "validation_metric": 0.0005831909365952015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5627249869994557
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.014377463900018483,
          "train_loss": 0.003612499451264739,
          "validation_loss": 0.003139699576422572,
          "train_metric": 0.0003638376947492361,
          "validation_metric": 0.00031621899688616395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5840065980009967
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.011999283102340996,
          "train_loss": 0.009721983224153519,
          "validation_loss": 0.005502407439053059,
          "train_metric": 0.0009791625197976828,
          "validation_metric": 0.0005541823338717222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.604015843000525
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.016813860973343253,
          "train_loss": 0.005427281837910414,
          "validation_loss": 0.0027574002742767334,
          "train_metric": 0.0005466159200295806,
          "validation_metric": 0.00027771524037234485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6213326140004938
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.014808754116529599,
          "train_loss": 0.0047049508430063725,
          "validation_loss": 0.0038559671957045794,
          "train_metric": 0.00047386542428284883,
          "validation_metric": 0.000388358865166083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.639210315001037
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.017020532744936645,
          "train_loss": 0.004727690480649471,
          "validation_loss": 0.003585625207051635,
          "train_metric": 0.0004761556047014892,
          "validation_metric": 0.00036113103851675987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6570638539997162
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.013561472500441596,
          "train_loss": 0.006196514703333378,
          "validation_loss": 0.0035665917675942183,
          "train_metric": 0.0006240902002900839,
          "validation_metric": 0.0003592140565160662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.775485657999525
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.01594212168129161,
          "train_loss": 0.006670842412859201,
          "validation_loss": 0.0042118183337152,
          "train_metric": 0.0006718627992086112,
          "validation_metric": 0.00042419895180501044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.795399807000649
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.01336584787350148,
          "train_loss": 0.0037581250071525574,
          "validation_loss": 0.0025063662324100733,
          "train_metric": 0.0003785045992117375,
          "validation_metric": 0.00025243201525881886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.815422245999798
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.011844955763081089,
          "train_loss": 0.004454830661416054,
          "validation_loss": 0.0030992061365395784,
          "train_metric": 0.0004486741963773966,
          "validation_metric": 0.00031214067712426186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.833227672999783
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.011948689993005246,
          "train_loss": 0.005078175570815802,
          "validation_loss": 0.004085762891918421,
          "train_metric": 0.0005114552332088351,
          "validation_metric": 0.0004115031333640218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8513675009999133
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.013733949686866254,
          "train_loss": 0.0037166974507272243,
          "validation_loss": 0.003508351743221283,
          "train_metric": 0.00037433215766213834,
          "validation_metric": 0.00035334835411049426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.869828539000082
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.012123144464567304,
          "train_loss": 0.003575362730771303,
          "validation_loss": 0.0022231251932680607,
          "train_metric": 0.000360097415978089,
          "validation_metric": 0.00022390503727365285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8887455870008125
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.01203954528318718,
          "train_loss": 0.002836600411683321,
          "validation_loss": 0.0018278748029842973,
          "train_metric": 0.0002856920182239264,
          "validation_metric": 0.00018409684707876295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9099860999995144
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.012868679681560025,
          "train_loss": 0.0027817951049655676,
          "validation_loss": 0.0018883872544392943,
          "train_metric": 0.0002801722439471632,
          "validation_metric": 0.00019019146566279233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.929475283001011
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.013812091405270621,
          "train_loss": 0.006651216186583042,
          "validation_loss": 0.005138416774570942,
          "train_metric": 0.0006698862416669726,
          "validation_metric": 0.0005175225669518113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.94902431599985
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.014030404679942876,
          "train_loss": 0.01292503997683525,
          "validation_loss": 0.012199703603982925,
          "train_metric": 0.0013017627643421292,
          "validation_metric": 0.0012287094723433256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.968033746999936
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.01324994774768129,
          "train_loss": 0.005902387667447329,
          "validation_loss": 0.0041593643836677074,
          "train_metric": 0.0005944668082520366,
          "validation_metric": 0.00041891593718901277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.990163112001028
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.011906570522114635,
          "train_loss": 0.005088765639811754,
          "validation_loss": 0.0027650478295981884,
          "train_metric": 0.0005125218885950744,
          "validation_metric": 0.00027848550234921277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0108756919999
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.011573890689760447,
          "train_loss": 0.004068304784595966,
          "validation_loss": 0.003237440250813961,
          "train_metric": 0.00040974473813548684,
          "validation_metric": 0.00032606313470751047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.028872246999526
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.01391709892777726,
          "train_loss": 0.005142729263752699,
          "validation_loss": 0.003324722871184349,
          "train_metric": 0.0005179569125175476,
          "validation_metric": 0.000334853888489306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0471476849998
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.014285419310908765,
          "train_loss": 0.0030822111293673515,
          "validation_loss": 0.0021392935886979103,
          "train_metric": 0.00031042902264744043,
          "validation_metric": 0.00021546182688325644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.065460006000649
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.012354841834167019,
          "train_loss": 0.004936872515827417,
          "validation_loss": 0.003049879800528288,
          "train_metric": 0.000497223692946136,
          "validation_metric": 0.0003071727405767888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.083618234999449
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.012099976738682017,
          "train_loss": 0.003753621829673648,
          "validation_loss": 0.003259923541918397,
          "train_metric": 0.0003780511033255607,
          "validation_metric": 0.00032832755823619664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.115346561000479
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.01198767256573774,
          "train_loss": 0.0027990336529910564,
          "validation_loss": 0.0017283010529354215,
          "train_metric": 0.0002819084329530597,
          "validation_metric": 0.00017406814731657505,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.133614921000117
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0126503593637608,
          "train_loss": 0.003650236874818802,
          "validation_loss": 0.002096449723467231,
          "train_metric": 0.0003676384803839028,
          "validation_metric": 0.00021114676201250404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.151863778999541
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.01402321649948135,
          "train_loss": 0.0031855457928031683,
          "validation_loss": 0.003153553232550621,
          "train_metric": 0.0003208364942111075,
          "validation_metric": 0.0003176142927259207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.169922858000064
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.01374273025430739,
          "train_loss": 0.0046165413223207,
          "validation_loss": 0.004009690135717392,
          "train_metric": 0.00046496119466610253,
          "validation_metric": 0.0004038412880618125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.187945191000836
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0138651542365551,
          "train_loss": 0.003838270902633667,
          "validation_loss": 0.0020640785805881023,
          "train_metric": 0.000386576633900404,
          "validation_metric": 0.00020788641995750368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.207002508001096
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.011927421670407057,
          "train_loss": 0.005996730178594589,
          "validation_loss": 0.0037298756651580334,
          "train_metric": 0.0006039686850272119,
          "validation_metric": 0.0003756594378501177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.224523603999842
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.014056780200917274,
          "train_loss": 0.0030838323291391134,
          "validation_loss": 0.0027008515316993,
          "train_metric": 0.0003105922951363027,
          "validation_metric": 0.0002720198535826057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.244531632999497
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.012588274606969208,
          "train_loss": 0.002398206852376461,
          "validation_loss": 0.0015708499122411013,
          "train_metric": 0.0002415386406937614,
          "validation_metric": 0.0001582102559041232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2638848919996235
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.012554739747429267,
          "train_loss": 0.003582750679925084,
          "validation_loss": 0.002099828328937292,
          "train_metric": 0.0003608415136113763,
          "validation_metric": 0.00021148701489437371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.281605534000846
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.011919241194846109,
          "train_loss": 0.003487691283226013,
          "validation_loss": 0.0024979894515126944,
          "train_metric": 0.0003512675466481596,
          "validation_metric": 0.00025158835342153907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.298895342999458
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.014406894595595077,
          "train_loss": 0.003867136547341943,
          "validation_loss": 0.003289777087047696,
          "train_metric": 0.000389483873732388,
          "validation_metric": 0.0003313342749606818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.316224303000126
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.012175677518825978,
          "train_loss": 0.003522458951920271,
          "validation_loss": 0.0021692393347620964,
          "train_metric": 0.00035476917400956154,
          "validation_metric": 0.0002184778277296573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.334705637000297
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.012224267004057765,
          "train_loss": 0.0035040881484746933,
          "validation_loss": 0.0028383336029946804,
          "train_metric": 0.0003529188979882747,
          "validation_metric": 0.00028586655389517546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.353592900999502
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.013344166451133788,
          "train_loss": 0.005079132970422506,
          "validation_loss": 0.002867239760234952,
          "train_metric": 0.0005115516833029687,
          "validation_metric": 0.0002887778973672539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.371663812000406
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.013445501594105735,
          "train_loss": 0.0027196696028113365,
          "validation_loss": 0.002138423267751932,
          "train_metric": 0.00027391515322960913,
          "validation_metric": 0.00021537415159400553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.39000760899944
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.01234035927336663,
          "train_loss": 0.005136807914823294,
          "validation_loss": 0.004696645308285952,
          "train_metric": 0.0005173605168238282,
          "validation_metric": 0.00047302889288403094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.408077043000958
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.012484832142945379,
          "train_loss": 0.00345787825062871,
          "validation_loss": 0.0022137167397886515,
          "train_metric": 0.00034826481714844704,
          "validation_metric": 0.00022295740200206637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.430569148000359
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.012720920494757593,
          "train_loss": 0.002348412061110139,
          "validation_loss": 0.0014582201838493347,
          "train_metric": 0.00023652345407754183,
          "validation_metric": 0.00014686660142615438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.448947602000771
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.013836864527547732,
          "train_loss": 0.00605508778244257,
          "validation_loss": 0.00449731620028615,
          "train_metric": 0.0006098462035879493,
          "validation_metric": 0.0004529531579464674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.469581649000247
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.01677886926336214,
          "train_loss": 0.004661740269511938,
          "validation_loss": 0.004615498706698418,
          "train_metric": 0.0004695133538916707,
          "validation_metric": 0.0004648561589419842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.487653611000496
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.01459421296021901,
          "train_loss": 0.0030542060267180204,
          "validation_loss": 0.002399216638877988,
          "train_metric": 0.00030760845402255654,
          "validation_metric": 0.000241640314925462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.506267194999964
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.012346214789431542,
          "train_loss": 0.002554923528805375,
          "validation_loss": 0.0015689227730035782,
          "train_metric": 0.000257322535617277,
          "validation_metric": 0.00015801616245880723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.524689203000889
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.012591579288709909,
          "train_loss": 0.006053201388567686,
          "validation_loss": 0.0030463445000350475,
          "train_metric": 0.0006096563301980495,
          "validation_metric": 0.00030681665521115065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.5420442190006725
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.012684728717431426,
          "train_loss": 0.0045836688950657845,
          "validation_loss": 0.003441988956183195,
          "train_metric": 0.0004616504011210054,
          "validation_metric": 0.00034666454303078353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.559500179000679
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.013100089854560792,
          "train_loss": 0.004588410723954439,
          "validation_loss": 0.0022832024842500687,
          "train_metric": 0.0004621279367711395,
          "validation_metric": 0.0002299558254890144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.576919217000977
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.013789190037641674,
          "train_loss": 0.0037130878772586584,
          "validation_loss": 0.0019031036645174026,
          "train_metric": 0.0003739685926120728,
          "validation_metric": 0.0001916736364364624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.594059691000439
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.012500525102950633,
          "train_loss": 0.004241492599248886,
          "validation_loss": 0.002051782561466098,
          "train_metric": 0.0004271876241546124,
          "validation_metric": 0.00020664800831582397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.611189336999814
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.013992963125929236,
          "train_loss": 0.0063552409410476685,
          "validation_loss": 0.006097977515310049,
          "train_metric": 0.0006400765851140022,
          "validation_metric": 0.0006141660269349813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.628541721000147
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.01554252445930615,
          "train_loss": 0.0032601829152554274,
          "validation_loss": 0.002408958738669753,
          "train_metric": 0.0003283536934759468,
          "validation_metric": 0.00024262149236164987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.647178453000379
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.012838723021559417,
          "train_loss": 0.004280546214431524,
          "validation_loss": 0.0039282008074223995,
          "train_metric": 0.00043112094863317907,
          "validation_metric": 0.00039563403697684407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.664641343000767
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0120983068482019,
          "train_loss": 0.003879191353917122,
          "validation_loss": 0.00218571862205863,
          "train_metric": 0.00039069794001989067,
          "validation_metric": 0.00022013759007677436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.683623459000955
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.012249497638549656,
          "train_loss": 0.0036998195573687553,
          "validation_loss": 0.003556748852133751,
          "train_metric": 0.00037263226113282144,
          "validation_metric": 0.00035822272184304893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.70176566400005
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.015563486434984952,
          "train_loss": 0.004337805323302746,
          "validation_loss": 0.00233467691577971,
          "train_metric": 0.00043688793084584177,
          "validation_metric": 0.00023514007625635713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.7197902080006315
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.012281253410037607,
          "train_loss": 0.0034237143117934465,
          "validation_loss": 0.002490678569301963,
          "train_metric": 0.0003448239585850388,
          "validation_metric": 0.0002508519683033228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.737547913000526
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.013391501037403941,
          "train_loss": 0.005516992881894112,
          "validation_loss": 0.003747228067368269,
          "train_metric": 0.0005556513206101954,
          "validation_metric": 0.00037740706466138363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.755188516999624
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.014229591353796422,
          "train_loss": 0.004213783424347639,
          "validation_loss": 0.0029008272103965282,
          "train_metric": 0.00042439685785211623,
          "validation_metric": 0.0002921606646850705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.772495602001072
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.01224180415738374,
          "train_loss": 0.0028095347806811333,
          "validation_loss": 0.0023621278814971447,
          "train_metric": 0.0002829660661518574,
          "validation_metric": 0.00023790485283825547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.790033606999714
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.012450661335606128,
          "train_loss": 0.0032191842328757048,
          "validation_loss": 0.002725656609982252,
          "train_metric": 0.0003242244420107454,
          "validation_metric": 0.00027451818459667265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.808731684999657
        }
      ],
      "best": {
        "epoch": 214,
        "updates": 3424
      },
      "best_validation_metric": 0.00014686660142615438,
      "config": {
        "label": "Dropout 0.1",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.1,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012118853139887222
    },
    "be44f6a2021f_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.17469824547879398,
          "train_loss": 0.08774837106466293,
          "validation_loss": 0.086602583527565,
          "train_metric": 0.008837694302201271,
          "validation_metric": 0.008722295053303242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08210091299952182
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07360453333240002,
          "train_loss": 0.06181218847632408,
          "validation_loss": 0.06474826484918594,
          "train_metric": 0.0062254974618554115,
          "validation_metric": 0.006521208211779594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15877179999915825
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.056057537789456546,
          "train_loss": 0.04291826859116554,
          "validation_loss": 0.046800144016742706,
          "train_metric": 0.004322570748627186,
          "validation_metric": 0.004713539034128189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23569728099937493
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.038856737315654755,
          "train_loss": 0.03218131884932518,
          "validation_loss": 0.0270075686275959,
          "train_metric": 0.0032411853317171335,
          "validation_metric": 0.0027201033663004637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31732826000006753
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.02863936615176499,
          "train_loss": 0.023211568593978882,
          "validation_loss": 0.02074548229575157,
          "train_metric": 0.0023377842735499144,
          "validation_metric": 0.002089408691972494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40151119099937205
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.020648730802349746,
          "train_loss": 0.021016666665673256,
          "validation_loss": 0.016835160553455353,
          "train_metric": 0.0021167215891182423,
          "validation_metric": 0.0016955757746472955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5056030670002656
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.017860822612419724,
          "train_loss": 0.015319674275815487,
          "validation_loss": 0.016857700422406197,
          "train_metric": 0.0015429414343088865,
          "validation_metric": 0.0016978459898382425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5883056910006417
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.01622412708820775,
          "train_loss": 0.019327949732542038,
          "validation_loss": 0.01822485774755478,
          "train_metric": 0.0019466402009129524,
          "validation_metric": 0.0018355409847572446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.664979465000215
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.015191223297733814,
          "train_loss": 0.00894016120582819,
          "validation_loss": 0.0067612831480801105,
          "train_metric": 0.0009004203020595014,
          "validation_metric": 0.0006809717160649598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7883385660006752
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.01109477988211438,
          "train_loss": 0.008814625442028046,
          "validation_loss": 0.00746913580223918,
          "train_metric": 0.0008877768414095044,
          "validation_metric": 0.0007522639352828264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.877447253000355
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.018847736122552305,
          "train_loss": 0.02247755043208599,
          "validation_loss": 0.015438548289239407,
          "train_metric": 0.0022638565860688686,
          "validation_metric": 0.0015549140516668558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8930670270001428
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.018711894837906584,
          "train_loss": 0.012158442288637161,
          "validation_loss": 0.008228214457631111,
          "train_metric": 0.0012245536781847477,
          "validation_metric": 0.000828715565148741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9077765290003299
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.010454037867020816,
          "train_loss": 0.008276775479316711,
          "validation_loss": 0.005153849255293608,
          "train_metric": 0.000833606522064656,
          "validation_metric": 0.0005190768279135227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9319520680001006
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.008383314532693475,
          "train_loss": 0.009224709123373032,
          "validation_loss": 0.010711843147873878,
          "train_metric": 0.000929078902117908,
          "validation_metric": 0.0010788575746119022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9462138069993671
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.011629403757979162,
          "train_loss": 0.012009760364890099,
          "validation_loss": 0.009356722235679626,
          "train_metric": 0.001209579175338149,
          "validation_metric": 0.0009423746960237622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9599637320006877
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.01043891730660107,
          "train_loss": 0.006014615297317505,
          "validation_loss": 0.0039059207774698734,
          "train_metric": 0.0006057699793018401,
          "validation_metric": 0.0003933900152333081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9749835439997696
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.00892931772978045,
          "train_loss": 0.008275914937257767,
          "validation_loss": 0.0037234548944979906,
          "train_metric": 0.000833519792649895,
          "validation_metric": 0.0003750127216335386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9903763349993824
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.016686872142599896,
          "train_loss": 0.0080908527597785,
          "validation_loss": 0.008409524336457253,
          "train_metric": 0.0008148810593411326,
          "validation_metric": 0.000846976472530514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0048519520005357
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.007357539521763101,
          "train_loss": 0.0050854990258812904,
          "validation_loss": 0.004172328859567642,
          "train_metric": 0.0005121927824802697,
          "validation_metric": 0.0004202217096462846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0192711880008574
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.007062774151563644,
          "train_loss": 0.013032539747655392,
          "validation_loss": 0.010284306481480598,
          "train_metric": 0.0013125897385179996,
          "validation_metric": 0.0010357977589592338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0327825269996538
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.013464375690091401,
          "train_loss": 0.01278369128704071,
          "validation_loss": 0.010428732261061668,
          "train_metric": 0.0012875264510512352,
          "validation_metric": 0.0010503436205908656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0465266259998316
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.007588754087919369,
          "train_loss": 0.005317991133779287,
          "validation_loss": 0.00449811015278101,
          "train_metric": 0.0005356085603125393,
          "validation_metric": 0.00045303325168788433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0608520540008612
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.005772954718850087,
          "train_loss": 0.004529461730271578,
          "validation_loss": 0.00530380941927433,
          "train_metric": 0.0004561908426694572,
          "validation_metric": 0.0005341802607290447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.159180237000328
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.003951733167923521,
          "train_loss": 0.0032853365410119295,
          "validation_loss": 0.0028877228032797575,
          "train_metric": 0.0003308870655018836,
          "validation_metric": 0.0002908408350776881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1742553580006643
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.005694857194612268,
          "train_loss": 0.012830698862671852,
          "validation_loss": 0.011499912478029728,
          "train_metric": 0.001292260829359293,
          "validation_metric": 0.001158229075372219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1884957549991668
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.011381016200175509,
          "train_loss": 0.016322482377290726,
          "validation_loss": 0.01610724814236164,
          "train_metric": 0.001643940806388855,
          "validation_metric": 0.0016222628764808178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2022705229992425
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.008822661446174607,
          "train_loss": 0.008458982221782207,
          "validation_loss": 0.006830762606114149,
          "train_metric": 0.0008519577095285058,
          "validation_metric": 0.0006879693828523159,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2163349279999238
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.0055612625073990785,
          "train_loss": 0.003302248427644372,
          "validation_loss": 0.003444814821705222,
          "train_metric": 0.00033259036717936397,
          "validation_metric": 0.00034694914938881993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2296494359998178
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.007565966276160907,
          "train_loss": 0.016289830207824707,
          "validation_loss": 0.011221380904316902,
          "train_metric": 0.001640652073547244,
          "validation_metric": 0.0011301763588562608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2436951090003276
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.013868523805285804,
          "train_loss": 0.01236303336918354,
          "validation_loss": 0.007309946231544018,
          "train_metric": 0.0012451594229787588,
          "validation_metric": 0.0007362310425378382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.257524132999606
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008568004821427166,
          "train_loss": 0.007436853833496571,
          "validation_loss": 0.006095121614634991,
          "train_metric": 0.0007490128045901656,
          "validation_metric": 0.000613878364674747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.271090161999382
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.011385732956114225,
          "train_loss": 0.006194804795086384,
          "validation_loss": 0.007034803740680218,
          "train_metric": 0.0006239180220291018,
          "validation_metric": 0.0007085197139531374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2848712119994161
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.005738529376685619,
          "train_loss": 0.00688930694013834,
          "validation_loss": 0.007670516613870859,
          "train_metric": 0.0006938657606951892,
          "validation_metric": 0.0007725464529357851,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2987571520006895
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.005401998438173905,
          "train_loss": 0.00349409319460392,
          "validation_loss": 0.002726582810282707,
          "train_metric": 0.0003519122547004372,
          "validation_metric": 0.00027461143326945603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.313048894999156
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.005597987765213475,
          "train_loss": 0.005804476328194141,
          "validation_loss": 0.005776972509920597,
          "train_metric": 0.0005846054991707206,
          "validation_metric": 0.0005818354547955096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3268845239999791
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.006143338585388847,
          "train_loss": 0.01148177683353424,
          "validation_loss": 0.009007774293422699,
          "train_metric": 0.0011564027518033981,
          "validation_metric": 0.0009072301327250898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3403846180008259
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.009111446946917567,
          "train_loss": 0.009930150583386421,
          "validation_loss": 0.010570409707725048,
          "train_metric": 0.0010001285700127482,
          "validation_metric": 0.0010646129958331585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3549817179991805
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.02144408703316003,
          "train_loss": 0.01571727730333805,
          "validation_loss": 0.009488346055150032,
          "train_metric": 0.0015829865587875247,
          "validation_metric": 0.0009556313161738217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3688571260008757
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01252610661322251,
          "train_loss": 0.00984510499984026,
          "validation_loss": 0.010160131379961967,
          "train_metric": 0.000991562963463366,
          "validation_metric": 0.0010232911445200443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.382655531000637
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.00745236522925552,
          "train_loss": 0.005789437796920538,
          "validation_loss": 0.005320598836988211,
          "train_metric": 0.0005830909358337522,
          "validation_metric": 0.0005358712514862418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.396727039000325
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.004231854632962495,
          "train_loss": 0.004675413481891155,
          "validation_loss": 0.004513945430517197,
          "train_metric": 0.00047089048894122243,
          "validation_metric": 0.0004546281124930829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.410595003000708
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.007189390700659715,
          "train_loss": 0.004233115818351507,
          "validation_loss": 0.00276603433303535,
          "train_metric": 0.00042634387500584126,
          "validation_metric": 0.0002785848337225616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4245028910008841
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.007435428553435486,
          "train_loss": 0.005871464032679796,
          "validation_loss": 0.003420013003051281,
          "train_metric": 0.0005913523491472006,
          "validation_metric": 0.0003444512258283794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4386754150000343
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.009470527118537575,
          "train_loss": 0.004352807067334652,
          "validation_loss": 0.0026026819832623005,
          "train_metric": 0.00043839873978868127,
          "validation_metric": 0.0002621326129883528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4537958529999742
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.005118723045598017,
          "train_loss": 0.00682386988773942,
          "validation_loss": 0.0061017912812530994,
          "train_metric": 0.000687275140080601,
          "validation_metric": 0.0006145501392893493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4680756729994755
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.005086616001790389,
          "train_loss": 0.005078463815152645,
          "validation_loss": 0.004163194447755814,
          "train_metric": 0.0005114842788316309,
          "validation_metric": 0.00041930173756554723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.481814401999145
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0032088879524962977,
          "train_loss": 0.00489709060639143,
          "validation_loss": 0.0024873053189367056,
          "train_metric": 0.0004932170268148184,
          "validation_metric": 0.0002505122683942318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4959702310006833
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0029426619221339934,
          "train_loss": 0.0027947891503572464,
          "validation_loss": 0.002015306381508708,
          "train_metric": 0.00028148089768365026,
          "validation_metric": 0.00020297427545301616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5102210759996524
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.004175201407633722,
          "train_loss": 0.020366378128528595,
          "validation_loss": 0.015269857831299305,
          "train_metric": 0.0020512270275503397,
          "validation_metric": 0.0015379241667687893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.523805283000911
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.017829494070610963,
          "train_loss": 0.00337790884077549,
          "validation_loss": 0.0019126050174236298,
          "train_metric": 0.00034021062310785055,
          "validation_metric": 0.00019263055582996458,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5382682769995881
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007706464035436511,
          "train_loss": 0.004119453486055136,
          "validation_loss": 0.004866297822445631,
          "train_metric": 0.000414896261645481,
          "validation_metric": 0.0004901156644336879,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5656222170000547
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.003865843806124758,
          "train_loss": 0.0030894684605300426,
          "validation_loss": 0.0026276682037860155,
          "train_metric": 0.0003111599071417004,
          "validation_metric": 0.00026464913389645517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5956395840003097
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.003945239663153188,
          "train_loss": 0.0032585756853222847,
          "validation_loss": 0.0023356659803539515,
          "train_metric": 0.0003281918470747769,
          "validation_metric": 0.000235239727771841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6254731490007543
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.005527104272914585,
          "train_loss": 0.005947537254542112,
          "validation_loss": 0.004396939184516668,
          "train_metric": 0.0005990142817609012,
          "validation_metric": 0.00044284359319135547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6465115819992207
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.006053970202628989,
          "train_loss": 0.0036551710218191147,
          "validation_loss": 0.002675779862329364,
          "train_metric": 0.0003681354282889515,
          "validation_metric": 0.00026949477614834905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6863924389999738
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.003581159413442947,
          "train_loss": 0.0024149450473487377,
          "validation_loss": 0.0019068822730332613,
          "train_metric": 0.0002432243782095611,
          "validation_metric": 0.00019205421267542988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7201387290006096
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008579878223827109,
          "train_loss": 0.007476395927369595,
          "validation_loss": 0.008935678750276566,
          "train_metric": 0.0007529952563345432,
          "validation_metric": 0.0008999688434414566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.751813213000787
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.004014553069282556,
          "train_loss": 0.003338887356221676,
          "validation_loss": 0.0019609611481428146,
          "train_metric": 0.0003362804709468037,
          "validation_metric": 0.00019750080537050962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7853055720006523
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.005568284024775494,
          "train_loss": 0.009784785099327564,
          "validation_loss": 0.009299298748373985,
          "train_metric": 0.0009854877134785056,
          "validation_metric": 0.000936591241043061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8134372900003655
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.008419338599196635,
          "train_loss": 0.006721618585288525,
          "validation_loss": 0.004362899344414473,
          "train_metric": 0.0006769767496734858,
          "validation_metric": 0.00043941522017121315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8470911929998692
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004889800802629907,
          "train_loss": 0.0033947343472391367,
          "validation_loss": 0.0016815293347463012,
          "train_metric": 0.00034190522274002433,
          "validation_metric": 0.0001693574886303395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8810928349994356
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.005923350501689129,
          "train_loss": 0.012823661789298058,
          "validation_loss": 0.01178542710840702,
          "train_metric": 0.0012915523257106543,
          "validation_metric": 0.001186985056847334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9151150159996178
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.007817275763954967,
          "train_loss": 0.0061249337159097195,
          "validation_loss": 0.006626131013035774,
          "train_metric": 0.0006168808904476464,
          "validation_metric": 0.0006673596799373627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9488706459997047
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0035805948573397473,
          "train_loss": 0.002339717233553529,
          "validation_loss": 0.0020089170429855585,
          "train_metric": 0.00023564770526718348,
          "validation_metric": 0.00020233076065778732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.978958394000074
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0028934544207004365,
          "train_loss": 0.0026947136502712965,
          "validation_loss": 0.002496195025742054,
          "train_metric": 0.00027140165911987424,
          "validation_metric": 0.0002514075895305723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0132968369998707
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0026622053919709288,
          "train_loss": 0.002684344770386815,
          "validation_loss": 0.0018297166097909212,
          "train_metric": 0.00027035741368308663,
          "validation_metric": 0.00018428235489409417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.045167684000262
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0030751192825846374,
          "train_loss": 0.0015367658343166113,
          "validation_loss": 0.0007716359687037766,
          "train_metric": 0.00015477740089409053,
          "validation_metric": 7.771634409436956e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0809541750004428
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0028058795105607715,
          "train_loss": 0.0019020121544599533,
          "validation_loss": 0.0013743768213316798,
          "train_metric": 0.0001915636967169121,
          "validation_metric": 0.0001384222268825397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.100017508999372
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.002211427541624289,
          "train_loss": 0.0022421651519834995,
          "validation_loss": 0.0025325354654341936,
          "train_metric": 0.00022582264500670135,
          "validation_metric": 0.0002550677163526416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.11350294300064
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.004392391329020029,
          "train_loss": 0.011406540870666504,
          "validation_loss": 0.007262186612933874,
          "train_metric": 0.0011488250456750393,
          "validation_metric": 0.0007314208196476102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.127218363999418
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.011072731198510155,
          "train_loss": 0.003727383678779006,
          "validation_loss": 0.002349167363718152,
          "train_metric": 0.0003754084464162588,
          "validation_metric": 0.00023659953149035573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.140772356000525
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00408680804684991,
          "train_loss": 0.005674190353602171,
          "validation_loss": 0.004940127022564411,
          "train_metric": 0.0005714837461709976,
          "validation_metric": 0.0004975515184924006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.153996769000514
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.004902782951830886,
          "train_loss": 0.005039860028773546,
          "validation_loss": 0.004312317352741957,
          "train_metric": 0.0005075962981209159,
          "validation_metric": 0.00043432082748040557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1680068869991374
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0032235977851087227,
          "train_loss": 0.008413396775722504,
          "validation_loss": 0.005775026511400938,
          "train_metric": 0.0008473665802739561,
          "validation_metric": 0.0005816395278088748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.181352830999458
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00586096119513968,
          "train_loss": 0.002768790815025568,
          "validation_loss": 0.002736549824476242,
          "train_metric": 0.00027886248426511884,
          "validation_metric": 0.00027561525348573923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1944186920009088
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.003293814654171001,
          "train_loss": 0.004184024874120951,
          "validation_loss": 0.002594922436401248,
          "train_metric": 0.00042139962897635996,
          "validation_metric": 0.0002613510878290981,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2079499320007017
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.009261588580557145,
          "train_loss": 0.009118457324802876,
          "validation_loss": 0.008033856749534607,
          "train_metric": 0.0009183775982819498,
          "validation_metric": 0.0008091406198218465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2217334260003554
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.007043960300507024,
          "train_loss": 0.00428296672180295,
          "validation_loss": 0.003561281831935048,
          "train_metric": 0.00043136466410942376,
          "validation_metric": 0.0003586792736314237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.235349649999989
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0025835381929937284,
          "train_loss": 0.0018728274153545499,
          "validation_loss": 0.0012331916950643063,
          "train_metric": 0.00018862432625610381,
          "validation_metric": 0.0001242025609826669,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.249152767000851
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004261667090759147,
          "train_loss": 0.003506213193759322,
          "validation_loss": 0.003425085684284568,
          "train_metric": 0.0003531329275574535,
          "validation_metric": 0.000344962056260556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2627073229996313
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.003228697612939868,
          "train_loss": 0.003436910454183817,
          "validation_loss": 0.0023535508662462234,
          "train_metric": 0.00034615304321050644,
          "validation_metric": 0.0002370410511502996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.276277540000592
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0023477348040614743,
          "train_loss": 0.0028101303614676,
          "validation_loss": 0.0027611437253654003,
          "train_metric": 0.0002830260491464287,
          "validation_metric": 0.0002780923095997423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2903162540005724
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.002771884508547373,
          "train_loss": 0.002052157651633024,
          "validation_loss": 0.0021283598616719246,
          "train_metric": 0.00020668579963967204,
          "validation_metric": 0.000214360625250265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3039742260007188
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.002379728917730972,
          "train_loss": 0.001543528283946216,
          "validation_loss": 0.0016648912569507957,
          "train_metric": 0.0001554585323901847,
          "validation_metric": 0.00016768173372838646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.31789633399967
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0056973773025674745,
          "train_loss": 0.003502016654238105,
          "validation_loss": 0.004245120100677013,
          "train_metric": 0.0003527103108353913,
          "validation_metric": 0.0004275529063306749,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.331504092999239
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.0036285217574913986,
          "train_loss": 0.005470584612339735,
          "validation_loss": 0.0037674789782613516,
          "train_metric": 0.0005509773036465049,
          "validation_metric": 0.0003794466610997915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.344777078000334
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.004111100519367028,
          "train_loss": 0.0026503419503569603,
          "validation_loss": 0.0017753037391230464,
          "train_metric": 0.00026693270774558187,
          "validation_metric": 0.00017880207451526076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3584728030000406
        }
      ],
      "best": {
        "epoch": 67,
        "updates": 1072
      },
      "best_validation_metric": 7.771634409436956e-05,
      "config": {
        "label": "LayerNorm + tanh + constant LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.00881568738637944
    },
    "ff9112ae3885_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3273510686121881,
          "train_loss": 0.11705735325813293,
          "validation_loss": 0.07973263412714005,
          "train_metric": 0.011789587326347828,
          "validation_metric": 0.008030379191040993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02232273899971915
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.08486356446519494,
          "train_loss": 0.05732471123337746,
          "validation_loss": 0.07251747697591782,
          "train_metric": 0.005773534998297691,
          "validation_metric": 0.007303694728761911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.043542046998481965
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.04616492288187146,
          "train_loss": 0.032562948763370514,
          "validation_loss": 0.03916358947753906,
          "train_metric": 0.0032796207815408707,
          "validation_metric": 0.003944413270801306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06545960999937961
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.029842033341992646,
          "train_loss": 0.025069084018468857,
          "validation_loss": 0.03333912417292595,
          "train_metric": 0.002524866256862879,
          "validation_metric": 0.003357794601470232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08626969499891857
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.0214119884185493,
          "train_loss": 0.017583012580871582,
          "validation_loss": 0.024185124784708023,
          "train_metric": 0.0017708966042846441,
          "validation_metric": 0.0024358374066650867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10754056299992953
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.016405746253440157,
          "train_loss": 0.013314454816281796,
          "validation_loss": 0.017470011487603188,
          "train_metric": 0.0013409830862656236,
          "validation_metric": 0.0017595157260075212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1294854949992441
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.014689415314933285,
          "train_loss": 0.015650659799575806,
          "validation_loss": 0.026589423418045044,
          "train_metric": 0.001576277194544673,
          "validation_metric": 0.0026779891923069954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1499798799995915
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.01493760384619236,
          "train_loss": 0.018550381064414978,
          "validation_loss": 0.023763317614793777,
          "train_metric": 0.001868326566182077,
          "validation_metric": 0.002393354196101427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16936328900010267
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.015906014014035463,
          "train_loss": 0.008983106352388859,
          "validation_loss": 0.01192385982722044,
          "train_metric": 0.0009047455387189984,
          "validation_metric": 0.001200927421450615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19032182499904593
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.009969578837626614,
          "train_loss": 0.011814270168542862,
          "validation_loss": 0.011308618821203709,
          "train_metric": 0.0011898900847882032,
          "validation_metric": 0.0011389626888558269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2121531220000179
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.01114819009671919,
          "train_loss": 0.011215937323868275,
          "validation_loss": 0.015419055707752705,
          "train_metric": 0.0011296281591057777,
          "validation_metric": 0.0015529508236795664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2325075709995872
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.011344271610141732,
          "train_loss": 0.00578324357047677,
          "validation_loss": 0.007838734425604343,
          "train_metric": 0.0005824671825394034,
          "validation_metric": 0.0007894885493442416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2531369789994642
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.007560007594292983,
          "train_loss": 0.005091303959488869,
          "validation_loss": 0.009888500906527042,
          "train_metric": 0.0005127774784341455,
          "validation_metric": 0.0009959336603060365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2730279339994013
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.007257838544319384,
          "train_loss": 0.005637935362756252,
          "validation_loss": 0.008928872644901276,
          "train_metric": 0.0005678322049789131,
          "validation_metric": 0.0008992831571958959,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29304252399924735
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.00623394575814018,
          "train_loss": 0.012971029616892338,
          "validation_loss": 0.0078086103312671185,
          "train_metric": 0.0013063945807516575,
          "validation_metric": 0.000786454533226788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3152415109998401
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.013430104008875787,
          "train_loss": 0.007163480389863253,
          "validation_loss": 0.009570866823196411,
          "train_metric": 0.0007214794750325382,
          "validation_metric": 0.0009639424388296902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3359098169985373
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.008007690135855228,
          "train_loss": 0.006668552756309509,
          "validation_loss": 0.007338789291679859,
          "train_metric": 0.0006716322968713939,
          "validation_metric": 0.0007391360122710466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3566731079990859
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.007334137088037096,
          "train_loss": 0.011812794022262096,
          "validation_loss": 0.011058768257498741,
          "train_metric": 0.0011897413060069084,
          "validation_metric": 0.001113798702135682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3763331169993762
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.006115007839980535,
          "train_loss": 0.0056520020589232445,
          "validation_loss": 0.010433555580675602,
          "train_metric": 0.0005692489794455469,
          "validation_metric": 0.0010508294217288494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39689741099937237
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.004616073609213345,
          "train_loss": 0.003704607952386141,
          "validation_loss": 0.0050687710754573345,
          "train_metric": 0.00037311456981115043,
          "validation_metric": 0.0005105080199427903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41828166199957195
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.0033525712133268826,
          "train_loss": 0.005505973938852549,
          "validation_loss": 0.005003985017538071,
          "train_metric": 0.0005545414751395583,
          "validation_metric": 0.0005039830575697124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43974632899880817
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.0048169265355682,
          "train_loss": 0.0031576580367982388,
          "validation_loss": 0.004121873993426561,
          "train_metric": 0.0003180277126375586,
          "validation_metric": 0.00041514006443321705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.561838222998631
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.004369251444586553,
          "train_loss": 0.006998151540756226,
          "validation_loss": 0.005888920743018389,
          "train_metric": 0.0007048281840980053,
          "validation_metric": 0.000593110453337431,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5825556459985819
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.006612358818529174,
          "train_loss": 0.004325545392930508,
          "validation_loss": 0.006397956516593695,
          "train_metric": 0.0004356531717348844,
          "validation_metric": 0.0006443787715397775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6026288109987945
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.005372359715693165,
          "train_loss": 0.01272423006594181,
          "validation_loss": 0.004804378841072321,
          "train_metric": 0.0012815379304811358,
          "validation_metric": 0.00048387947026640177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.62231353199968
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.013801360037177801,
          "train_loss": 0.010914361104369164,
          "validation_loss": 0.009418616071343422,
          "train_metric": 0.0010992544703185558,
          "validation_metric": 0.0009486085036769509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6419695329987007
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.023850818717619404,
          "train_loss": 0.017270682379603386,
          "validation_loss": 0.013394763693213463,
          "train_metric": 0.001739440020173788,
          "validation_metric": 0.0013490712735801935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6621901879989309
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.011587520159082487,
          "train_loss": 0.006311653647571802,
          "validation_loss": 0.006933359429240227,
          "train_metric": 0.0006356866215355694,
          "validation_metric": 0.0006983026396483183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6822788529989339
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.007428047931171022,
          "train_loss": 0.008943111635744572,
          "validation_loss": 0.010714882984757423,
          "train_metric": 0.0009007173939608037,
          "validation_metric": 0.001079163863323629,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7026801579995663
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.008485349782858975,
          "train_loss": 0.004578865133225918,
          "validation_loss": 0.006117227487266064,
          "train_metric": 0.00046116652083583176,
          "validation_metric": 0.0006161048077046871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.722274412999468
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.005771150841610506,
          "train_loss": 0.008173983544111252,
          "validation_loss": 0.00479541951790452,
          "train_metric": 0.0008232537657022476,
          "validation_metric": 0.0004829770477954298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7421414139989793
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.007144150244130287,
          "train_loss": 0.012213042005896568,
          "validation_loss": 0.01698092184960842,
          "train_metric": 0.0012300527887418866,
          "validation_metric": 0.0017102562123909593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.762679126999501
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.005899935073102824,
          "train_loss": 0.0051960451528429985,
          "validation_loss": 0.003369505750015378,
          "train_metric": 0.0005233266274444759,
          "validation_metric": 0.0003393642255105078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7841801039994607
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.005917149548622547,
          "train_loss": 0.007233455311506987,
          "validation_loss": 0.007564662955701351,
          "train_metric": 0.000728527142200619,
          "validation_metric": 0.0007618850795552135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.804600075000053
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.004864623650064459,
          "train_loss": 0.0051497905515134335,
          "validation_loss": 0.00282531906850636,
          "train_metric": 0.0005186679773032665,
          "validation_metric": 0.0002845557755790651,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8256465019985626
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.00409684213809669,
          "train_loss": 0.006532116327434778,
          "validation_loss": 0.007584122009575367,
          "train_metric": 0.0006578908069059253,
          "validation_metric": 0.0007638449314981699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8452044339992426
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.009312566791777499,
          "train_loss": 0.005541311576962471,
          "validation_loss": 0.005547116044908762,
          "train_metric": 0.0005581005825661123,
          "validation_metric": 0.0005586851621046662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8648469469990232
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.017059682053513825,
          "train_loss": 0.008513408713042736,
          "validation_loss": 0.006189574953168631,
          "train_metric": 0.0008574392995797098,
          "validation_metric": 0.0006233913591131568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.884398282998518
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.0177556861017365,
          "train_loss": 0.019069751724600792,
          "validation_loss": 0.025645239278674126,
          "train_metric": 0.0019206355791538954,
          "validation_metric": 0.0025828946381807327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9060657439986244
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012423908628989011,
          "train_loss": 0.0077280704863369465,
          "validation_loss": 0.007089395076036453,
          "train_metric": 0.0007783429464325309,
          "validation_metric": 0.0007140179513953626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.925747640998452
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.0052849889907520264,
          "train_loss": 0.003457134822383523,
          "validation_loss": 0.0031268971506506205,
          "train_metric": 0.0003481899912003428,
          "validation_metric": 0.00031492963898926973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9461699329986004
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.003456670143350493,
          "train_loss": 0.0027233094442635775,
          "validation_loss": 0.0024531311355531216,
          "train_metric": 0.0002742817741818726,
          "validation_metric": 0.000247070420300588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9668697540000721
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.0057835834377328865,
          "train_loss": 0.004220916423946619,
          "validation_loss": 0.0033909312915056944,
          "train_metric": 0.0004251152276992798,
          "validation_metric": 0.0003415220999158919,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9870584419986699
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.004961211503541563,
          "train_loss": 0.0027626005467027426,
          "validation_loss": 0.0027505706530064344,
          "train_metric": 0.0002782390220090747,
          "validation_metric": 0.00027702737133949995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0066730749986164
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.004748126368212979,
          "train_loss": 0.00945703499019146,
          "validation_loss": 0.007011713460087776,
          "train_metric": 0.0009524778579361737,
          "validation_metric": 0.0007061940850690007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0260498529987672
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.004740721691632643,
          "train_loss": 0.0028950944542884827,
          "validation_loss": 0.0023027523420751095,
          "train_metric": 0.00029158330289646983,
          "validation_metric": 0.0002319247869309038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0468955669985007
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.002202396521170158,
          "train_loss": 0.001396849867887795,
          "validation_loss": 0.001685641473159194,
          "train_metric": 0.0001406856026733294,
          "validation_metric": 0.00016977165068965405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0682675579992065
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0018137458646378946,
          "train_loss": 0.0035003693774342537,
          "validation_loss": 0.002999718999490142,
          "train_metric": 0.0003525444190017879,
          "validation_metric": 0.0003021206648554653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0890990759999113
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00825509933201829,
          "train_loss": 0.017938241362571716,
          "validation_loss": 0.021720224991440773,
          "train_metric": 0.0018066740594804287,
          "validation_metric": 0.0021875817328691483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1100707719997445
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.01623914416995831,
          "train_loss": 0.010971436277031898,
          "validation_loss": 0.008731418289244175,
          "train_metric": 0.001105002942495048,
          "validation_metric": 0.0008793964516371489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1305326109995804
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007084169847075827,
          "train_loss": 0.004891781136393547,
          "validation_loss": 0.005322495009750128,
          "train_metric": 0.0004926822730340064,
          "validation_metric": 0.0005360622308216989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1510440959991683
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.004670059774070978,
          "train_loss": 0.0025661683175712824,
          "validation_loss": 0.004384555388242006,
          "train_metric": 0.0002584551111795008,
          "validation_metric": 0.0004415963776409626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1710314439987997
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.0026889196306001395,
          "train_loss": 0.001824769889935851,
          "validation_loss": 0.0024213145952671766,
          "train_metric": 0.00018378414097242057,
          "validation_metric": 0.0002438659139443189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1914562459987792
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.002969987181131728,
          "train_loss": 0.0016826242208480835,
          "validation_loss": 0.002400194527581334,
          "train_metric": 0.00016946776304394007,
          "validation_metric": 0.0002417388022877276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2120697840000503
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.002957138916826807,
          "train_loss": 0.0018714220495894551,
          "validation_loss": 0.001718224142678082,
          "train_metric": 0.00018848275067284703,
          "validation_metric": 0.00017305325309280306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2324509079990094
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.002361739312618738,
          "train_loss": 0.0021901200525462627,
          "validation_loss": 0.0017026650020852685,
          "train_metric": 0.0002205808414146304,
          "validation_metric": 0.0001714861864456907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2523802959985915
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.0040892575270845555,
          "train_loss": 0.004727373830974102,
          "validation_loss": 0.006496707908809185,
          "train_metric": 0.00047612376511096954,
          "validation_metric": 0.0006543245981447399,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2720253099996626
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.002474357552273432,
          "train_loss": 0.001369926263578236,
          "validation_loss": 0.0023739011958241463,
          "train_metric": 0.00013797394058201462,
          "validation_metric": 0.00023909061565063894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3899000319997867
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.006354807257594075,
          "train_loss": 0.015139594674110413,
          "validation_loss": 0.01125525962561369,
          "train_metric": 0.0015248045092448592,
          "validation_metric": 0.00113358860835433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4106396489987674
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.010932616889476776,
          "train_loss": 0.007124463561922312,
          "validation_loss": 0.008845338597893715,
          "train_metric": 0.00071754987584427,
          "validation_metric": 0.0008908701129257679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4308519269998214
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004576077175443061,
          "train_loss": 0.0023673835676163435,
          "validation_loss": 0.0022435954306274652,
          "train_metric": 0.000238434222410433,
          "validation_metric": 0.00022596673807129264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4511632659996394
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.004789368216734147,
          "train_loss": 0.004700355231761932,
          "validation_loss": 0.006186458747833967,
          "train_metric": 0.00047340255696326494,
          "validation_metric": 0.0006230775034055114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.471592777999831
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.004132697693421505,
          "train_loss": 0.003298452589660883,
          "validation_loss": 0.005111518781632185,
          "train_metric": 0.00033220808836631477,
          "validation_metric": 0.0005148134659975767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4926019639988226
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0029204953170847148,
          "train_loss": 0.0020211266819387674,
          "validation_loss": 0.0029522827826440334,
          "train_metric": 0.00020356047025416046,
          "validation_metric": 0.00029734312556684017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5123988739997003
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0023652956588193774,
          "train_loss": 0.00262121157720685,
          "validation_loss": 0.002600692445412278,
          "train_metric": 0.000263998779701069,
          "validation_metric": 0.00026193217490799725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5326619269999355
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0034190102960565127,
          "train_loss": 0.004208575934171677,
          "validation_loss": 0.005060605704784393,
          "train_metric": 0.0004238723195157945,
          "validation_metric": 0.0005096857203170657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5532517829997232
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.008092854994174559,
          "train_loss": 0.010237257927656174,
          "validation_loss": 0.00962753500789404,
          "train_metric": 0.0010310590732842684,
          "validation_metric": 0.0009696499910205603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5736341179999727
        }
      ],
      "best": {
        "epoch": 47,
        "updates": 752
      },
      "best_validation_metric": 0.00016977165068965405,
      "config": {
        "label": "LayerNorm + relu + constant LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.013029645071514958
    },
    "9d4c0f0efefb_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.5774121936410666,
          "train_loss": 0.294059157371521,
          "validation_loss": 0.18514592945575714,
          "train_metric": 0.02961656264960766,
          "validation_metric": 0.018647219985723495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1309335549995012
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.18550898879766464,
          "train_loss": 0.11212970316410065,
          "validation_loss": 0.08571828901767731,
          "train_metric": 0.011293292976915836,
          "validation_metric": 0.008633231744170189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24739159499949892
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.10298393480479717,
          "train_loss": 0.07773282378911972,
          "validation_loss": 0.07188624143600464,
          "train_metric": 0.007828965783119202,
          "validation_metric": 0.00724011892452836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4002511590006179
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.06412105564959347,
          "train_loss": 0.05094965919852257,
          "validation_loss": 0.04597301781177521,
          "train_metric": 0.005131462588906288,
          "validation_metric": 0.004630234092473984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5444296919995395
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.04303465294651687,
          "train_loss": 0.04300175979733467,
          "validation_loss": 0.04528236761689186,
          "train_metric": 0.004330980125814676,
          "validation_metric": 0.004560674540698528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6937595500003226
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.03830924828071147,
          "train_loss": 0.033392999321222305,
          "validation_loss": 0.03401530534029007,
          "train_metric": 0.0033632207196205854,
          "validation_metric": 0.0034258970990777016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7952175150003313
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.03397042921278626,
          "train_loss": 0.02902095392346382,
          "validation_loss": 0.0294038038700819,
          "train_metric": 0.002922884188592434,
          "validation_metric": 0.002961443504318595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8632255630000145
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.028644307632930577,
          "train_loss": 0.024392686784267426,
          "validation_loss": 0.024763032793998718,
          "train_metric": 0.0024567418731749058,
          "validation_metric": 0.0024940420407801867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9286228540004231
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.025791987834963948,
          "train_loss": 0.022494914010167122,
          "validation_loss": 0.021305473521351814,
          "train_metric": 0.002265605144202709,
          "validation_metric": 0.002145809354260564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9960852860003797
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.022267098247539252,
          "train_loss": 0.020542671903967857,
          "validation_loss": 0.018898980692029,
          "train_metric": 0.002068982692435384,
          "validation_metric": 0.0019034361466765404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.063274674999775
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.019326402485603467,
          "train_loss": 0.017142336815595627,
          "validation_loss": 0.016364816576242447,
          "train_metric": 0.0017265132628381252,
          "validation_metric": 0.0016482046339660883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1294850110007246
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.01831330923596397,
          "train_loss": 0.017703048884868622,
          "validation_loss": 0.01770043559372425,
          "train_metric": 0.0017829864518716931,
          "validation_metric": 0.0017827230039983988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1962291489999188
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.015926301362924278,
          "train_loss": 0.014232463203370571,
          "validation_loss": 0.012841502204537392,
          "train_metric": 0.001433441648259759,
          "validation_metric": 0.0012933490797877312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2623425309993763
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.014752681818208657,
          "train_loss": 0.014692189171910286,
          "validation_loss": 0.013351271860301495,
          "train_metric": 0.0014797435142099857,
          "validation_metric": 0.001344691263511777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3268712949993642
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.013684301731700543,
          "train_loss": 0.011956298723816872,
          "validation_loss": 0.011078478768467903,
          "train_metric": 0.0012041946174576879,
          "validation_metric": 0.001115783816203475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3928929230005451
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.013400915369857103,
          "train_loss": 0.012751118279993534,
          "validation_loss": 0.010455927811563015,
          "train_metric": 0.0012842458672821522,
          "validation_metric": 0.0010530828731134534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4599509899999248
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.013569886461482383,
          "train_loss": 0.010416709817945957,
          "validation_loss": 0.009776536375284195,
          "train_metric": 0.0010491330176591873,
          "validation_metric": 0.0009846568573266268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5273108239998692
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.012626554729649797,
          "train_loss": 0.011327638290822506,
          "validation_loss": 0.010171906091272831,
          "train_metric": 0.0011408781865611672,
          "validation_metric": 0.0010244771838188171,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6199694520000776
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.010195287963142619,
          "train_loss": 0.009805411100387573,
          "validation_loss": 0.008421489968895912,
          "train_metric": 0.000987565261311829,
          "validation_metric": 0.0008481815457344055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7290309889995115
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.009596882882760838,
          "train_loss": 0.011243026703596115,
          "validation_loss": 0.009939208626747131,
          "train_metric": 0.0011323565850034356,
          "validation_metric": 0.0010010406840592623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8112878230003844
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.009566899665514939,
          "train_loss": 0.008505701087415218,
          "validation_loss": 0.0083831287920475,
          "train_metric": 0.0008566630422137678,
          "validation_metric": 0.0008443179540336132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.884986002000005
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.008996637450763956,
          "train_loss": 0.007998993620276451,
          "validation_loss": 0.007226749788969755,
          "train_metric": 0.0008056294173002243,
          "validation_metric": 0.0007278518751263618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9961116650010808
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.008599064211011864,
          "train_loss": 0.00857299193739891,
          "validation_loss": 0.00689657311886549,
          "train_metric": 0.0008634403347969055,
          "validation_metric": 0.0006945976056158543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0559092399998917
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.00932047635433264,
          "train_loss": 0.010407332330942154,
          "validation_loss": 0.00988768506795168,
          "train_metric": 0.0010481884237378836,
          "validation_metric": 0.0009958514710888267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0870484269998997
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.012891171863884665,
          "train_loss": 0.010646202601492405,
          "validation_loss": 0.007915755733847618,
          "train_metric": 0.0010722465813159943,
          "validation_metric": 0.0007972458843141794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.11276167700089
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.00863263870269293,
          "train_loss": 0.007655404508113861,
          "validation_loss": 0.007404357194900513,
          "train_metric": 0.0007710243226028979,
          "validation_metric": 0.0007457397296093404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1331396990008216
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.00772539585159393,
          "train_loss": 0.005938301328569651,
          "validation_loss": 0.005495866760611534,
          "train_metric": 0.0005980839487165213,
          "validation_metric": 0.0005535235395655036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1544100049995905
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.006553043029271066,
          "train_loss": 0.005648879334330559,
          "validation_loss": 0.0051649645902216434,
          "train_metric": 0.0005689343670383096,
          "validation_metric": 0.0005201962776482105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1787175049994403
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.00929182502295589,
          "train_loss": 0.00946560688316822,
          "validation_loss": 0.00940791703760624,
          "train_metric": 0.0009533411939628422,
          "validation_metric": 0.0009475309634581208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1953361930009123
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.010813940956722945,
          "train_loss": 0.013771352358162403,
          "validation_loss": 0.00959679763764143,
          "train_metric": 0.001386999967508018,
          "validation_metric": 0.0009665543329901993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2122049179997703
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.012622722424566746,
          "train_loss": 0.02096780203282833,
          "validation_loss": 0.01925402693450451,
          "train_metric": 0.002111800480633974,
          "validation_metric": 0.0019391949754208326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2286268379994
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.011696335786837153,
          "train_loss": 0.006010301411151886,
          "validation_loss": 0.006196456961333752,
          "train_metric": 0.000605335517320782,
          "validation_metric": 0.0006240844959393144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.245352231999277
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.007980397160281427,
          "train_loss": 0.006278416141867638,
          "validation_loss": 0.005658129695802927,
          "train_metric": 0.000632339040748775,
          "validation_metric": 0.0005698660388588905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.261705109000104
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.0065069579795817845,
          "train_loss": 0.00524216890335083,
          "validation_loss": 0.004888336174190044,
          "train_metric": 0.0005279720644466579,
          "validation_metric": 0.0004923353553749621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.379810065000129
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.005760406238550786,
          "train_loss": 0.004586071241647005,
          "validation_loss": 0.004119434859603643,
          "train_metric": 0.00046189234126359224,
          "validation_metric": 0.0004148944281041622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4004605850004737
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.0045585226216644514,
          "train_loss": 0.005266420543193817,
          "validation_loss": 0.004402578342705965,
          "train_metric": 0.0005304146907292306,
          "validation_metric": 0.0004434115835465491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.421595633999459
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.006553589082614053,
          "train_loss": 0.005288745742291212,
          "validation_loss": 0.004342237953096628,
          "train_metric": 0.0005326630780473351,
          "validation_metric": 0.0004373343545012176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4530196680007066
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.008085504625341855,
          "train_loss": 0.006113831419497728,
          "validation_loss": 0.00677781505510211,
          "train_metric": 0.0006157626630738378,
          "validation_metric": 0.0006826367462053895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4778469530010625
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.008388264235691167,
          "train_loss": 0.006022359244525433,
          "validation_loss": 0.006610931362956762,
          "train_metric": 0.0006065500201657414,
          "validation_metric": 0.0006658287020400167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4969321270000364
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.006271929800277576,
          "train_loss": 0.004784443881362677,
          "validation_loss": 0.005455212201923132,
          "train_metric": 0.0004818716843146831,
          "validation_metric": 0.0005494290380738676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.516492037999342
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.004426749670528807,
          "train_loss": 0.004869739059358835,
          "validation_loss": 0.004735659807920456,
          "train_metric": 0.0004904622910544276,
          "validation_metric": 0.00047695834655314684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5360821230005968
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.00503799160651397,
          "train_loss": 0.00582069531083107,
          "validation_loss": 0.0053982301615178585,
          "train_metric": 0.000586239097174257,
          "validation_metric": 0.0005436899373307824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5544909050004208
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.004446250888577197,
          "train_loss": 0.004337501246482134,
          "validation_loss": 0.0032029259018599987,
          "train_metric": 0.0004368572263047099,
          "validation_metric": 0.00032258694409392774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5733760069997516
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.004557450061838608,
          "train_loss": 0.004902039188891649,
          "validation_loss": 0.00567815313115716,
          "train_metric": 0.0004937154008075595,
          "validation_metric": 0.0005718828178942204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.597712691000197
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.004582483365084045,
          "train_loss": 0.005697045009583235,
          "validation_loss": 0.005711521487683058,
          "train_metric": 0.0005737855099141598,
          "validation_metric": 0.0005752436118200421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6259058180003194
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.004090232592716347,
          "train_loss": 0.003039372619241476,
          "validation_loss": 0.003294390859082341,
          "train_metric": 0.00030611446709372103,
          "validation_metric": 0.0003317990049254149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6378535160001775
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.003721215463883709,
          "train_loss": 0.0045801871456205845,
          "validation_loss": 0.0031514237634837627,
          "train_metric": 0.00046129972906783223,
          "validation_metric": 0.000317399826599285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6499923880001006
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0034876778154284693,
          "train_loss": 0.002759418683126569,
          "validation_loss": 0.0028712067287415266,
          "train_metric": 0.00027791858883574605,
          "validation_metric": 0.00028917743475176394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6624948930002574
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.004967396831489168,
          "train_loss": 0.007330173160880804,
          "validation_loss": 0.007194535806775093,
          "train_metric": 0.0007382681360468268,
          "validation_metric": 0.0007246073218993843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.674210232000405
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.007276508884388022,
          "train_loss": 0.0035923055838793516,
          "validation_loss": 0.002843968803063035,
          "train_metric": 0.00036180386086925864,
          "validation_metric": 0.0002864341076929122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6865638559993386
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.004195019148028223,
          "train_loss": 0.0026765677612274885,
          "validation_loss": 0.002929920796304941,
          "train_metric": 0.00026957408408634365,
          "validation_metric": 0.0002950908965431154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.698294564999742
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.002959191991976695,
          "train_loss": 0.0028001193422824144,
          "validation_loss": 0.003547939006239176,
          "train_metric": 0.00028201774694025517,
          "validation_metric": 0.0003573354333639145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7100819980005326
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.0034247991388838273,
          "train_loss": 0.00523747131228447,
          "validation_loss": 0.0036896278616040945,
          "train_metric": 0.0005274988361634314,
          "validation_metric": 0.00037160576903261244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.721378382999319
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.004353811600594781,
          "train_loss": 0.0026113770436495543,
          "validation_loss": 0.0026402571238577366,
          "train_metric": 0.00026300831814296544,
          "validation_metric": 0.0002659170131664723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.733457188000102
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.003151958269882016,
          "train_loss": 0.0036025396548211575,
          "validation_loss": 0.004043330438435078,
          "train_metric": 0.0003628346021287143,
          "validation_metric": 0.0004072293813806027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.745523024999784
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0032559617884544423,
          "train_loss": 0.002714689588174224,
          "validation_loss": 0.0033075218088924885,
          "train_metric": 0.0002734135778155178,
          "validation_metric": 0.00033312145387753844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.757119772000806
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.004565099468891276,
          "train_loss": 0.004512454383075237,
          "validation_loss": 0.005886324681341648,
          "train_metric": 0.0004544778785202652,
          "validation_metric": 0.0005928490427322686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7687503670003935
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0032115350622916594,
          "train_loss": 0.0026501223910599947,
          "validation_loss": 0.002537619788199663,
          "train_metric": 0.0002669106761459261,
          "validation_metric": 0.0002555797982495278,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.780885185000443
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.005201598945859587,
          "train_loss": 0.005197241902351379,
          "validation_loss": 0.007132071070373058,
          "train_metric": 0.0005234472337178886,
          "validation_metric": 0.0007183161214925349,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7924760730002163
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0042155816918239,
          "train_loss": 0.0025268460158258677,
          "validation_loss": 0.002925586886703968,
          "train_metric": 0.00025449469103477895,
          "validation_metric": 0.00029465442639775574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8041968630004703
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0024052051012404263,
          "train_loss": 0.0026287937071174383,
          "validation_loss": 0.0022091022692620754,
          "train_metric": 0.0002647624642122537,
          "validation_metric": 0.0002224926429335028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.816297409000981
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0031572032894473523,
          "train_loss": 0.0037825582548975945,
          "validation_loss": 0.004320094827562571,
          "train_metric": 0.0003809654444921762,
          "validation_metric": 0.0004351041279733181,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8282274220000545
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.002708098301809514,
          "train_loss": 0.002279422013089061,
          "validation_loss": 0.00315505126491189,
          "train_metric": 0.0002295750455232337,
          "validation_metric": 0.0003177651669830084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.839902183999584
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.002843628612026805,
          "train_loss": 0.0028118446934968233,
          "validation_loss": 0.003716384992003441,
          "train_metric": 0.0002831987221725285,
          "validation_metric": 0.0003743006382137537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8517385440009093
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0035039567155763507,
          "train_loss": 0.003793995128944516,
          "validation_loss": 0.005207617301493883,
          "train_metric": 0.0003821173158939928,
          "validation_metric": 0.0005244921776466072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8633749870004976
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0029094393612467684,
          "train_loss": 0.002119696931913495,
          "validation_loss": 0.002887309528887272,
          "train_metric": 0.00021348810696508735,
          "validation_metric": 0.0002907992457039654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8747762530001637
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.002472376305377111,
          "train_loss": 0.0017696898430585861,
          "validation_loss": 0.002405409933999181,
          "train_metric": 0.0001782366744009778,
          "validation_metric": 0.00024226406821981072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8862316430004284
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0022274950351857115,
          "train_loss": 0.002916512778028846,
          "validation_loss": 0.0020778505131602287,
          "train_metric": 0.00029374047880992293,
          "validation_metric": 0.00020927347941324115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.89845927100032
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0024405498734267894,
          "train_loss": 0.001757952501066029,
          "validation_loss": 0.0027827334124594927,
          "train_metric": 0.00017705450591165572,
          "validation_metric": 0.00028026674408465624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9105498970002373
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.002558383597715874,
          "train_loss": 0.00683883810415864,
          "validation_loss": 0.0037830271758139133,
          "train_metric": 0.0006887827767059207,
          "validation_metric": 0.00038101262180134654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9221343330009404
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0030386707730940543,
          "train_loss": 0.0021613386925309896,
          "validation_loss": 0.004035930149257183,
          "train_metric": 0.0002176820853492245,
          "validation_metric": 0.00040648411959409714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9336348720007663
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.002486728481017053,
          "train_loss": 0.004119814373552799,
          "validation_loss": 0.0035093501210212708,
          "train_metric": 0.00041493267053738236,
          "validation_metric": 0.0003534488787408918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9450867800005653
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.003095766016485868,
          "train_loss": 0.0033013192005455494,
          "validation_loss": 0.0047364914789795876,
          "train_metric": 0.0003324967110529542,
          "validation_metric": 0.0004770420491695404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.956156860000192
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0040786805911920965,
          "train_loss": 0.007741731591522694,
          "validation_loss": 0.004469695966690779,
          "train_metric": 0.0007797188591212034,
          "validation_metric": 0.0004501714138314128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9676790539997455
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0056443685607519,
          "train_loss": 0.0029131171759217978,
          "validation_loss": 0.004777350462973118,
          "train_metric": 0.0002933985088020563,
          "validation_metric": 0.0004811572143808007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9792141460002313
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0030780239358136896,
          "train_loss": 0.0035530428867787123,
          "validation_loss": 0.003245867323130369,
          "train_metric": 0.0003578494652174413,
          "validation_metric": 0.00032691186061128974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9909744809992844
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.0035905873446608894,
          "train_loss": 0.0047118403017520905,
          "validation_loss": 0.004935065284371376,
          "train_metric": 0.00047455934691242874,
          "validation_metric": 0.0004970416775904596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0028057979998266
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.003453753361100098,
          "train_loss": 0.0024621791671961546,
          "validation_loss": 0.004497029818594456,
          "train_metric": 0.00024798166123218834,
          "validation_metric": 0.00045292434515431523,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0146310960008122
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0019890131188731175,
          "train_loss": 0.0018999604508280754,
          "validation_loss": 0.002862765220925212,
          "train_metric": 0.00019135705952066928,
          "validation_metric": 0.0002883272245526314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.026284638999641
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0032120241594384424,
          "train_loss": 0.0034094289876520634,
          "validation_loss": 0.005056427326053381,
          "train_metric": 0.00034338521072641015,
          "validation_metric": 0.0005092648207210004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.037489655000172
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.003336558824230451,
          "train_loss": 0.0043709068559110165,
          "validation_loss": 0.0027086969930678606,
          "train_metric": 0.00044022174552083015,
          "validation_metric": 0.0002728100516833365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.052499502000501
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0030609259229095187,
          "train_loss": 0.003848441643640399,
          "validation_loss": 0.005840881261974573,
          "train_metric": 0.0003876009432133287,
          "validation_metric": 0.0005882722325623035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0809331399996154
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.004044876302941702,
          "train_loss": 0.003150316420942545,
          "validation_loss": 0.0024858368560671806,
          "train_metric": 0.0003172883589286357,
          "validation_metric": 0.00025036439183168113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1030398580005567
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0023190899737528525,
          "train_loss": 0.00216874061152339,
          "validation_loss": 0.0035732698161154985,
          "train_metric": 0.0002184276090702042,
          "validation_metric": 0.0003598867042455822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.114280793999569
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.003590360101952683,
          "train_loss": 0.0018713378813117743,
          "validation_loss": 0.003382258815690875,
          "train_metric": 0.00018847431056201458,
          "validation_metric": 0.0003406486939638853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1259755619994394
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.002627471461892128,
          "train_loss": 0.002222866052761674,
          "validation_loss": 0.0034534898586571217,
          "train_metric": 0.00022387891658581793,
          "validation_metric": 0.0003478228463791311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1375858610008436
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0022042072341719177,
          "train_loss": 0.0031039670575410128,
          "validation_loss": 0.0016907412791624665,
          "train_metric": 0.000312620191834867,
          "validation_metric": 0.0001702852750895545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1498361119993206
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0024366691141040064,
          "train_loss": 0.001888641738332808,
          "validation_loss": 0.003663487732410431,
          "train_metric": 0.00019021707703359425,
          "validation_metric": 0.0003689730365294963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1618330990004324
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.00476427480498387,
          "train_loss": 0.003981828223913908,
          "validation_loss": 0.006923143286257982,
          "train_metric": 0.00040103515493683517,
          "validation_metric": 0.00069727364461869,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.173487654999917
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.004823016977752559,
          "train_loss": 0.0052878353744745255,
          "validation_loss": 0.005548713728785515,
          "train_metric": 0.0005325714591890574,
          "validation_metric": 0.0005588461644947529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1848930640007893
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.004909237039100844,
          "train_loss": 0.002157811773940921,
          "validation_loss": 0.0019987064879387617,
          "train_metric": 0.0002173268876504153,
          "validation_metric": 0.0002013024059124291,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1968027820003044
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0026352187069278443,
          "train_loss": 0.009383774362504482,
          "validation_loss": 0.011356707662343979,
          "train_metric": 0.0009450993966311216,
          "validation_metric": 0.0011438061483204365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2086237460007396
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004923979904560838,
          "train_loss": 0.005509258713573217,
          "validation_loss": 0.0029702011961489916,
          "train_metric": 0.0005548723274841905,
          "validation_metric": 0.0002991477958858013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.220258979999926
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.005007215127989184,
          "train_loss": 0.00580101041123271,
          "validation_loss": 0.007136151194572449,
          "train_metric": 0.0005842566024512053,
          "validation_metric": 0.000718727009370923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.231750238001041
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0030804566813458223,
          "train_loss": 0.002407587133347988,
          "validation_loss": 0.0028656218200922012,
          "train_metric": 0.0002424833655823022,
          "validation_metric": 0.00028861494502052665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.243778141999428
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0017192719496961217,
          "train_loss": 0.0014442848041653633,
          "validation_loss": 0.0016772303497418761,
          "train_metric": 0.0001454630692023784,
          "validation_metric": 0.00016892449639271945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.255907709000894
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0019828030326607404,
          "train_loss": 0.0012381428387016058,
          "validation_loss": 0.002003708854317665,
          "train_metric": 0.00012470119690988213,
          "validation_metric": 0.00020180619321763515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2674264190009126
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0031226869468810037,
          "train_loss": 0.005759143736213446,
          "validation_loss": 0.008629643358290195,
          "train_metric": 0.0005800397484563291,
          "validation_metric": 0.0008691459661349654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.280939035999836
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0029638198975590058,
          "train_loss": 0.0014986784663051367,
          "validation_loss": 0.0017633503302931786,
          "train_metric": 0.00015094141417648643,
          "validation_metric": 0.00017759819456841797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2944278190007026
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.002100569570757216,
          "train_loss": 0.002503404626622796,
          "validation_loss": 0.003953541629016399,
          "train_metric": 0.0002521337301004678,
          "validation_metric": 0.00039818621007725596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3088125659996876
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.002240743655420374,
          "train_loss": 0.0022745169699192047,
          "validation_loss": 0.0016846813960000873,
          "train_metric": 0.00022908103710506111,
          "validation_metric": 0.0001696749241091311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.322799630999725
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0017987300216191215,
          "train_loss": 0.0013937337789684534,
          "validation_loss": 0.002752286847680807,
          "train_metric": 0.00014037176151759923,
          "validation_metric": 0.0002772002771962434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.338959413000339
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0015646795418433612,
          "train_loss": 0.0037900833413004875,
          "validation_loss": 0.0019015310099348426,
          "train_metric": 0.0003817233082372695,
          "validation_metric": 0.00019151523883920163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3535111940000206
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.002880423911847174,
          "train_loss": 0.001885311445221305,
          "validation_loss": 0.002280502812936902,
          "train_metric": 0.0001898816553875804,
          "validation_metric": 0.00022968387929722667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4692543870005466
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.003936280139896553,
          "train_loss": 0.004738356918096542,
          "validation_loss": 0.00599646894261241,
          "train_metric": 0.0004772299434989691,
          "validation_metric": 0.0006039424915798008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4821763830004784
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0028263399908610154,
          "train_loss": 0.0013622217811644077,
          "validation_loss": 0.0019364948384463787,
          "train_metric": 0.00013719798880629241,
          "validation_metric": 0.00019503665680531412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.497220651999669
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.00160372248137719,
          "train_loss": 0.005502783693373203,
          "validation_loss": 0.008414536714553833,
          "train_metric": 0.0005542202852666378,
          "validation_metric": 0.0008474813657812774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.513094211999487
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.003067993857257534,
          "train_loss": 0.0018564235651865602,
          "validation_loss": 0.0017830643337219954,
          "train_metric": 0.0001869721891125664,
          "validation_metric": 0.00017958373064175248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5382606980001583
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.00274656202600454,
          "train_loss": 0.0021645219530910254,
          "validation_loss": 0.002295372076332569,
          "train_metric": 0.00021800270769745111,
          "validation_metric": 0.00023118150420486927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5560361679999914
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.002225847139925463,
          "train_loss": 0.00217449595220387,
          "validation_loss": 0.002930194139480591,
          "train_metric": 0.00021900727006141096,
          "validation_metric": 0.00029511837055906653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5779096109999955
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.0020853535970672965,
          "train_loss": 0.0028707794845104218,
          "validation_loss": 0.004462416283786297,
          "train_metric": 0.000289134361082688,
          "validation_metric": 0.00044943817192688584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5978747400004067
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.002370399883147911,
          "train_loss": 0.0029257566202431917,
          "validation_loss": 0.0037477570585906506,
          "train_metric": 0.00029467151034623384,
          "validation_metric": 0.00037746032467111945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6139397870010725
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0018251223791594384,
          "train_loss": 0.0032321754842996597,
          "validation_loss": 0.001961696892976761,
          "train_metric": 0.0003255328629165888,
          "validation_metric": 0.00019757493282668293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6281338339995273
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.004322474851505831,
          "train_loss": 0.002477467991411686,
          "validation_loss": 0.003668795106932521,
          "train_metric": 0.0002495214866939932,
          "validation_metric": 0.00036950758658349514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6404465329997038
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.002239982750324998,
          "train_loss": 0.0018282387172803283,
          "validation_loss": 0.0028423776384443045,
          "train_metric": 0.00018413351790513843,
          "validation_metric": 0.00028627386200241745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.661135460000878
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.001816645542930928,
          "train_loss": 0.0012853818479925394,
          "validation_loss": 0.0013589411973953247,
          "train_metric": 0.00012945894559379667,
          "validation_metric": 0.0001368675584672019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.67837808799959
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0022456374081230024,
          "train_loss": 0.002932559931650758,
          "validation_loss": 0.0044084349647164345,
          "train_metric": 0.0002953567018266767,
          "validation_metric": 0.0004440014890860766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6953446900006384
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.002309491472260561,
          "train_loss": 0.0015450005885213614,
          "validation_loss": 0.0013614309718832374,
          "train_metric": 0.00015560680185444653,
          "validation_metric": 0.00013711834617424756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7117904030001228
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0028746211719408166,
          "train_loss": 0.0028911081608384848,
          "validation_loss": 0.003177341539412737,
          "train_metric": 0.0002911818155553192,
          "validation_metric": 0.0003200102364644408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.725421921000816
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.001682287398580229,
          "train_loss": 0.0013422915944829583,
          "validation_loss": 0.0017795178573578596,
          "train_metric": 0.00013519071217160672,
          "validation_metric": 0.00017922652477864176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.738792320000357
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0018040796239802148,
          "train_loss": 0.002127530286088586,
          "validation_loss": 0.0035116730723530054,
          "train_metric": 0.0002142770536011085,
          "validation_metric": 0.00035368281533010304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7508177290001186
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0035885336856154026,
          "train_loss": 0.0024930397048592567,
          "validation_loss": 0.003582203993573785,
          "train_metric": 0.0002510898048058152,
          "validation_metric": 0.00036078644916415215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.762592434000908
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.002440041942463722,
          "train_loss": 0.002071328228339553,
          "validation_loss": 0.003137587569653988,
          "train_metric": 0.0002086165768560022,
          "validation_metric": 0.0003160063351970166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.774297674999616
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0015053635343065253,
          "train_loss": 0.0013379281153902411,
          "validation_loss": 0.00139755685813725,
          "train_metric": 0.00013475121522787958,
          "validation_metric": 0.0001407568051945418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7858932969993475
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0011997421679552644,
          "train_loss": 0.003375915577635169,
          "validation_loss": 0.006228860467672348,
          "train_metric": 0.00034000986488536,
          "validation_metric": 0.0006273479666560888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.797163178000119
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0016255568552878685,
          "train_loss": 0.001654527266509831,
          "validation_loss": 0.003184559755027294,
          "train_metric": 0.00016663791029714048,
          "validation_metric": 0.0003207371919415891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8088241790010215
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0016060371508501703,
          "train_loss": 0.0013549907598644495,
          "validation_loss": 0.0032393913716077805,
          "train_metric": 0.00013646972365677357,
          "validation_metric": 0.0003262596146669239,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8206138530003955
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0013997153928357875,
          "train_loss": 0.001190817798487842,
          "validation_loss": 0.0032645934261381626,
          "train_metric": 0.00011993480438832194,
          "validation_metric": 0.00032879787613637745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8319814030001
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0008378774891752983,
          "train_loss": 0.0025929270777851343,
          "validation_loss": 0.0011361116776242852,
          "train_metric": 0.00026115012587979436,
          "validation_metric": 0.00011442499817349017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.844190752999566
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0021985557832522318,
          "train_loss": 0.002255830680951476,
          "validation_loss": 0.00395776005461812,
          "train_metric": 0.00022719899425283074,
          "validation_metric": 0.0003986110968980938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8558562119997077
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.002652918206877075,
          "train_loss": 0.0019042760832235217,
          "validation_loss": 0.0037489726673811674,
          "train_metric": 0.00019179171067662537,
          "validation_metric": 0.0003775827353820205,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8672501370001555
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0041515064585837536,
          "train_loss": 0.003259880468249321,
          "validation_loss": 0.00412637647241354,
          "train_metric": 0.0003283232799731195,
          "validation_metric": 0.0004155935603193939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.878715873999681
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.004385384756460553,
          "train_loss": 0.004532619845122099,
          "validation_loss": 0.004682764410972595,
          "train_metric": 0.0004565088893286884,
          "validation_metric": 0.0004716308612842113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.890097980000064
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.006797151028877124,
          "train_loss": 0.005427464377135038,
          "validation_loss": 0.0031488772947341204,
          "train_metric": 0.0005466343136504292,
          "validation_metric": 0.0003171433927491307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9017502829992736
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0033017042915162165,
          "train_loss": 0.0015741141978651285,
          "validation_loss": 0.0030499815475195646,
          "train_metric": 0.00015853899822104722,
          "validation_metric": 0.0003071829560212791,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9135589390007226
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.003241788515879307,
          "train_loss": 0.005246582441031933,
          "validation_loss": 0.006868756376206875,
          "train_metric": 0.0005284165381453931,
          "validation_metric": 0.0006917960126884282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9252189290000388
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0031301887611334678,
          "train_loss": 0.0013254234800115228,
          "validation_loss": 0.002793177030980587,
          "train_metric": 0.00013349179062061012,
          "validation_metric": 0.00028131852741353214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9367342090008606
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0018194815893366467,
          "train_loss": 0.002093493938446045,
          "validation_loss": 0.002503735711798072,
          "train_metric": 0.00021084904437884688,
          "validation_metric": 0.0002521670830901712,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9482009170005767
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.002162508813853492,
          "train_loss": 0.0017664936603978276,
          "validation_loss": 0.0030482665169984102,
          "train_metric": 0.0001779147714842111,
          "validation_metric": 0.00030701025389134884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.959823568000502
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0017917106324603083,
          "train_loss": 0.0011647827923297882,
          "validation_loss": 0.00321584171615541,
          "train_metric": 0.00011731266567949206,
          "validation_metric": 0.0003238878271076828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.971478681000008
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0014370912813319592,
          "train_loss": 0.0015338376397266984,
          "validation_loss": 0.0032264720648527145,
          "train_metric": 0.00015448249178007245,
          "validation_metric": 0.0003249584697186947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.98304319700037
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.002439240171952406,
          "train_loss": 0.004098306410014629,
          "validation_loss": 0.001488276175223291,
          "train_metric": 0.00041276641422882676,
          "validation_metric": 0.0001498937199357897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9947760159993777
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.002375960913923336,
          "train_loss": 0.003497058991342783,
          "validation_loss": 0.006043639034032822,
          "train_metric": 0.0003522109764162451,
          "validation_metric": 0.0006086931098252535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.006205972000316
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0014905951647961047,
          "train_loss": 0.001169357681646943,
          "validation_loss": 0.0017256271094083786,
          "train_metric": 0.00011777342297136784,
          "validation_metric": 0.00017379883502144367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.017770593000023
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.001261683684788295,
          "train_loss": 0.0013655294897034764,
          "validation_loss": 0.0033411383628845215,
          "train_metric": 0.00013753112580161542,
          "validation_metric": 0.0003365072188898921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.029271864999828
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0008869130106177181,
          "train_loss": 0.0010938713094219565,
          "validation_loss": 0.0032618262339383364,
          "train_metric": 0.00011017070210073143,
          "validation_metric": 0.00032851920695975423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0409929910001665
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0013422430547507247,
          "train_loss": 0.0014236595015972853,
          "validation_loss": 0.001303720404393971,
          "train_metric": 0.0001433857687516138,
          "validation_metric": 0.00013130594743415713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.052524872000504
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0013685834746866021,
          "train_loss": 0.0016441475600004196,
          "validation_loss": 0.002029459225013852,
          "train_metric": 0.0001655925443628803,
          "validation_metric": 0.00020439970830921084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.064031939000415
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.002770109706034418,
          "train_loss": 0.0043099853210151196,
          "validation_loss": 0.006131511181592941,
          "train_metric": 0.00043408593046478927,
          "validation_metric": 0.0006175433518365026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.075469629000509
        }
      ],
      "best": {
        "epoch": 129,
        "updates": 2064
      },
      "best_validation_metric": 0.00011442499817349017,
      "config": {
        "label": "ReLU",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010696962100217528
    },
    "be44f6a2021f_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.19096134742721915,
          "train_loss": 0.07242120057344437,
          "validation_loss": 0.0500011146068573,
          "train_metric": 0.007293997798115015,
          "validation_metric": 0.005035929381847382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.015692491999288904
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.05849547719117254,
          "train_loss": 0.038079965859651566,
          "validation_loss": 0.02615676075220108,
          "train_metric": 0.003835274837911129,
          "validation_metric": 0.0026344130747020245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.030759475999730057
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.03498553414829075,
          "train_loss": 0.02507447637617588,
          "validation_loss": 0.02035215124487877,
          "train_metric": 0.0025254094507545233,
          "validation_metric": 0.002049794187769294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04576137100048072
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.023305899696424603,
          "train_loss": 0.019549626857042313,
          "validation_loss": 0.014797462150454521,
          "train_metric": 0.0019689667969942093,
          "validation_metric": 0.0014903461560606956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06034782999995514
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.03700241423211992,
          "train_loss": 0.023921405896544456,
          "validation_loss": 0.018056131899356842,
          "train_metric": 0.002409276319667697,
          "validation_metric": 0.0018185474909842014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07432291899931442
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.028315134666627273,
          "train_loss": 0.022200098261237144,
          "validation_loss": 0.014051035977900028,
          "train_metric": 0.0022359120193868876,
          "validation_metric": 0.0014151687500998378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0894214740001189
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.03941670642234385,
          "train_loss": 0.047134917229413986,
          "validation_loss": 0.04291398823261261,
          "train_metric": 0.004747255705296993,
          "validation_metric": 0.004322139546275139,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10353084700000181
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.03125104267382994,
          "train_loss": 0.025500396266579628,
          "validation_loss": 0.018091389909386635,
          "train_metric": 0.002568306401371956,
          "validation_metric": 0.0018220983911305666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1177432200001931
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.01859823355334811,
          "train_loss": 0.014343221671879292,
          "validation_loss": 0.014042512513697147,
          "train_metric": 0.0014445967972278595,
          "validation_metric": 0.0014143104199320078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13213671399898885
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.01476419367827475,
          "train_loss": 0.011451940052211285,
          "validation_loss": 0.011174211278557777,
          "train_metric": 0.0011533974902704358,
          "validation_metric": 0.0011254255659878254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14750525099952938
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.013102602475555614,
          "train_loss": 0.009634414687752724,
          "validation_loss": 0.009227216243743896,
          "train_metric": 0.000970343011431396,
          "validation_metric": 0.0009293314651586115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16236421800022072
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.01181740511674434,
          "train_loss": 0.027752339839935303,
          "validation_loss": 0.021214960142970085,
          "train_metric": 0.002795114181935787,
          "validation_metric": 0.0021366931032389402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17674406100013584
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.024153543607098982,
          "train_loss": 0.010684475302696228,
          "validation_loss": 0.0072398437187075615,
          "train_metric": 0.001076101209037006,
          "validation_metric": 0.0007291705696843565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19203906599977927
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.020216720033204183,
          "train_loss": 0.012447123415768147,
          "validation_loss": 0.013840309344232082,
          "train_metric": 0.0012536287540569901,
          "validation_metric": 0.0013939451891928911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2056627080000908
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.0160851250984706,
          "train_loss": 0.020313572138547897,
          "validation_loss": 0.014188272878527641,
          "train_metric": 0.002045908709987998,
          "validation_metric": 0.0014289908576756716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21971880700039037
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.01265521373716183,
          "train_loss": 0.01249550748616457,
          "validation_loss": 0.008259164169430733,
          "train_metric": 0.001258501666598022,
          "validation_metric": 0.000831832701805979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23368911400029901
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.008663570624776185,
          "train_loss": 0.008145946078002453,
          "validation_loss": 0.005313525442034006,
          "train_metric": 0.0008204297628253698,
          "validation_metric": 0.0005351588479243219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24843305600006715
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.010117787925992161,
          "train_loss": 0.006461773533374071,
          "validation_loss": 0.004270083270967007,
          "train_metric": 0.0006508060614578426,
          "validation_metric": 0.0004300671280361712,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26303127000028326
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.011943971047003288,
          "train_loss": 0.006918198429048061,
          "validation_loss": 0.005362193565815687,
          "train_metric": 0.0006967756198719144,
          "validation_metric": 0.0005400604568421841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27655419099937717
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.007399222507956438,
          "train_loss": 0.005954752676188946,
          "validation_loss": 0.00478105153888464,
          "train_metric": 0.000599740887992084,
          "validation_metric": 0.0004815299471374601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2901779259991599
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.007807362839230336,
          "train_loss": 0.005216581281274557,
          "validation_loss": 0.003698101732879877,
          "train_metric": 0.0005253948620520532,
          "validation_metric": 0.00037245929706841707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30429019199982577
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.007189471689343918,
          "train_loss": 0.005692244507372379,
          "validation_loss": 0.006105087231844664,
          "train_metric": 0.0005733020952902734,
          "validation_metric": 0.000614881981164217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3179714059988328
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01745169445348438,
          "train_loss": 0.02158437669277191,
          "validation_loss": 0.019700421020388603,
          "train_metric": 0.0021738994400948286,
          "validation_metric": 0.0019841541070491076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33130762099972344
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.016935257881414145,
          "train_loss": 0.011732752434909344,
          "validation_loss": 0.015319348312914371,
          "train_metric": 0.0011816798942163587,
          "validation_metric": 0.0015429084887728095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3448070809990895
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.008408344830968417,
          "train_loss": 0.007929813116788864,
          "validation_loss": 0.005407745484262705,
          "train_metric": 0.0007986616692505777,
          "validation_metric": 0.0005446483264677227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35842033799963247
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.007061928350594826,
          "train_loss": 0.006596588063985109,
          "validation_loss": 0.006484700832515955,
          "train_metric": 0.0006643841043114662,
          "validation_metric": 0.0006531153339892626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3718919379989529
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.012771852227160707,
          "train_loss": 0.007400437258183956,
          "validation_loss": 0.008532926440238953,
          "train_metric": 0.0007453449070453644,
          "validation_metric": 0.0008594051469117403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3856367289990885
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.017370946967275813,
          "train_loss": 0.011434217914938927,
          "validation_loss": 0.008804738521575928,
          "train_metric": 0.0011516126105561852,
          "validation_metric": 0.0008867810829542577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39917980899917893
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.02165296656312421,
          "train_loss": 0.010835460387170315,
          "validation_loss": 0.009168559685349464,
          "train_metric": 0.0010913078440353274,
          "validation_metric": 0.0009234237368218601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4131084170003305
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.015899773105047643,
          "train_loss": 0.007458425592631102,
          "validation_loss": 0.005842586047947407,
          "train_metric": 0.0007511853473260999,
          "validation_metric": 0.0005884440615773201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4271963049995975
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008075340345385484,
          "train_loss": 0.011918378062546253,
          "validation_loss": 0.0076078916899859905,
          "train_metric": 0.0012003753799945116,
          "validation_metric": 0.0007662390125915408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4408182629995281
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.014930206874851137,
          "train_loss": 0.01142839714884758,
          "validation_loss": 0.011296026408672333,
          "train_metric": 0.0011510264594107866,
          "validation_metric": 0.0011376944603398442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4547697719990538
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.010705090418923646,
          "train_loss": 0.01712808571755886,
          "validation_loss": 0.012351919896900654,
          "train_metric": 0.001725077978335321,
          "validation_metric": 0.0012440400896593928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46840383100061445
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01758727661217563,
          "train_loss": 0.012319177389144897,
          "validation_loss": 0.010577203705906868,
          "train_metric": 0.0012407422764226794,
          "validation_metric": 0.0010652970522642136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4823560009990615
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.00873475598928053,
          "train_loss": 0.005653894040733576,
          "validation_loss": 0.005312873981893063,
          "train_metric": 0.0005694394931197166,
          "validation_metric": 0.0005350932478904724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4961626799995429
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.007662540898309089,
          "train_loss": 0.01242721639573574,
          "validation_loss": 0.010057749226689339,
          "train_metric": 0.001251623616553843,
          "validation_metric": 0.0010129796573892236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5099731890004477
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.009284880899940617,
          "train_loss": 0.006296744104474783,
          "validation_loss": 0.004750421736389399,
          "train_metric": 0.0006341850385069847,
          "validation_metric": 0.0004784450866281986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5238916199996311
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.0056952688755700365,
          "train_loss": 0.004445214290171862,
          "validation_loss": 0.0042821867391467094,
          "train_metric": 0.0004477056791074574,
          "validation_metric": 0.0004312862001825124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5375538070002222
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.004478971393837128,
          "train_loss": 0.005975057836622,
          "validation_loss": 0.004622609354555607,
          "train_metric": 0.0006017859559506178,
          "validation_metric": 0.0004655722586903721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5512521359996754
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.0062652170527144335,
          "train_loss": 0.0042640212923288345,
          "validation_loss": 0.005077602341771126,
          "train_metric": 0.0004294565296731889,
          "validation_metric": 0.000511397433001548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5650447759999224
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.005206328321946785,
          "train_loss": 0.003924076911062002,
          "validation_loss": 0.003398760687559843,
          "train_metric": 0.00039521866710856557,
          "validation_metric": 0.00034231075551360846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5797929219988873
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.007653391556232236,
          "train_loss": 0.00622263178229332,
          "validation_loss": 0.005182477179914713,
          "train_metric": 0.0006267207209020853,
          "validation_metric": 0.0005219600861892104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5933517789999314
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.01490420536720194,
          "train_loss": 0.027935141697525978,
          "validation_loss": 0.027941394597291946,
          "train_metric": 0.002813525265082717,
          "validation_metric": 0.0028141550719738007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6067683879991819
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.009290112473536283,
          "train_loss": 0.005340253468602896,
          "validation_loss": 0.0038507459685206413,
          "train_metric": 0.000537850777618587,
          "validation_metric": 0.0003878330171573907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6206555739991018
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.005773664073785767,
          "train_loss": 0.004043286666274071,
          "validation_loss": 0.0033850104082375765,
          "train_metric": 0.00040722498670220375,
          "validation_metric": 0.0003409258497413248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6347308709991921
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.00846997057669796,
          "train_loss": 0.011231747455894947,
          "validation_loss": 0.00984361581504345,
          "train_metric": 0.0011312204878777266,
          "validation_metric": 0.000991413020528853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6483703379999497
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.007452010904671624,
          "train_loss": 0.004567780066281557,
          "validation_loss": 0.005149920005351305,
          "train_metric": 0.0004600501270033419,
          "validation_metric": 0.0005186810158193111,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6625426020000305
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.004643035572371446,
          "train_loss": 0.008115299046039581,
          "validation_loss": 0.006411276292055845,
          "train_metric": 0.0008173431269824505,
          "validation_metric": 0.0006457202835008502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6761226679991523
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.006204894780239556,
          "train_loss": 0.004889220930635929,
          "validation_loss": 0.004092714749276638,
          "train_metric": 0.0004924244130961597,
          "validation_metric": 0.0004122032260056585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6898994620005396
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.005712241669243667,
          "train_loss": 0.006028683390468359,
          "validation_loss": 0.003571887267753482,
          "train_metric": 0.0006071868701837957,
          "validation_metric": 0.0003597473551053554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.703665674000149
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007643929682672024,
          "train_loss": 0.010405834764242172,
          "validation_loss": 0.011365502141416073,
          "train_metric": 0.0010480375494807959,
          "validation_metric": 0.001144691719673574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7180795639997086
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.014325551019283012,
          "train_loss": 0.008878681808710098,
          "validation_loss": 0.008927879855036736,
          "train_metric": 0.0008942283457145095,
          "validation_metric": 0.0008991833310574293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7321713580004143
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.007739027278148569,
          "train_loss": 0.0036479460541158915,
          "validation_loss": 0.002636530203744769,
          "train_metric": 0.00036740777431987226,
          "validation_metric": 0.0002655416610650718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7501777559991751
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.006650625742622651,
          "train_loss": 0.009196588769555092,
          "validation_loss": 0.008692975156009197,
          "train_metric": 0.0009262466337531805,
          "validation_metric": 0.0008755245362408459,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7676444360004098
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.005418006141553633,
          "train_loss": 0.008098069578409195,
          "validation_loss": 0.005245323292911053,
          "train_metric": 0.0008156078401952982,
          "validation_metric": 0.0005282897036522627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7850586080003268
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.004031276435853215,
          "train_loss": 0.0024014683440327644,
          "validation_loss": 0.0022977094631642103,
          "train_metric": 0.0002418671065242961,
          "validation_metric": 0.00023141685232985765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8173563189993729
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.006051160591596272,
          "train_loss": 0.00354887917637825,
          "validation_loss": 0.0021738363429903984,
          "train_metric": 0.0003574301372282207,
          "validation_metric": 0.00021894084056839347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8405839499991998
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.021515964770514984,
          "train_loss": 0.022570086643099785,
          "validation_loss": 0.022444233298301697,
          "train_metric": 0.0022731765639036894,
          "validation_metric": 0.0022605007980018854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8547246950001863
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.019276428647572175,
          "train_loss": 0.013129238039255142,
          "validation_loss": 0.012197362259030342,
          "train_metric": 0.0013223285786807537,
          "validation_metric": 0.001228473731316626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8725110960003803
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.007101236100425012,
          "train_loss": 0.006458396557718515,
          "validation_loss": 0.0058199018239974976,
          "train_metric": 0.000650466070510447,
          "validation_metric": 0.0005861591780558228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8901861079993978
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004585860078805126,
          "train_loss": 0.004941094666719437,
          "validation_loss": 0.004171045497059822,
          "train_metric": 0.0004976489581167698,
          "validation_metric": 0.0004200924595352262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9055258899988985
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0034054697607643902,
          "train_loss": 0.005723410286009312,
          "validation_loss": 0.005059115123003721,
          "train_metric": 0.0005764408851973712,
          "validation_metric": 0.0005095354863442481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9192826700000296
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.006766643738956191,
          "train_loss": 0.0043543921783566475,
          "validation_loss": 0.004314951132982969,
          "train_metric": 0.00043855843250639737,
          "validation_metric": 0.00043458607979118824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9328714700004639
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.005852240006788634,
          "train_loss": 0.009078649803996086,
          "validation_loss": 0.00744221918284893,
          "train_metric": 0.0009143684292212129,
          "validation_metric": 0.0007495530880987644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.946992113998931
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.004776617875904776,
          "train_loss": 0.01593680866062641,
          "validation_loss": 0.013017169199883938,
          "train_metric": 0.001605096971616149,
          "validation_metric": 0.001311041647568345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9613107440000022
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.012038450367981568,
          "train_loss": 0.004420083947479725,
          "validation_loss": 0.002657263306900859,
          "train_metric": 0.00044517460628412664,
          "validation_metric": 0.00026762980269268155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9751894660003018
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.007768053750623949,
          "train_loss": 0.007285573054105043,
          "validation_loss": 0.006151596549898386,
          "train_metric": 0.0007337763090617955,
          "validation_metric": 0.0006195663008838892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9886091549997218
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.007226608126075007,
          "train_loss": 0.007201131898909807,
          "validation_loss": 0.005738956853747368,
          "train_metric": 0.0007252716459333897,
          "validation_metric": 0.0005780067294836044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0025312279994978
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.004565104773064377,
          "train_loss": 0.003592194290831685,
          "validation_loss": 0.0032982334960252047,
          "train_metric": 0.0003617926558945328,
          "validation_metric": 0.0003321859403513372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0163893259996257
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0043839493082487024,
          "train_loss": 0.004496533423662186,
          "validation_loss": 0.0034993495792150497,
          "train_metric": 0.000452874373877421,
          "validation_metric": 0.00035244168248027563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0303747420002765
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.002848995183740044,
          "train_loss": 0.0023259848821908236,
          "validation_loss": 0.0018966331845149398,
          "train_metric": 0.0002342646475881338,
          "validation_metric": 0.00019102194346487522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.050508242999058
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0038086834028945304,
          "train_loss": 0.013368995860219002,
          "validation_loss": 0.01452961191534996,
          "train_metric": 0.0013464763760566711,
          "validation_metric": 0.0014633694663643837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0685229030004848
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.00995112769305706,
          "train_loss": 0.014141579158604145,
          "validation_loss": 0.012196009978652,
          "train_metric": 0.001424288027919829,
          "validation_metric": 0.0012283374089747667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.084860081999068
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.006706159314489923,
          "train_loss": 0.007830265909433365,
          "validation_loss": 0.006172288674861193,
          "train_metric": 0.0007886356906965375,
          "validation_metric": 0.0006216503097675741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1012364779999189
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.008659465456730686,
          "train_loss": 0.012232339009642601,
          "validation_loss": 0.008769161067903042,
          "train_metric": 0.0012319964589551091,
          "validation_metric": 0.0008831977611407638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1195408049989055
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.010430926064145751,
          "train_loss": 0.01546911709010601,
          "validation_loss": 0.012524712830781937,
          "train_metric": 0.001557992771267891,
          "validation_metric": 0.0012614431325346231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1472476969993295
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.01052079837245401,
          "train_loss": 0.005902647506445646,
          "validation_loss": 0.005649357568472624,
          "train_metric": 0.0005944930599071085,
          "validation_metric": 0.0005689826211892068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1694248000003427
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.005875552931684069,
          "train_loss": 0.0038188451435416937,
          "validation_loss": 0.0035810552071779966,
          "train_metric": 0.0003846200997941196,
          "validation_metric": 0.00036067076143808663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1989689229994838
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.005087846257083584,
          "train_loss": 0.0038356517907232046,
          "validation_loss": 0.002553624100983143,
          "train_metric": 0.00038631274946965277,
          "validation_metric": 0.00025719168479554355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2205741069992655
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0080159234785242,
          "train_loss": 0.005641886964440346,
          "validation_loss": 0.0037425935734063387,
          "train_metric": 0.0005682301707565784,
          "validation_metric": 0.000376940326532349,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3828190980002546
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005474936231621541,
          "train_loss": 0.0049967411905527115,
          "validation_loss": 0.003871488617733121,
          "train_metric": 0.0005032534245401621,
          "validation_metric": 0.00038992217741906643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.397316778999084
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.006959793754504062,
          "train_loss": 0.007020993158221245,
          "validation_loss": 0.007475012913346291,
          "train_metric": 0.0007071287836879492,
          "validation_metric": 0.0007528559071943164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4112174259989843
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.010578843735856935,
          "train_loss": 0.011548236943781376,
          "validation_loss": 0.010483461432158947,
          "train_metric": 0.0011630962835624814,
          "validation_metric": 0.001055855886079371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4247839549989294
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.006364221655530855,
          "train_loss": 0.006330887787044048,
          "validation_loss": 0.006707367021590471,
          "train_metric": 0.0006376238306984305,
          "validation_metric": 0.0006755414651706815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4387072850004188
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.004168871259025764,
          "train_loss": 0.00514680240303278,
          "validation_loss": 0.0030304265674203634,
          "train_metric": 0.0005183671019040048,
          "validation_metric": 0.00030521347071044147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.452474962999986
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005112245904456358,
          "train_loss": 0.002527973148971796,
          "validation_loss": 0.0017807470867410302,
          "train_metric": 0.00025460822507739067,
          "validation_metric": 0.00017935031792148948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.467904223000005
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0025165708429994993,
          "train_loss": 0.002781960181891918,
          "validation_loss": 0.003041807096451521,
          "train_metric": 0.0002801888040266931,
          "validation_metric": 0.000306359666865319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4825568619999103
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.00491555334883742,
          "train_loss": 0.003914761822670698,
          "validation_loss": 0.002786326454952359,
          "train_metric": 0.00039428044692613184,
          "validation_metric": 0.00028062862111255527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4965185660003044
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.002354760479647666,
          "train_loss": 0.0017621538136154413,
          "validation_loss": 0.0014821243239566684,
          "train_metric": 0.00017747767560649663,
          "validation_metric": 0.0001492741284891963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5112570649998815
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0025322262845293153,
          "train_loss": 0.0028939952608197927,
          "validation_loss": 0.002731239888817072,
          "train_metric": 0.0002914725919254124,
          "validation_metric": 0.0002750804997049272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5258160199991835
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0018616534362081438,
          "train_loss": 0.0011780669447034597,
          "validation_loss": 0.0008902671979740262,
          "train_metric": 0.00011865058331750333,
          "validation_metric": 8.966443419922143e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.555554074999236
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0015732529700471787,
          "train_loss": 0.0013709714403375983,
          "validation_loss": 0.0012966964859515429,
          "train_metric": 0.00013807922368869185,
          "validation_metric": 0.00013059853517916054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5841273030000593
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.002125462833646452,
          "train_loss": 0.007780477870255709,
          "validation_loss": 0.0046353209763765335,
          "train_metric": 0.0007836212171241641,
          "validation_metric": 0.0004668526235036552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6045991119990504
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0037950047189951874,
          "train_loss": 0.005985003896057606,
          "validation_loss": 0.003948509693145752,
          "train_metric": 0.0006027877680025995,
          "validation_metric": 0.0003976793959736824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6265430580006068
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0019286813076178078,
          "train_loss": 0.0015709116123616695,
          "validation_loss": 0.0017604497261345387,
          "train_metric": 0.00015821645502001047,
          "validation_metric": 0.00017730605031829327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6426906689994212
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.001973107788217021,
          "train_loss": 0.005932911764830351,
          "validation_loss": 0.004535900428891182,
          "train_metric": 0.0005975410458631814,
          "validation_metric": 0.0004568392760120332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7314136940003664
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.007166496441641357,
          "train_loss": 0.012348899617791176,
          "validation_loss": 0.008688892237842083,
          "train_metric": 0.001243735896423459,
          "validation_metric": 0.0008751134155318141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.768680162000237
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.007633354893187061,
          "train_loss": 0.0062506855465471745,
          "validation_loss": 0.004773764871060848,
          "train_metric": 0.0006295461789704859,
          "validation_metric": 0.0004807961522601545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.845512771000358
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.003695114850415848,
          "train_loss": 0.004371763207018375,
          "validation_loss": 0.003149190219119191,
          "train_metric": 0.00044030798017047346,
          "validation_metric": 0.0003171748830936849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9238305859998945
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0031724011569167487,
          "train_loss": 0.004557460080832243,
          "validation_loss": 0.003674623090773821,
          "train_metric": 0.0004590106545947492,
          "validation_metric": 0.0003700945235323161,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9999019209990365
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.005244883639534237,
          "train_loss": 0.0026669742073863745,
          "validation_loss": 0.002124050399288535,
          "train_metric": 0.000268607895122841,
          "validation_metric": 0.00021392657072283328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0770858169998974
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0025669708084024023,
          "train_loss": 0.005898982286453247,
          "validation_loss": 0.005176703445613384,
          "train_metric": 0.0005941238487139344,
          "validation_metric": 0.0005213785916566849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1542597869993187
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.003071530576562509,
          "train_loss": 0.0016834059497341514,
          "validation_loss": 0.0014910846948623657,
          "train_metric": 0.00016954648890532553,
          "validation_metric": 0.0001501765800639987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.229562686999998
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0030939307871449273,
          "train_loss": 0.0016892739804461598,
          "validation_loss": 0.0013086769031360745,
          "train_metric": 0.00017013750039041042,
          "validation_metric": 0.00013180515088606626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.303911341999992
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0033565580670256168,
          "train_loss": 0.002626169938594103,
          "validation_loss": 0.0025459269527345896,
          "train_metric": 0.000264498230535537,
          "validation_metric": 0.00025641644606366754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.381455050999648
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.005015744871343486,
          "train_loss": 0.008482285775244236,
          "validation_loss": 0.009002014994621277,
          "train_metric": 0.0008543047588318586,
          "validation_metric": 0.0009066499187611043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.456920425000135
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.004886253111180849,
          "train_loss": 0.002289703581482172,
          "validation_loss": 0.002145282458513975,
          "train_metric": 0.00023061058891471475,
          "validation_metric": 0.00021606500376947224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5363825539989193
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0025834226871666033,
          "train_loss": 0.006932929158210754,
          "validation_loss": 0.005609591957181692,
          "train_metric": 0.0006982592167332768,
          "validation_metric": 0.0005649775848723948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.611879832998966
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.005086877470603213,
          "train_loss": 0.003777578938752413,
          "validation_loss": 0.004498353693634272,
          "train_metric": 0.00038046392728574574,
          "validation_metric": 0.00045305766980163753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6882954520006024
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.00354289139067987,
          "train_loss": 0.002508274745196104,
          "validation_loss": 0.0019564232788980007,
          "train_metric": 0.0002526242460589856,
          "validation_metric": 0.00019704378792084754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7613448390002304
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.0016869029022927862,
          "train_loss": 0.001545896870084107,
          "validation_loss": 0.0014121183194220066,
          "train_metric": 0.00015569705283269286,
          "validation_metric": 0.00014222337631508708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8444559889994707
        }
      ],
      "best": {
        "epoch": 91,
        "updates": 1456
      },
      "best_validation_metric": 8.966443419922143e-05,
      "config": {
        "label": "LayerNorm + tanh + constant LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009469130593630096
    },
    "ff9112ae3885_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3071288261562586,
          "train_loss": 0.11203804612159729,
          "validation_loss": 0.07620196044445038,
          "train_metric": 0.011284060776233673,
          "validation_metric": 0.007674782536923885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01797514000099909
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.074585021706298,
          "train_loss": 0.05086174234747887,
          "validation_loss": 0.04995014891028404,
          "train_metric": 0.005122608505189419,
          "validation_metric": 0.005030796863138676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.032244421001450974
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.04231863818131387,
          "train_loss": 0.032734062522649765,
          "validation_loss": 0.024340001866221428,
          "train_metric": 0.0032968551386147738,
          "validation_metric": 0.0024514361284673214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.046584416000769124
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.028199329390190542,
          "train_loss": 0.02129916287958622,
          "validation_loss": 0.020750556141138077,
          "train_metric": 0.0021451739594340324,
          "validation_metric": 0.0020899202208966017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06106502500006172
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.02146986947627738,
          "train_loss": 0.02125341072678566,
          "validation_loss": 0.033918291330337524,
          "train_metric": 0.002140565775334835,
          "validation_metric": 0.0034161261282861233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07465022200085514
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.019407856394536793,
          "train_loss": 0.012668851763010025,
          "validation_loss": 0.009883912280201912,
          "train_metric": 0.001275960123166442,
          "validation_metric": 0.0009954714914783835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08843468600025517
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.024012377456529066,
          "train_loss": 0.02390378527343273,
          "validation_loss": 0.021115930750966072,
          "train_metric": 0.002407501917332411,
          "validation_metric": 0.0021267193369567394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10213265000129468
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.025285185838583857,
          "train_loss": 0.01872871071100235,
          "validation_loss": 0.012351857498288155,
          "train_metric": 0.0018862871220335364,
          "validation_metric": 0.001244033919647336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11564734000057797
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.012040380737744272,
          "train_loss": 0.011232025921344757,
          "validation_loss": 0.006161433178931475,
          "train_metric": 0.0011312485439702868,
          "validation_metric": 0.0006205569952726364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13003625700002885
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.009275261327275075,
          "train_loss": 0.009033228270709515,
          "validation_loss": 0.006427751388400793,
          "train_metric": 0.0009097935981117189,
          "validation_metric": 0.0006473796092905104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14389230500091799
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.011373822577297688,
          "train_loss": 0.0066519249230623245,
          "validation_loss": 0.005146905779838562,
          "train_metric": 0.0006699575460515916,
          "validation_metric": 0.0005183775210753083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15813882800102874
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.008848477293213364,
          "train_loss": 0.016483120620250702,
          "validation_loss": 0.012250147759914398,
          "train_metric": 0.0016601196257397532,
          "validation_metric": 0.0012337899534031749,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1715578520015697
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.01770462401327677,
          "train_loss": 0.0076902518048882484,
          "validation_loss": 0.0066915531642735004,
          "train_metric": 0.0007745339535176754,
          "validation_metric": 0.0006739486125297844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18473612400157435
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.01383943113614805,
          "train_loss": 0.007693617604672909,
          "validation_loss": 0.008185774087905884,
          "train_metric": 0.0007748730131424963,
          "validation_metric": 0.0008244412601925433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1981585690009524
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.013440196213196032,
          "train_loss": 0.020549535751342773,
          "validation_loss": 0.015472847037017345,
          "train_metric": 0.002069673966616392,
          "validation_metric": 0.00155836867634207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21153606900043087
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.015208673314191401,
          "train_loss": 0.02309454418718815,
          "validation_loss": 0.015796467661857605,
          "train_metric": 0.002325997920706868,
          "validation_metric": 0.0015909625217318535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22474355600024865
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.01357882475713268,
          "train_loss": 0.024356551468372345,
          "validation_loss": 0.019021105021238327,
          "train_metric": 0.002453102497383952,
          "validation_metric": 0.0019157358910888433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23801686000115296
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.01771878165891394,
          "train_loss": 0.016155371442437172,
          "validation_loss": 0.01659294404089451,
          "train_metric": 0.0016271100612357259,
          "validation_metric": 0.0016711805947124958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2508024350008782
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.01291148022573907,
          "train_loss": 0.009475353173911572,
          "validation_loss": 0.00587684940546751,
          "train_metric": 0.0009543228079564869,
          "validation_metric": 0.0005918947863392532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2641445400004159
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.009440544323297217,
          "train_loss": 0.012063981965184212,
          "validation_loss": 0.010029615834355354,
          "train_metric": 0.001215039985254407,
          "validation_metric": 0.001010146108455956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2772463239998615
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.006840501737315208,
          "train_loss": 0.006719559896737337,
          "validation_loss": 0.0064447373151779175,
          "train_metric": 0.0006767694721929729,
          "validation_metric": 0.0006490902742370963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29058244800035027
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.006070428840757813,
          "train_loss": 0.004238975699990988,
          "validation_loss": 0.0047693317756056786,
          "train_metric": 0.00042693412979133427,
          "validation_metric": 0.0004803496412932873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38024680400121724
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.009627404615457635,
          "train_loss": 0.009205630980432034,
          "validation_loss": 0.009717017412185669,
          "train_metric": 0.0009271574672311544,
          "validation_metric": 0.0009786623995751143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39407289500013576
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.014047909789951518,
          "train_loss": 0.013953428715467453,
          "validation_loss": 0.015782037749886513,
          "train_metric": 0.001405338174663484,
          "validation_metric": 0.0015895090764388442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.407140478000656
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.010443313098221552,
          "train_loss": 0.006825929507613182,
          "validation_loss": 0.006799179129302502,
          "train_metric": 0.0006874826503917575,
          "validation_metric": 0.0006847883923910558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4230557980008598
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.011437548062531278,
          "train_loss": 0.011943532153964043,
          "validation_loss": 0.008627168834209442,
          "train_metric": 0.0012029088102281094,
          "validation_metric": 0.0008688968955539167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43609302600088995
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.006769984611310065,
          "train_loss": 0.004456540569663048,
          "validation_loss": 0.004082114901393652,
          "train_metric": 0.00044884643284603953,
          "validation_metric": 0.0004111356975045055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45009932200082403
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01576909028517548,
          "train_loss": 0.01666870154440403,
          "validation_loss": 0.014270413666963577,
          "train_metric": 0.0016788106877356768,
          "validation_metric": 0.001437263679690659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46352306400149246
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.014986050591687672,
          "train_loss": 0.01137054618448019,
          "validation_loss": 0.009255947545170784,
          "train_metric": 0.0011451997561380267,
          "validation_metric": 0.0009322250261902809,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4766336450011295
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.013578396625234745,
          "train_loss": 0.006296770181506872,
          "validation_loss": 0.0036617377772927284,
          "train_metric": 0.0006341876578517258,
          "validation_metric": 0.00036879684193991125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.490517334001197
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.009990488950279541,
          "train_loss": 0.019505172967910767,
          "validation_loss": 0.015952683985233307,
          "train_metric": 0.001964489696547389,
          "validation_metric": 0.001606695936061442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5038625680008408
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01945007365429774,
          "train_loss": 0.009888243861496449,
          "validation_loss": 0.006357372738420963,
          "train_metric": 0.0009959075832739472,
          "validation_metric": 0.000640291313175112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5172224990001268
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.010607568343402818,
          "train_loss": 0.013945555314421654,
          "validation_loss": 0.012985952198505402,
          "train_metric": 0.0014045452699065208,
          "validation_metric": 0.0013078978518024087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5307944380001572
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.011888013665156905,
          "train_loss": 0.007090152706950903,
          "validation_loss": 0.006639047991484404,
          "train_metric": 0.0007140942034311593,
          "validation_metric": 0.0006686606211587787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5443901940016076
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.0071046433949959464,
          "train_loss": 0.004070295486599207,
          "validation_loss": 0.0030450946651399136,
          "train_metric": 0.0004099452344235033,
          "validation_metric": 0.0003066907811444253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5581630169999698
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.008023191941902041,
          "train_loss": 0.004413176327943802,
          "validation_loss": 0.0034315665252506733,
          "train_metric": 0.00044447893742471933,
          "validation_metric": 0.00034561482607387006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5715105039998889
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.006907705130288377,
          "train_loss": 0.005530179478228092,
          "validation_loss": 0.003692805767059326,
          "train_metric": 0.000556979444809258,
          "validation_metric": 0.0003719258529599756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.585188298000503
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.0040482903641532175,
          "train_loss": 0.002839048858731985,
          "validation_loss": 0.0027677796315401793,
          "train_metric": 0.0002859385858755559,
          "validation_metric": 0.00027876062085852027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5992943190012738
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01328999012184795,
          "train_loss": 0.035411953926086426,
          "validation_loss": 0.02749413251876831,
          "train_metric": 0.0035665626637637615,
          "validation_metric": 0.0027691081631928682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6128472870004771
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.022517298988532275,
          "train_loss": 0.011345796287059784,
          "validation_loss": 0.013017391785979271,
          "train_metric": 0.0011427070712670684,
          "validation_metric": 0.0013110639993101358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6261112680003862
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.008080109604634345,
          "train_loss": 0.00588337192311883,
          "validation_loss": 0.004020294640213251,
          "train_metric": 0.0005925517180003226,
          "validation_metric": 0.00040490939863957465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6395866729999398
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.005963767820503563,
          "train_loss": 0.004166283644735813,
          "validation_loss": 0.003553282469511032,
          "train_metric": 0.0004196127993054688,
          "validation_metric": 0.0003578736213967204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6530116419999104
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009335822789580561,
          "train_loss": 0.01667610928416252,
          "validation_loss": 0.014476744458079338,
          "train_metric": 0.0016795566771179438,
          "validation_metric": 0.0014580447459593415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6661675900013506
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01296637166524306,
          "train_loss": 0.008167150430381298,
          "validation_loss": 0.007495296187698841,
          "train_metric": 0.0008225654601119459,
          "validation_metric": 0.0007548988214693964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6795432770013576
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.006371240058797412,
          "train_loss": 0.007321102544665337,
          "validation_loss": 0.007356789894402027,
          "train_metric": 0.0007373546250164509,
          "validation_metric": 0.0007409490062855184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6928575020010612
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.005271240741421934,
          "train_loss": 0.004778060130774975,
          "validation_loss": 0.005544597283005714,
          "train_metric": 0.00048122869338840246,
          "validation_metric": 0.0005584315513260663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.705939532001139
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.00498176013934426,
          "train_loss": 0.002443666337057948,
          "validation_loss": 0.0022033751010894775,
          "train_metric": 0.00024611709523014724,
          "validation_metric": 0.00022191590687725693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7200790150000103
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0035049753241764847,
          "train_loss": 0.004942765925079584,
          "validation_loss": 0.004673020448535681,
          "train_metric": 0.0004978172364644706,
          "validation_metric": 0.0004706494801212102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7333482310004911
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.004495720146223903,
          "train_loss": 0.004035751335322857,
          "validation_loss": 0.0031157019548118114,
          "train_metric": 0.0004064661043230444,
          "validation_metric": 0.00031380209838971496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7470562520011299
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.004526940640062094,
          "train_loss": 0.003896507667377591,
          "validation_loss": 0.003825979772955179,
          "train_metric": 0.00039244198706001043,
          "validation_metric": 0.0003853386442642659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7602994440003386
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.005048604900366627,
          "train_loss": 0.008761128410696983,
          "validation_loss": 0.00579631794244051,
          "train_metric": 0.0008823888492770493,
          "validation_metric": 0.0005837838398292661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.773471745000279
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008277216817077715,
          "train_loss": 0.005358617752790451,
          "validation_loss": 0.0038503266405314207,
          "train_metric": 0.0005397003842517734,
          "validation_metric": 0.00038779081660322845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.786674105000202
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008937666658312082,
          "train_loss": 0.004297289066016674,
          "validation_loss": 0.003845198079943657,
          "train_metric": 0.0004328072536736727,
          "validation_metric": 0.00038727428182028234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8001196580007672
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.004864150418143254,
          "train_loss": 0.003194466233253479,
          "validation_loss": 0.003566374769434333,
          "train_metric": 0.00032173492945730686,
          "validation_metric": 0.0003591921995393932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.813240522000342
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.004782680254720617,
          "train_loss": 0.004385479260236025,
          "validation_loss": 0.004400455392897129,
          "train_metric": 0.00044168936437927186,
          "validation_metric": 0.000443197728600353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8268251170011354
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.003914387518307194,
          "train_loss": 0.0026528919115662575,
          "validation_loss": 0.002810490084812045,
          "train_metric": 0.00026718954904936254,
          "validation_metric": 0.0002830622543115169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8401788530009071
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.004540577207080787,
          "train_loss": 0.002056040335446596,
          "validation_loss": 0.0017707201186567545,
          "train_metric": 0.00020707685325760394,
          "validation_metric": 0.00017834047321230173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8543908730007388
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.018147993556340225,
          "train_loss": 0.02214161306619644,
          "validation_loss": 0.015251225791871548,
          "train_metric": 0.0022300221025943756,
          "validation_metric": 0.001536047668196261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9466027880007459
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.01538866848568432,
          "train_loss": 0.012630714103579521,
          "validation_loss": 0.01062769629061222,
          "train_metric": 0.001272119116038084,
          "validation_metric": 0.0010703826555982232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9599581560014485
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0071873173146741465,
          "train_loss": 0.006293626967817545,
          "validation_loss": 0.0052931103855371475,
          "train_metric": 0.0006338709499686956,
          "validation_metric": 0.0005331026623025537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9731206360011129
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004570935227093287,
          "train_loss": 0.00463627465069294,
          "validation_loss": 0.0032044495455920696,
          "train_metric": 0.00046694857883267105,
          "validation_metric": 0.0003227404085919261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.986409643001025
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0039688462420599535,
          "train_loss": 0.006133649498224258,
          "validation_loss": 0.0034299264661967754,
          "train_metric": 0.0006177587201818824,
          "validation_metric": 0.0003454496036283672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.999400055001388
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.010329144039133098,
          "train_loss": 0.009874947369098663,
          "validation_loss": 0.007222643122076988,
          "train_metric": 0.0009945685742422938,
          "validation_metric": 0.0007274382514879107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.012267990001419
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007473938385373913,
          "train_loss": 0.003380092326551676,
          "validation_loss": 0.002962950151413679,
          "train_metric": 0.00034043053165078163,
          "validation_metric": 0.00029841752257198095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0254101080008695
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.007203828259662259,
          "train_loss": 0.00945940613746643,
          "validation_loss": 0.006850644946098328,
          "train_metric": 0.0009527167421765625,
          "validation_metric": 0.000689971842803061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0386222940014704
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.006102527371695032,
          "train_loss": 0.004661147017031908,
          "validation_loss": 0.0043709771707654,
          "train_metric": 0.00046945360372774303,
          "validation_metric": 0.0004402288468554616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.051978372001031
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0037031408755865414,
          "train_loss": 0.0022908838000148535,
          "validation_loss": 0.0015214508166536689,
          "train_metric": 0.00023072943440638483,
          "validation_metric": 0.0001532349269837141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0659499050016166
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0028230119569343515,
          "train_loss": 0.0017397275660187006,
          "validation_loss": 0.002219205955043435,
          "train_metric": 0.00017521898553241044,
          "validation_metric": 0.0002235103165730834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0791559640001651
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.008306211986564449,
          "train_loss": 0.0058769723400473595,
          "validation_loss": 0.004003970883786678,
          "train_metric": 0.0005919070681557059,
          "validation_metric": 0.00040326532325707376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0946734839999408
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.010768441410618834,
          "train_loss": 0.005165383219718933,
          "validation_loss": 0.004679038655012846,
          "train_metric": 0.0005202384199947119,
          "validation_metric": 0.0004712556255981326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1077004020007735
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0041675820320961066,
          "train_loss": 0.0027609083335846663,
          "validation_loss": 0.00321635277941823,
          "train_metric": 0.0002780685608740896,
          "validation_metric": 0.00032393925357609987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1211152040013985
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.005793672764411895,
          "train_loss": 0.009301627986133099,
          "validation_loss": 0.005919034592807293,
          "train_metric": 0.0009368258761242032,
          "validation_metric": 0.0005961434217169881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1345165260008798
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.01884086456266232,
          "train_loss": 0.008877714164555073,
          "validation_loss": 0.010837715119123459,
          "train_metric": 0.0008941309060901403,
          "validation_metric": 0.0010915349703282118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1477790330009157
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.008154400333296508,
          "train_loss": 0.0052547999657690525,
          "validation_loss": 0.006944365333765745,
          "train_metric": 0.0005292442510835826,
          "validation_metric": 0.0006994110881350935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1607410740016348
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006463957513915375,
          "train_loss": 0.014513191767036915,
          "validation_loss": 0.014055795967578888,
          "train_metric": 0.0014617154374718666,
          "validation_metric": 0.0014156483812257648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.17420985300123
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.01248252679943107,
          "train_loss": 0.022043246775865555,
          "validation_loss": 0.021901432424783707,
          "train_metric": 0.0022201151587069035,
          "validation_metric": 0.002205832162871957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.187418115001492
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.014458285921136849,
          "train_loss": 0.007923133671283722,
          "validation_loss": 0.008137684315443039,
          "train_metric": 0.0007979889633134007,
          "validation_metric": 0.0008195977425202727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2008031020013732
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006276465603150427,
          "train_loss": 0.004931956063956022,
          "validation_loss": 0.0049920761957764626,
          "train_metric": 0.000496728578582406,
          "validation_metric": 0.0005027836887165904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.213880434001112
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.007936062829685397,
          "train_loss": 0.006534598767757416,
          "validation_loss": 0.005528176669031382,
          "train_metric": 0.0006581409252248704,
          "validation_metric": 0.0005567776970565319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2272465170008218
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.004230338196066441,
          "train_loss": 0.0021906602196395397,
          "validation_loss": 0.0022253997158259153,
          "train_metric": 0.00022063528012949973,
          "validation_metric": 0.00022413412807509303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2401487050010473
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.004086529286723817,
          "train_loss": 0.007844879291951656,
          "validation_loss": 0.006103771273046732,
          "train_metric": 0.0007901075296103954,
          "validation_metric": 0.0006147495587356389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2534552450015326
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.010943995686830021,
          "train_loss": 0.02295665442943573,
          "validation_loss": 0.018512284383177757,
          "train_metric": 0.002312110038474202,
          "validation_metric": 0.001864489633589983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2668868440014194
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.017220832727616653,
          "train_loss": 0.01076485961675644,
          "validation_loss": 0.007450628560036421,
          "train_metric": 0.00108419731259346,
          "validation_metric": 0.0007504000677727163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2803273870013072
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.007787665133946575,
          "train_loss": 0.005587470252066851,
          "validation_loss": 0.006476185750216246,
          "train_metric": 0.0005627496284432709,
          "validation_metric": 0.0006522577023133636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.293606268000076
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0053450265259016305,
          "train_loss": 0.005946928635239601,
          "validation_loss": 0.005091794300824404,
          "train_metric": 0.0005989529308862984,
          "validation_metric": 0.0005128269549459219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3070785580002848
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.0061574064020533115,
          "train_loss": 0.003973843529820442,
          "validation_loss": 0.0032734577544033527,
          "train_metric": 0.0004002310161013156,
          "validation_metric": 0.0003296906943432987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.320556619000854
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0031939352556946687,
          "train_loss": 0.0022564323153346777,
          "validation_loss": 0.0015776599757373333,
          "train_metric": 0.00022725961753167212,
          "validation_metric": 0.00015889614587649703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3338318330006587
        }
      ],
      "best": {
        "epoch": 67,
        "updates": 1072
      },
      "best_validation_metric": 0.0001532349269837141,
      "config": {
        "label": "LayerNorm + relu + constant LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012378809594775829
    },
    "7a196ed2cdbd_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3594899591989815,
          "train_loss": 0.10689476132392883,
          "validation_loss": 0.10651864856481552,
          "train_metric": 0.010766048915684223,
          "validation_metric": 0.010728168301284313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03154845300014131
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.08005768503062427,
          "train_loss": 0.05876293405890465,
          "validation_loss": 0.06888415664434433,
          "train_metric": 0.0059183877892792225,
          "validation_metric": 0.006937759928405285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04783711399977619
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.047159415553323925,
          "train_loss": 0.04439746215939522,
          "validation_loss": 0.04207828640937805,
          "train_metric": 0.004471549764275551,
          "validation_metric": 0.004237971268594265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06558077999943634
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.03627265279646963,
          "train_loss": 0.036621857434511185,
          "validation_loss": 0.0362851545214653,
          "train_metric": 0.0036884187720716,
          "validation_metric": 0.003654508152976632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0831082749991765
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.02609884802950546,
          "train_loss": 0.02378969080746174,
          "validation_loss": 0.019558057188987732,
          "train_metric": 0.0023960103280842304,
          "validation_metric": 0.0019698156975209713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10533328800011077
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.020956436870619655,
          "train_loss": 0.027749165892601013,
          "validation_loss": 0.02930276282131672,
          "train_metric": 0.002794794039800763,
          "validation_metric": 0.0029512671753764153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13842030399973737
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.017646923603024334,
          "train_loss": 0.0202290341258049,
          "validation_loss": 0.020458824932575226,
          "train_metric": 0.0020373943261802197,
          "validation_metric": 0.002060537924990058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17342509199988854
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.013338577758986503,
          "train_loss": 0.02697916328907013,
          "validation_loss": 0.02338114194571972,
          "train_metric": 0.0027172425761818886,
          "validation_metric": 0.0023548630997538567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2065049269986048
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.03040843659255188,
          "train_loss": 0.02058366686105728,
          "validation_loss": 0.021947549656033516,
          "train_metric": 0.002073111478239298,
          "validation_metric": 0.002210476668551564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2257348479997745
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.02013958990573883,
          "train_loss": 0.014335322193801403,
          "validation_loss": 0.011812934651970863,
          "train_metric": 0.0014438010985031724,
          "validation_metric": 0.0011897553922608495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2511611769987212
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.02076608722563833,
          "train_loss": 0.014632223173975945,
          "validation_loss": 0.018382254987955093,
          "train_metric": 0.0014737040037289262,
          "validation_metric": 0.0018513933755457401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26659763799943903
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.01418373743945267,
          "train_loss": 0.00860548485070467,
          "validation_loss": 0.010486365295946598,
          "train_metric": 0.0008667129441164434,
          "validation_metric": 0.0010561482049524784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2844496799989429
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.0173728795198258,
          "train_loss": 0.014441790990531445,
          "validation_loss": 0.011986266821622849,
          "train_metric": 0.0014545242302119732,
          "validation_metric": 0.0012072130339220166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30008942499989644
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.014862617274047807,
          "train_loss": 0.0154472840949893,
          "validation_loss": 0.010748198255896568,
          "train_metric": 0.001555793802253902,
          "validation_metric": 0.0010825189528986812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3167895279984805
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.025403889041626826,
          "train_loss": 0.041615959256887436,
          "validation_loss": 0.03319608047604561,
          "train_metric": 0.00419140700250864,
          "validation_metric": 0.003343387506902218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33223549499962246
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.021664694766514003,
          "train_loss": 0.01344259362667799,
          "validation_loss": 0.01134704053401947,
          "train_metric": 0.001353888656012714,
          "validation_metric": 0.0011428323341533542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34774372199899517
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.014402267828700133,
          "train_loss": 0.01665353961288929,
          "validation_loss": 0.022806476801633835,
          "train_metric": 0.0016772834351286292,
          "validation_metric": 0.0022969848942011595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36340271799963375
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.013661834585946053,
          "train_loss": 0.009685010649263859,
          "validation_loss": 0.011342911049723625,
          "train_metric": 0.0009754388011060655,
          "validation_metric": 0.0011424164986237884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37913224099975196
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.015104164893273264,
          "train_loss": 0.010758163407444954,
          "validation_loss": 0.009069183841347694,
          "train_metric": 0.0010835228022187948,
          "validation_metric": 0.0009134150459431112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39496929599954456
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.009322157100541517,
          "train_loss": 0.009018831886351109,
          "validation_loss": 0.0059044333174824715,
          "train_metric": 0.0009083438199013472,
          "validation_metric": 0.0005946729215793312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4127001669985475
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.006748501502443105,
          "train_loss": 0.005163723137229681,
          "validation_loss": 0.012859508395195007,
          "train_metric": 0.0005200713058002293,
          "validation_metric": 0.0012951624812558293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4278299279994826
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.012611139762157109,
          "train_loss": 0.007669901009649038,
          "validation_loss": 0.007750188000500202,
          "train_metric": 0.0007724842871539295,
          "validation_metric": 0.0007805704954080284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4430303319986706
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01289312489097938,
          "train_loss": 0.010503493249416351,
          "validation_loss": 0.02323693037033081,
          "train_metric": 0.00105787324719131,
          "validation_metric": 0.0023403384257107973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4591382029993838
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.019172809479641728,
          "train_loss": 0.012283912859857082,
          "validation_loss": 0.01608487218618393,
          "train_metric": 0.0012371905613690615,
          "validation_metric": 0.0016200094250962138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47456527999929676
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.010826966870808974,
          "train_loss": 0.009194883517920971,
          "validation_loss": 0.009451521560549736,
          "train_metric": 0.0009260750375688076,
          "validation_metric": 0.0009519225568510592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5755919749990426
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.008604193280916661,
          "train_loss": 0.005289078690111637,
          "validation_loss": 0.007678186986595392,
          "train_metric": 0.0005326967220753431,
          "validation_metric": 0.0007733188685961068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5913289089985483
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01894094643648714,
          "train_loss": 0.02910364232957363,
          "validation_loss": 0.04844817891716957,
          "train_metric": 0.0029312120750546455,
          "validation_metric": 0.004879523068666458,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6065485669987538
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.020172985445242375,
          "train_loss": 0.013426442630589008,
          "validation_loss": 0.014586572535336018,
          "train_metric": 0.001352262101136148,
          "validation_metric": 0.0014691061805933714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.622258494999187
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.012159680161857978,
          "train_loss": 0.010088502429425716,
          "validation_loss": 0.016597436740994453,
          "train_metric": 0.001016077003441751,
          "validation_metric": 0.0016716329846531153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6383964589986135
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.008545430930098519,
          "train_loss": 0.005742490291595459,
          "validation_loss": 0.012234696187078953,
          "train_metric": 0.0005783626111224294,
          "validation_metric": 0.0012322337133809924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6540665550000995
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.01068879006197676,
          "train_loss": 0.030793415382504463,
          "validation_loss": 0.04476664587855339,
          "train_metric": 0.003101400099694729,
          "validation_metric": 0.004508732818067074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6690088949999335
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01573269465006888,
          "train_loss": 0.00725203100591898,
          "validation_loss": 0.008648212067782879,
          "train_metric": 0.0007303980528376997,
          "validation_metric": 0.0008710161782801151,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.683810287999222
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.007807671136106364,
          "train_loss": 0.004955038893967867,
          "validation_loss": 0.014485729858279228,
          "train_metric": 0.000499053334351629,
          "validation_metric": 0.0014589495258405805,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.698268938000183
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.005399884026701329,
          "train_loss": 0.005224285181611776,
          "validation_loss": 0.014489397406578064,
          "train_metric": 0.0005261709447950125,
          "validation_metric": 0.001459319144487381,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.7135649109986844
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.004446482009370811,
          "train_loss": 0.0038121750112622976,
          "validation_loss": 0.009451248683035374,
          "train_metric": 0.00038394832517951727,
          "validation_metric": 0.000951895082835108,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.7271991699999489
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.009215873302309774,
          "train_loss": 0.014284037984907627,
          "validation_loss": 0.02633497305214405,
          "train_metric": 0.001438635983504355,
          "validation_metric": 0.0026523617561906576,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.7414090189995477
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.006258717046875972,
          "train_loss": 0.004671803675591946,
          "validation_loss": 0.009857463650405407,
          "train_metric": 0.00047052695299498737,
          "validation_metric": 0.0009928077924996614,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.7554487139987032
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.005093592808407266,
          "train_loss": 0.0033986996859312057,
          "validation_loss": 0.013448094949126244,
          "train_metric": 0.00034230458550155163,
          "validation_metric": 0.001354442909359932,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.7698852519988577
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.005586413637502119,
          "train_loss": 0.003398810513317585,
          "validation_loss": 0.014533723704516888,
          "train_metric": 0.00034231579047627747,
          "validation_metric": 0.001463783672079444,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.7857615619996068
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.005749243820901029,
          "train_loss": 0.004765740130096674,
          "validation_loss": 0.0053887092508375645,
          "train_metric": 0.00047998782247304916,
          "validation_metric": 0.000542731024324894,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8013502169997082
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.006106412933149841,
          "train_loss": 0.002671936061233282,
          "validation_loss": 0.010833938606083393,
          "train_metric": 0.0002691076078917831,
          "validation_metric": 0.001091154757887125,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8148341989999608
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.0036993084468122106,
          "train_loss": 0.002380888909101486,
          "validation_loss": 0.00993440579622984,
          "train_metric": 0.00023979440447874367,
          "validation_metric": 0.0010005569783970714,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8287963289985782
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.0026995613807230256,
          "train_loss": 0.002683021128177643,
          "validation_loss": 0.012494150549173355,
          "train_metric": 0.0002702240599319339,
          "validation_metric": 0.0012583649950101972,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8429483479994815
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.002449960451485822,
          "train_loss": 0.002062493236735463,
          "validation_loss": 0.011542617343366146,
          "train_metric": 0.00020772677089553326,
          "validation_metric": 0.0011625302722677588,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8569292969987146
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.00343519244233903,
          "train_loss": 0.002355591394007206,
          "validation_loss": 0.008457519114017487,
          "train_metric": 0.00023724653874523938,
          "validation_metric": 0.0008518103859387338,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8706362399989303
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.012224615074956091,
          "train_loss": 0.004575670231133699,
          "validation_loss": 0.012193112634122372,
          "train_metric": 0.00046084471978247166,
          "validation_metric": 0.0012280457885935903,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8848354000001564
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.007198922423413023,
          "train_loss": 0.004952558316290379,
          "validation_loss": 0.0059533496387302876,
          "train_metric": 0.0004988035652786493,
          "validation_metric": 0.000599599676206708,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8992599960001826
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.004574588332616258,
          "train_loss": 0.008102833293378353,
          "validation_loss": 0.0036516161635518074,
          "train_metric": 0.0008160877041518688,
          "validation_metric": 0.0003677773929666728,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9142775210002583
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.004114874667720869,
          "train_loss": 0.0027797252405434847,
          "validation_loss": 0.011636012233793736,
          "train_metric": 0.00027996368589811027,
          "validation_metric": 0.0011719363974407315,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9279379709987552
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.0028922135152242845,
          "train_loss": 0.006677310913801193,
          "validation_loss": 0.013914878480136395,
          "train_metric": 0.0006725142011418939,
          "validation_metric": 0.0014014554908499122,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9423347510000895
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008213680539483903,
          "train_loss": 0.0030607839580625296,
          "validation_loss": 0.012357104569673538,
          "train_metric": 0.00030827094451524317,
          "validation_metric": 0.0012445623287931085,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9562685059991054
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.00333169008808909,
          "train_loss": 0.004170401021838188,
          "validation_loss": 0.005583195481449366,
          "train_metric": 0.00042002752888947725,
          "validation_metric": 0.0005623189499601722,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9700924919998215
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.0026944439887301996,
          "train_loss": 0.002650755923241377,
          "validation_loss": 0.01314778346568346,
          "train_metric": 0.00026697441353462636,
          "validation_metric": 0.0013241965789347887,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.983287698998538
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.002719895896007074,
          "train_loss": 0.003546384396031499,
          "validation_loss": 0.006438537500798702,
          "train_metric": 0.0003571789129637182,
          "validation_metric": 0.0006484658224508166,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9971232709995093
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.004830198678973829,
          "train_loss": 0.0038257238920778036,
          "validation_loss": 0.0068717109970748425,
          "train_metric": 0.00038531291647814214,
          "validation_metric": 0.0006920936284586787,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.010612478999974
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.003925781038560672,
          "train_loss": 0.004397534765303135,
          "validation_loss": 0.01739174872636795,
          "train_metric": 0.00044290366349741817,
          "validation_metric": 0.0017516330117359757,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0243441789989447
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.00702551580616273,
          "train_loss": 0.005306765902787447,
          "validation_loss": 0.007704962510615587,
          "train_metric": 0.000534478051122278,
          "validation_metric": 0.0007760155713185668,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0381226609988516
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0036716123522637645,
          "train_loss": 0.004096755757927895,
          "validation_loss": 0.006455617491155863,
          "train_metric": 0.00041261024307459593,
          "validation_metric": 0.0006501862080767751,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0517814340000768
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.0031072380661498755,
          "train_loss": 0.001971377059817314,
          "validation_loss": 0.008943586610257626,
          "train_metric": 0.00019854986749123782,
          "validation_metric": 0.0009007652406580746,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0649184079993574
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0020024125115014613,
          "train_loss": 0.001408973941579461,
          "validation_loss": 0.009472215548157692,
          "train_metric": 0.00014190668298397213,
          "validation_metric": 0.0009540069149807096,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.0781472660000873
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.001719378940833849,
          "train_loss": 0.0016232072375714779,
          "validation_loss": 0.007954126223921776,
          "train_metric": 0.00016348350618500262,
          "validation_metric": 0.0008011105237528682,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.0918792140000733
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.001791495704310364,
          "train_loss": 0.0012808110332116485,
          "validation_loss": 0.009574257768690586,
          "train_metric": 0.00012899859575554729,
          "validation_metric": 0.0009642841177992523,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.1055612499985727
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.0015006730100139976,
          "train_loss": 0.0015882713487371802,
          "validation_loss": 0.007926885969936848,
          "train_metric": 0.00015996488218661398,
          "validation_metric": 0.0007983669056557119,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.119909633998759
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0016352456041204277,
          "train_loss": 0.0016078423941507936,
          "validation_loss": 0.008539360016584396,
          "train_metric": 0.0001619360118638724,
          "validation_metric": 0.0008600529399700463,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.1402663989993016
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0013124882789270487,
          "train_loss": 0.0010964293032884598,
          "validation_loss": 0.009672518819570541,
          "train_metric": 0.00011042833648389205,
          "validation_metric": 0.0009741806425154209,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.1569607080000424
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0016128199367813068,
          "train_loss": 0.0022245259024202824,
          "validation_loss": 0.014530581422150135,
          "train_metric": 0.00022404605988413095,
          "validation_metric": 0.0014634669059887528,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.175369739999951
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0023149457556428388,
          "train_loss": 0.0011563189327716827,
          "validation_loss": 0.009921922348439693,
          "train_metric": 0.00011646020720945671,
          "validation_metric": 0.0009992994600906968,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.2056150070002332
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.002749934379608021,
          "train_loss": 0.0015958426520228386,
          "validation_loss": 0.012820369563996792,
          "train_metric": 0.00016072741709649563,
          "validation_metric": 0.0012912207748740911,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.235326661999352
        }
      ],
      "best": {
        "epoch": 48,
        "updates": 768
      },
      "best_validation_metric": 0.0003677773929666728,
      "config": {
        "label": "LayerNorm + relu + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.01917752311865828
    },
    "5bc849f25b25_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.4533875067718327,
          "train_loss": 0.23607385158538818,
          "validation_loss": 0.23469389975070953,
          "train_metric": 0.023776492103934288,
          "validation_metric": 0.02363750897347927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07313907999923686
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.18792315386235714,
          "train_loss": 0.1515727937221527,
          "validation_loss": 0.11647375673055649,
          "train_metric": 0.015265856869518757,
          "validation_metric": 0.011730808764696121,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1413543500002561
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.13304669689387083,
          "train_loss": 0.11357603222131729,
          "validation_loss": 0.09504947811365128,
          "train_metric": 0.011438963003456593,
          "validation_metric": 0.009573034942150116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20888659199954418
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.10178157920017838,
          "train_loss": 0.08968152105808258,
          "validation_loss": 0.07504606992006302,
          "train_metric": 0.00903239380568266,
          "validation_metric": 0.007558364886790514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33088823199977924
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.08518321870360523,
          "train_loss": 0.07935658097267151,
          "validation_loss": 0.07560079544782639,
          "train_metric": 0.00799250416457653,
          "validation_metric": 0.007614235393702984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4313141940001515
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.07195879612118006,
          "train_loss": 0.06808119267225266,
          "validation_loss": 0.06190333515405655,
          "train_metric": 0.006856888998299837,
          "validation_metric": 0.006234677508473396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4990419699988706
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.06672494462691247,
          "train_loss": 0.060704514384269714,
          "validation_loss": 0.05149659514427185,
          "train_metric": 0.006113936193287373,
          "validation_metric": 0.005186547990888357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6349346349998086
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.06260837381705642,
          "train_loss": 0.05938337370753288,
          "validation_loss": 0.049449171870946884,
          "train_metric": 0.005980874877423048,
          "validation_metric": 0.004980339203029871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7036732109991135
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.06097837444394827,
          "train_loss": 0.05915136635303497,
          "validation_loss": 0.050153087824583054,
          "train_metric": 0.005957508459687233,
          "validation_metric": 0.005051236134022474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7692358650001552
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.052757218945771456,
          "train_loss": 0.05054059252142906,
          "validation_loss": 0.04282967373728752,
          "train_metric": 0.005090263672173023,
          "validation_metric": 0.004313648212701082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8370654119989922
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.049022101797163486,
          "train_loss": 0.04715494066476822,
          "validation_loss": 0.03893263638019562,
          "train_metric": 0.00474927294999361,
          "validation_metric": 0.003921152092516422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9053161209994869
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04612647625617683,
          "train_loss": 0.04525791481137276,
          "validation_loss": 0.0385778434574604,
          "train_metric": 0.004558211658149958,
          "validation_metric": 0.003885418875142932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9738797770005476
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.04362151655368507,
          "train_loss": 0.04062950611114502,
          "validation_loss": 0.03383733704686165,
          "train_metric": 0.004092054907232523,
          "validation_metric": 0.003407972864806652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0405336080002598
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.04155640339013189,
          "train_loss": 0.04026982933282852,
          "validation_loss": 0.033952053636312485,
          "train_metric": 0.004055829253047705,
          "validation_metric": 0.003419526619836688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1075863739988563
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.0389854412060231,
          "train_loss": 0.046169765293598175,
          "validation_loss": 0.039701227098703384,
          "train_metric": 0.00465004937723279,
          "validation_metric": 0.003998562227934599,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1733329409998987
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.03759062208700925,
          "train_loss": 0.03777801990509033,
          "validation_loss": 0.031983327120542526,
          "train_metric": 0.0038048638962209225,
          "validation_metric": 0.003221243154257536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2422873390005407
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03390645200852305,
          "train_loss": 0.03163185343146324,
          "validation_loss": 0.025477733463048935,
          "train_metric": 0.0031858442816883326,
          "validation_metric": 0.0025660244282335043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.503508141000566
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.03297593432944268,
          "train_loss": 0.03252008929848671,
          "validation_loss": 0.025172283872961998,
          "train_metric": 0.0032753043342381716,
          "validation_metric": 0.002535260282456875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6474537459998828
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.03237103816354647,
          "train_loss": 0.029383456334471703,
          "validation_loss": 0.021564599126577377,
          "train_metric": 0.0029593941289931536,
          "validation_metric": 0.0021719078067690134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7852621479996742
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.026710529753472656,
          "train_loss": 0.025510035455226898,
          "validation_loss": 0.02122308686375618,
          "train_metric": 0.0025692773051559925,
          "validation_metric": 0.00213751127012074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8815731499998947
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.026009427790995687,
          "train_loss": 0.026116108521819115,
          "validation_loss": 0.021488625556230545,
          "train_metric": 0.0026303185150027275,
          "validation_metric": 0.002164255827665329,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9877929390004283
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.024129130237270147,
          "train_loss": 0.02176036685705185,
          "validation_loss": 0.017171265557408333,
          "train_metric": 0.002191624604165554,
          "validation_metric": 0.0017294271383434534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.100082229999316
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.023260375950485468,
          "train_loss": 0.02014264091849327,
          "validation_loss": 0.015213078819215298,
          "train_metric": 0.0020286934450268745,
          "validation_metric": 0.0015322054969146848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1679877339993254
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.02255199186038226,
          "train_loss": 0.020951731130480766,
          "validation_loss": 0.017546795308589935,
          "train_metric": 0.002110181376338005,
          "validation_metric": 0.001767248846590519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2484993750003923
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.019907136273104697,
          "train_loss": 0.020050933584570885,
          "validation_loss": 0.014397621154785156,
          "train_metric": 0.002019456820562482,
          "validation_metric": 0.0014500754186883569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.270874718000414
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.019589825882576406,
          "train_loss": 0.018015272915363312,
          "validation_loss": 0.015123320743441582,
          "train_metric": 0.0018144322093576193,
          "validation_metric": 0.0015231654979288578,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.290956491000543
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.017295620578806847,
          "train_loss": 0.015368524007499218,
          "validation_loss": 0.012418915517628193,
          "train_metric": 0.0015478613786399364,
          "validation_metric": 0.0012507877545431256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.309923216000243
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.0154280370043125,
          "train_loss": 0.015355197712779045,
          "validation_loss": 0.012731400318443775,
          "train_metric": 0.0015465191099792719,
          "validation_metric": 0.0012822600547224283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.327041742999427
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.016729709721403196,
          "train_loss": 0.015733517706394196,
          "validation_loss": 0.01229722611606121,
          "train_metric": 0.0015846219612285495,
          "validation_metric": 0.0012385316658765078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3452743679990817
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.01713228109292686,
          "train_loss": 0.015824001282453537,
          "validation_loss": 0.01352560892701149,
          "train_metric": 0.0015937354182824492,
          "validation_metric": 0.0013622497208416462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.362462634999247
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.020328038604930043,
          "train_loss": 0.014254491776227951,
          "validation_loss": 0.01142613310366869,
          "train_metric": 0.0014356601750478148,
          "validation_metric": 0.0011507982853800058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3802188169993315
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.016385189519496635,
          "train_loss": 0.012247339822351933,
          "validation_loss": 0.008907568641006947,
          "train_metric": 0.001233507297001779,
          "validation_metric": 0.0008971377392299473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3990673560001596
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.014947676099836826,
          "train_loss": 0.01345524750649929,
          "validation_loss": 0.010800512507557869,
          "train_metric": 0.001355163287371397,
          "validation_metric": 0.00108778802677989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4162008709990914
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.01746969239320606,
          "train_loss": 0.01597156934440136,
          "validation_loss": 0.012431350536644459,
          "train_metric": 0.0016085980460047722,
          "validation_metric": 0.0012520401505753398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.433903281000312
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014921038688044064,
          "train_loss": 0.015928193926811218,
          "validation_loss": 0.014794972725212574,
          "train_metric": 0.0016042294446378946,
          "validation_metric": 0.0014900953974574804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.451105660000394
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.018117933708708733,
          "train_loss": 0.02529345080256462,
          "validation_loss": 0.022420872002840042,
          "train_metric": 0.0025474634021520615,
          "validation_metric": 0.002258148044347763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4690336479998223
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.015057234646519646,
          "train_loss": 0.012036655098199844,
          "validation_loss": 0.011001024395227432,
          "train_metric": 0.0012122879270464182,
          "validation_metric": 0.001107982825487852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4863659009988623
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.013253879238618538,
          "train_loss": 0.01084258034825325,
          "validation_loss": 0.009163621813058853,
          "train_metric": 0.0010920249624177814,
          "validation_metric": 0.0009229264105670154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.503389543000594
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.010740257654106244,
          "train_loss": 0.010983876883983612,
          "validation_loss": 0.007662314921617508,
          "train_metric": 0.0011062559206038713,
          "validation_metric": 0.0007717203116044402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5219770319999952
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.011668395760352723,
          "train_loss": 0.013261827640235424,
          "validation_loss": 0.010143599472939968,
          "train_metric": 0.0013356826966628432,
          "validation_metric": 0.0010216259397566319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.540026582999417
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.010171889152843505,
          "train_loss": 0.008394639007747173,
          "validation_loss": 0.006462304852902889,
          "train_metric": 0.0008454773342236876,
          "validation_metric": 0.0006508596707135439,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5580245679993823
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.00945035154290963,
          "train_loss": 0.01014809962362051,
          "validation_loss": 0.009052149020135403,
          "train_metric": 0.0010220793774351478,
          "validation_metric": 0.0009116992587223649,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5755714699989767
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009885470208246261,
          "train_loss": 0.008425256237387657,
          "validation_loss": 0.005771612748503685,
          "train_metric": 0.0008485609432682395,
          "validation_metric": 0.0005812955787405372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5936712909988273
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.008287601260235533,
          "train_loss": 0.008009068667888641,
          "validation_loss": 0.006281142123043537,
          "train_metric": 0.000806643976829946,
          "validation_metric": 0.0006326136644929647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.611312404000273
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.012470217305235565,
          "train_loss": 0.009016468189656734,
          "validation_loss": 0.0080361757427454,
          "train_metric": 0.0009081056341528893,
          "validation_metric": 0.0008093741489574313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.629173470999376
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.010691074712667614,
          "train_loss": 0.00837999302893877,
          "validation_loss": 0.006612769328057766,
          "train_metric": 0.0008440022938884795,
          "validation_metric": 0.0006660138606093824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6467965520005237
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.009084325531148352,
          "train_loss": 0.008290590718388557,
          "validation_loss": 0.0060104141011834145,
          "train_metric": 0.0008349979761987925,
          "validation_metric": 0.0006053468678146601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6639602100003685
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.007990338111994788,
          "train_loss": 0.01227379310876131,
          "validation_loss": 0.011215022765100002,
          "train_metric": 0.0012361713452264667,
          "validation_metric": 0.0011295359581708908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6810124319999886
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.010695809818571433,
          "train_loss": 0.008311442099511623,
          "validation_loss": 0.006138735916465521,
          "train_metric": 0.0008370979339815676,
          "validation_metric": 0.0006182710640132427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6975771649995295
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.009366934529680293,
          "train_loss": 0.00727478601038456,
          "validation_loss": 0.004861930850893259,
          "train_metric": 0.0007326898630708456,
          "validation_metric": 0.0004896759055554867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7153401909999957
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.008054503327002749,
          "train_loss": 0.008607103489339352,
          "validation_loss": 0.006797916255891323,
          "train_metric": 0.0008668758673593402,
          "validation_metric": 0.0006846612668596208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.732658315999288
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0070726188714616,
          "train_loss": 0.006159184966236353,
          "validation_loss": 0.004700029268860817,
          "train_metric": 0.0006203305092640221,
          "validation_metric": 0.00047336978605017066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7501213870000356
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.007670263643376529,
          "train_loss": 0.009581781923770905,
          "validation_loss": 0.007878285832703114,
          "train_metric": 0.0009650419233366847,
          "validation_metric": 0.0007934721070341766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.865433947999918
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.009720426896819845,
          "train_loss": 0.015219557099044323,
          "validation_loss": 0.01584332063794136,
          "train_metric": 0.0015328580047935247,
          "validation_metric": 0.0015956811839714646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.883349822999662
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.016255740949418396,
          "train_loss": 0.01885773055255413,
          "validation_loss": 0.019417332485318184,
          "train_metric": 0.0018992816330865026,
          "validation_metric": 0.0019556425977498293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9004522719988017
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.011518082421389408,
          "train_loss": 0.011556210927665234,
          "validation_loss": 0.008255921304225922,
          "train_metric": 0.0011638993164524436,
          "validation_metric": 0.0008315060986205935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.917683625999416
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.009397679910762236,
          "train_loss": 0.005451057571917772,
          "validation_loss": 0.003757721744477749,
          "train_metric": 0.0005490105249918997,
          "validation_metric": 0.0003784639702644199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.935827466999399
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.005856432529981248,
          "train_loss": 0.005307372193783522,
          "validation_loss": 0.0038755061104893684,
          "train_metric": 0.0005345390527509153,
          "validation_metric": 0.00039032677887007594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9535239369997726
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.005967337630863767,
          "train_loss": 0.005051338113844395,
          "validation_loss": 0.004096639808267355,
          "train_metric": 0.0005087523022666574,
          "validation_metric": 0.0004125985433347523,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.971120176000113
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.00570713150955271,
          "train_loss": 0.005845473613590002,
          "validation_loss": 0.005170540418475866,
          "train_metric": 0.0005887348088435829,
          "validation_metric": 0.0005207578651607037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9884947840000677
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0065462313941679895,
          "train_loss": 0.005330507177859545,
          "validation_loss": 0.003239354817196727,
          "train_metric": 0.0005368692218326032,
          "validation_metric": 0.0003262559184804559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0068295460005174
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.00611052056774497,
          "train_loss": 0.004719524644315243,
          "validation_loss": 0.003180380677804351,
          "train_metric": 0.00047533318866044283,
          "validation_metric": 0.00032031629234552383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.025431086000026
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.005034669891756494,
          "train_loss": 0.0048412783071398735,
          "validation_loss": 0.003743167035281658,
          "train_metric": 0.000487595796585083,
          "validation_metric": 0.0003769980394281447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0429361249989597
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.005226737375778612,
          "train_loss": 0.005342314951121807,
          "validation_loss": 0.003920630551874638,
          "train_metric": 0.0005380582879297435,
          "validation_metric": 0.0003948715457227081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.060299380998913
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.00489465642021969,
          "train_loss": 0.00503171980381012,
          "validation_loss": 0.003621733048930764,
          "train_metric": 0.0005067764432169497,
          "validation_metric": 0.0003647676785476506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0772826800002804
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.005051887943409383,
          "train_loss": 0.006425080355256796,
          "validation_loss": 0.004099943209439516,
          "train_metric": 0.0006471105152741075,
          "validation_metric": 0.0004129313165321946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0947212769988255
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0053282281733118,
          "train_loss": 0.0042839739471673965,
          "validation_loss": 0.003196745179593563,
          "train_metric": 0.00043146612006239593,
          "validation_metric": 0.0003219644713681191,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.111612527000034
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.006007408715959173,
          "train_loss": 0.006090410519391298,
          "validation_loss": 0.005225718021392822,
          "train_metric": 0.0006134039140306413,
          "validation_metric": 0.000526315183378756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1286266699989937
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.004928238282445818,
          "train_loss": 0.006109215784817934,
          "validation_loss": 0.006595205049961805,
          "train_metric": 0.0006152978749014437,
          "validation_metric": 0.0006642448715865612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.145234501998857
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.004878773310338147,
          "train_loss": 0.004665357060730457,
          "validation_loss": 0.0028979070484638214,
          "train_metric": 0.00046987776295281947,
          "validation_metric": 0.00029186662868596613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1627519680005207
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.005996902516926639,
          "train_loss": 0.012495739385485649,
          "validation_loss": 0.009249559603631496,
          "train_metric": 0.0012585250660777092,
          "validation_metric": 0.0009315817733295262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1799967650003964
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.009059346659341827,
          "train_loss": 0.007268242537975311,
          "validation_loss": 0.005201374646276236,
          "train_metric": 0.0007320307195186615,
          "validation_metric": 0.0005238634184934199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.197047952999128
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0072491949395043775,
          "train_loss": 0.0072659337893128395,
          "validation_loss": 0.005736661609262228,
          "train_metric": 0.0007317982381209731,
          "validation_metric": 0.0005777755286544561,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2138613059996715
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.006947963709535543,
          "train_loss": 0.004304398782551289,
          "validation_loss": 0.0031396036501973867,
          "train_metric": 0.00043352326611056924,
          "validation_metric": 0.00031620939262211323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2311238199999934
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.004752985296363477,
          "train_loss": 0.0041091241873800755,
          "validation_loss": 0.0029762170743197203,
          "train_metric": 0.0004138559161219746,
          "validation_metric": 0.00029975365032441914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2480900609989476
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.008463983576803003,
          "train_loss": 0.00544814532622695,
          "validation_loss": 0.0038710974622517824,
          "train_metric": 0.0005487171583808959,
          "validation_metric": 0.000389882770832628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2654826660000253
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.006841984868515283,
          "train_loss": 0.003808444831520319,
          "validation_loss": 0.0030762404203414917,
          "train_metric": 0.0003835725656244904,
          "validation_metric": 0.0003098276793025434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.282696751999538
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006057678139768541,
          "train_loss": 0.006204069592058659,
          "validation_loss": 0.005628124810755253,
          "train_metric": 0.0006248512072488666,
          "validation_metric": 0.000566844129934907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3003135039998597
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.004686398409830872,
          "train_loss": 0.004135404713451862,
          "validation_loss": 0.0025229970924556255,
          "train_metric": 0.0004165028512943536,
          "validation_metric": 0.0002541069989092648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.317916451000201
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.003738334766239859,
          "train_loss": 0.0032569661270827055,
          "validation_loss": 0.002777101704850793,
          "train_metric": 0.00032802968053147197,
          "validation_metric": 0.0002796995104290545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.335041962000105
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.005030309803260025,
          "train_loss": 0.00613615196198225,
          "validation_loss": 0.006478056777268648,
          "train_metric": 0.0006180107593536377,
          "validation_metric": 0.0006524462369270623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.352355560999058
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.004809984740859363,
          "train_loss": 0.0033369562588632107,
          "validation_loss": 0.002576836384832859,
          "train_metric": 0.00033608597004786134,
          "validation_metric": 0.00025952953728847206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3694223729999067
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.003906025674950797,
          "train_loss": 0.0048045264557003975,
          "validation_loss": 0.004254537634551525,
          "train_metric": 0.0004838943132199347,
          "validation_metric": 0.00042850145837292075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3866050540000288
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.005448948410048615,
          "train_loss": 0.005645789206027985,
          "validation_loss": 0.004060148261487484,
          "train_metric": 0.0005686232470907271,
          "validation_metric": 0.0004089232243131846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.403371230999255
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005252400515018962,
          "train_loss": 0.005102651659399271,
          "validation_loss": 0.005062987562268972,
          "train_metric": 0.0005139203276485205,
          "validation_metric": 0.0005099255358800292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.419666451000012
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.00440666491704178,
          "train_loss": 0.004108186811208725,
          "validation_loss": 0.0029798434115946293,
          "train_metric": 0.0004137615906074643,
          "validation_metric": 0.0003001189325004816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.436242436999237
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.00479293245007284,
          "train_loss": 0.0036091520451009274,
          "validation_loss": 0.003237219760194421,
          "train_metric": 0.0003635005559772253,
          "validation_metric": 0.00032604095758870244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4529666580001503
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0037656951681128703,
          "train_loss": 0.0033405404537916183,
          "validation_loss": 0.002654704498127103,
          "train_metric": 0.00033644697396084666,
          "validation_metric": 0.00026737211737781763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4696686400002363
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0032498820437467657,
          "train_loss": 0.0028385049663484097,
          "validation_loss": 0.0022084307856857777,
          "train_metric": 0.00028588378336280584,
          "validation_metric": 0.00022242506383918226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.589296727999681
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0030899235789547674,
          "train_loss": 0.0026463421527296305,
          "validation_loss": 0.0016440614126622677,
          "train_metric": 0.0002665298816282302,
          "validation_metric": 0.00016558385686948895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6081908989999647
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0029565395852841903,
          "train_loss": 0.004266181029379368,
          "validation_loss": 0.0029921522364020348,
          "train_metric": 0.00042967419722117484,
          "validation_metric": 0.00030135863926261663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.625784178999311
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.004501827483181842,
          "train_loss": 0.002989942440763116,
          "validation_loss": 0.0019288909388706088,
          "train_metric": 0.0003011360240634531,
          "validation_metric": 0.0001942708477145061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.660838784999214
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0048313349834643304,
          "train_loss": 0.002921847626566887,
          "validation_loss": 0.0016070499550551176,
          "train_metric": 0.00029427779372781515,
          "validation_metric": 0.00016185620916076005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.709222100998886
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0032557330014242325,
          "train_loss": 0.003302534343674779,
          "validation_loss": 0.003115077270194888,
          "train_metric": 0.0003326191508676857,
          "validation_metric": 0.00031373914680443704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.752606181000374
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0035759690144914202,
          "train_loss": 0.002626328496262431,
          "validation_loss": 0.0016278572147712111,
          "train_metric": 0.0002645141794346273,
          "validation_metric": 0.00016395181592088193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7701587909996306
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0033441103078075685,
          "train_loss": 0.0024596580769866705,
          "validation_loss": 0.0017771890852600336,
          "train_metric": 0.00024772778851911426,
          "validation_metric": 0.0001789920061128214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7876186710000184
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.005403652357927058,
          "train_loss": 0.005403873510658741,
          "validation_loss": 0.005336983595043421,
          "train_metric": 0.0005442583351396024,
          "validation_metric": 0.0005375214968807995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8043117489996803
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.010415417462354526,
          "train_loss": 0.005917325150221586,
          "validation_loss": 0.005376921501010656,
          "train_metric": 0.000595971301663667,
          "validation_metric": 0.0005415438208729029,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.821325700999296
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.006867525691632181,
          "train_loss": 0.007065956946462393,
          "validation_loss": 0.005118570290505886,
          "train_metric": 0.000711657281499356,
          "validation_metric": 0.000515523599460721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8382450529989
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.007654631088371389,
          "train_loss": 0.006965443957597017,
          "validation_loss": 0.005810673348605633,
          "train_metric": 0.0007015339797362685,
          "validation_metric": 0.0005852297763340175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8557506239994837
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.004485361117986031,
          "train_loss": 0.0030118217691779137,
          "validation_loss": 0.0028321505524218082,
          "train_metric": 0.00030333962058648467,
          "validation_metric": 0.00028524379013106227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.873467499999606
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0030661513956147246,
          "train_loss": 0.003227348206564784,
          "validation_loss": 0.0026655802503228188,
          "train_metric": 0.00032504668342880905,
          "validation_metric": 0.00026846746914088726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8913278719992377
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.002665597276063636,
          "train_loss": 0.002407657215371728,
          "validation_loss": 0.0013992124004289508,
          "train_metric": 0.00024249040870927274,
          "validation_metric": 0.0001409235701430589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.909367033998933
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.00261260278057307,
          "train_loss": 0.002531382953748107,
          "validation_loss": 0.001400784240104258,
          "train_metric": 0.00025495162117294967,
          "validation_metric": 0.0001410818804288283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.926614993999465
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.00264340846115374,
          "train_loss": 0.0032702216412872076,
          "validation_loss": 0.0029423993546515703,
          "train_metric": 0.00032936478964984417,
          "validation_metric": 0.00029634771635755897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9433637210004235
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0026679264701670036,
          "train_loss": 0.002671831287443638,
          "validation_loss": 0.0021777537185698748,
          "train_metric": 0.00026909710140898824,
          "validation_metric": 0.00021933538664598018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9610246539996297
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0025181930031976663,
          "train_loss": 0.002438729163259268,
          "validation_loss": 0.0019511622376739979,
          "train_metric": 0.0002456198853906244,
          "validation_metric": 0.00019651392358355224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.978402868999183
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0029138306417735294,
          "train_loss": 0.005897448398172855,
          "validation_loss": 0.005751435179263353,
          "train_metric": 0.0005939693073742092,
          "validation_metric": 0.0005792634328827262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.995680038999126
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.005577649957558606,
          "train_loss": 0.002466340083628893,
          "validation_loss": 0.0017874768236652017,
          "train_metric": 0.00024840072728693485,
          "validation_metric": 0.00018002811702899635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.012804698999389
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.00573070678365184,
          "train_loss": 0.006290717050433159,
          "validation_loss": 0.00643183384090662,
          "train_metric": 0.0006335780490189791,
          "validation_metric": 0.0006477907299995422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.029607720000058
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.005476280512084486,
          "train_loss": 0.0045829345472157,
          "validation_loss": 0.003017950803041458,
          "train_metric": 0.0004615763609763235,
          "validation_metric": 0.000303956912830472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.047270192999349
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.003168667064528563,
          "train_loss": 0.0031393817625939846,
          "validation_loss": 0.0030156481079757214,
          "train_metric": 0.000316187011776492,
          "validation_metric": 0.0003037250426132232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.065162075999979
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.003031807736988412,
          "train_loss": 0.0025101318024098873,
          "validation_loss": 0.0020753375720232725,
          "train_metric": 0.00025281126727350056,
          "validation_metric": 0.00020902040705550462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.082658007999271
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.002249700224638218,
          "train_loss": 0.0018499811412766576,
          "validation_loss": 0.0013398014707490802,
          "train_metric": 0.00018632333376444876,
          "validation_metric": 0.0001349398953607306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.100395199999184
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0020899204082525102,
          "train_loss": 0.004798028618097305,
          "validation_loss": 0.004584485664963722,
          "train_metric": 0.00048323985538445413,
          "validation_metric": 0.00046173264854587615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.1172057969997695
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0035895399960281793,
          "train_loss": 0.005140263587236404,
          "validation_loss": 0.005124194547533989,
          "train_metric": 0.0005177085404284298,
          "validation_metric": 0.0005160900764167309,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.134073303999685
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0033274536399403587,
          "train_loss": 0.003378899535164237,
          "validation_loss": 0.002768230391666293,
          "train_metric": 0.0003403104201424867,
          "validation_metric": 0.0002788060228340328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.150759518999621
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.003609997766034212,
          "train_loss": 0.003766664769500494,
          "validation_loss": 0.0038273020181804895,
          "train_metric": 0.0003793646756093949,
          "validation_metric": 0.0003854718233924359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.167712446000223
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.003287670719146263,
          "train_loss": 0.004469671286642551,
          "validation_loss": 0.004026212729513645,
          "train_metric": 0.0004501689691096544,
          "validation_metric": 0.00040550538687966764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.184123512999577
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.003759970139071811,
          "train_loss": 0.004752433858811855,
          "validation_loss": 0.005170000717043877,
          "train_metric": 0.00047864782391116023,
          "validation_metric": 0.0005207035574130714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.201021209999453
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0030612527443736326,
          "train_loss": 0.002212698571383953,
          "validation_loss": 0.0015920678852126002,
          "train_metric": 0.00022285491286311299,
          "validation_metric": 0.00016034726286306977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.217497509000168
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.00435583637226955,
          "train_loss": 0.004133421927690506,
          "validation_loss": 0.003583188634365797,
          "train_metric": 0.0004163031408097595,
          "validation_metric": 0.00036088566412217915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.234423211999456
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.003578058196580969,
          "train_loss": 0.0020004971884191036,
          "validation_loss": 0.001784461084753275,
          "train_metric": 0.00020148276234976947,
          "validation_metric": 0.00017972438945434988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2513932259989815
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0026060053023684304,
          "train_loss": 0.0021818976383656263,
          "validation_loss": 0.0018673925660550594,
          "train_metric": 0.0002197527210228145,
          "validation_metric": 0.00018807692686095834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2682452730005025
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.002625781577080488,
          "train_loss": 0.0025918714236468077,
          "validation_loss": 0.002174088265746832,
          "train_metric": 0.00026104378048330545,
          "validation_metric": 0.00021896621910855174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.382449572000041
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.001903202491121192,
          "train_loss": 0.0015606095548719168,
          "validation_loss": 0.0010293308878317475,
          "train_metric": 0.00015717888891231269,
          "validation_metric": 0.00010367044160375372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.400775635998798
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.002938209468993591,
          "train_loss": 0.0028978371992707253,
          "validation_loss": 0.0018301713280379772,
          "train_metric": 0.0002918595855589956,
          "validation_metric": 0.00018432814977131784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.417436961999556
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0027406745139160194,
          "train_loss": 0.002567772753536701,
          "validation_loss": 0.002477837959304452,
          "train_metric": 0.0002586166374385357,
          "validation_metric": 0.00024955873959697783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.434332369999538
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0031034180283313617,
          "train_loss": 0.0022959173657000065,
          "validation_loss": 0.0020760272163897753,
          "train_metric": 0.00023123642313294113,
          "validation_metric": 0.00020908983424305916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.451183020999451
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0020835576215176843,
          "train_loss": 0.001434429781511426,
          "validation_loss": 0.0009832908399403095,
          "train_metric": 0.00014447051216848195,
          "validation_metric": 9.903345926431939e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.471265815000152
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.004515076576353749,
          "train_loss": 0.0021404733415693045,
          "validation_loss": 0.0018725330010056496,
          "train_metric": 0.00021558065782301128,
          "validation_metric": 0.00018859468400478363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.489080676999947
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0022267352615017444,
          "train_loss": 0.002635679440572858,
          "validation_loss": 0.002244101371616125,
          "train_metric": 0.0002654559793882072,
          "validation_metric": 0.00022601768432650715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.50649073599925
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0021948056455585174,
          "train_loss": 0.0017627379857003689,
          "validation_loss": 0.0010068250121548772,
          "train_metric": 0.0001775364944478497,
          "validation_metric": 0.00010140372614841908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.523995120000109
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0019648598463390954,
          "train_loss": 0.002116976771503687,
          "validation_loss": 0.0014492467744275928,
          "train_metric": 0.0002132141380570829,
          "validation_metric": 0.0001459628256270662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.5414407110001775
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0021465518802870065,
          "train_loss": 0.001607498386874795,
          "validation_loss": 0.001149441348388791,
          "train_metric": 0.00016190136375371367,
          "validation_metric": 0.00011576752876862884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.558179233999908
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0018135067184630316,
          "train_loss": 0.0016954857856035233,
          "validation_loss": 0.001424490357749164,
          "train_metric": 0.00017076311632990837,
          "validation_metric": 0.0001434694422641769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.5753351650000695
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.002125087754393462,
          "train_loss": 0.0033307718113064766,
          "validation_loss": 0.0024017037358134985,
          "train_metric": 0.00033546314807608724,
          "validation_metric": 0.00024189082614611834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.592138985999554
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.006833330222434597,
          "train_loss": 0.004477747716009617,
          "validation_loss": 0.0038435833994299173,
          "train_metric": 0.0004509823629632592,
          "validation_metric": 0.0003871116496156901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.609326632000375
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0029655855760211125,
          "train_loss": 0.0033502138685435057,
          "validation_loss": 0.0025363194290548563,
          "train_metric": 0.0003374212537892163,
          "validation_metric": 0.00025544880190864205,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.626240780000444
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0020473326039791573,
          "train_loss": 0.002128754509612918,
          "validation_loss": 0.0012995712459087372,
          "train_metric": 0.00021440035197883844,
          "validation_metric": 0.00013088806008454412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.643284911999217
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0035319341186550446,
          "train_loss": 0.0024205376394093037,
          "validation_loss": 0.0017557241953909397,
          "train_metric": 0.00024378769739996642,
          "validation_metric": 0.0001768301153788343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.660843120998834
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.002245432471681852,
          "train_loss": 0.002478276612237096,
          "validation_loss": 0.0019130428554490209,
          "train_metric": 0.0002496029483154416,
          "validation_metric": 0.00019267466268502176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.678043374000481
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0020329057369963266,
          "train_loss": 0.003525164909660816,
          "validation_loss": 0.003382037626579404,
          "train_metric": 0.00035504173138178885,
          "validation_metric": 0.0003406264295335859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.69476376899911
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.00248192797153024,
          "train_loss": 0.0014074752107262611,
          "validation_loss": 0.0009456945699639618,
          "train_metric": 0.0001417557505192235,
          "validation_metric": 9.524689812678844e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.712666514999
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0014384709829755593,
          "train_loss": 0.0022625387646257877,
          "validation_loss": 0.00218959734775126,
          "train_metric": 0.0002278745814692229,
          "validation_metric": 0.00022052823624107987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.729756966000423
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.004275443998267292,
          "train_loss": 0.004239027854055166,
          "validation_loss": 0.0039953854866325855,
          "train_metric": 0.000426939339376986,
          "validation_metric": 0.0004024005902465433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.74700730099903
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0027245093224337325,
          "train_loss": 0.0017387941479682922,
          "validation_loss": 0.0011526519665494561,
          "train_metric": 0.00017512496560811996,
          "validation_metric": 0.00011609087232500315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.7641148479997355
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0018652441431186162,
          "train_loss": 0.0018250684952363372,
          "validation_loss": 0.0009578680037520826,
          "train_metric": 0.00018381419067736715,
          "validation_metric": 9.647294791648164e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.781335820000095
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0022047787315386813,
          "train_loss": 0.00763572147116065,
          "validation_loss": 0.006257608067244291,
          "train_metric": 0.0007690420607104897,
          "validation_metric": 0.0006302432739175856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.797793726000236
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.005657732275722083,
          "train_loss": 0.005602178629487753,
          "validation_loss": 0.004461579956114292,
          "train_metric": 0.0005642309552058578,
          "validation_metric": 0.0004493539163377136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.814526288999332
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.005245436012046412,
          "train_loss": 0.0015944303013384342,
          "validation_loss": 0.0009763073176145554,
          "train_metric": 0.00016058518667705357,
          "validation_metric": 9.83301069936715e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.830811492000066
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.002275213657412678,
          "train_loss": 0.0015102248871698976,
          "validation_loss": 0.0012591993436217308,
          "train_metric": 0.00015210433048196137,
          "validation_metric": 0.00012682196393143386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.8482834309998
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.001517502259957837,
          "train_loss": 0.0016329508507624269,
          "validation_loss": 0.001277376082725823,
          "train_metric": 0.000164464843692258,
          "validation_metric": 0.00012865265307482332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.8652836849996675
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0013887211043766001,
          "train_loss": 0.0011386564001441002,
          "validation_loss": 0.0007162950350902975,
          "train_metric": 0.00011468132288428023,
          "validation_metric": 7.214260403998196e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.8837004079996404
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0018280936474184273,
          "train_loss": 0.0030964903999119997,
          "validation_loss": 0.0022518460173159838,
          "train_metric": 0.00031186715932562947,
          "validation_metric": 0.00022679765243083239,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.901595125998938
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.00625359287369065,
          "train_loss": 0.009855058044195175,
          "validation_loss": 0.00972878560423851,
          "train_metric": 0.0009925651829689741,
          "validation_metric": 0.000979847740381956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.918842451999808
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.006092103096307255,
          "train_loss": 0.004092697985470295,
          "validation_loss": 0.003581519704312086,
          "train_metric": 0.00041220159619115293,
          "validation_metric": 0.00036071750218980014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.935651768999378
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0025830627128016204,
          "train_loss": 0.0015254061436280608,
          "validation_loss": 0.0010523379314690828,
          "train_metric": 0.0001536333147669211,
          "validation_metric": 0.00010598763765301555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.952157398000054
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0023916380614537047,
          "train_loss": 0.004739873111248016,
          "validation_loss": 0.004851037170737982,
          "train_metric": 0.0004773827095050365,
          "validation_metric": 0.0004885786911472678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.969068401000186
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.002629463788252906,
          "train_loss": 0.0032812159042805433,
          "validation_loss": 0.0022281124256551266,
          "train_metric": 0.0003304720448795706,
          "validation_metric": 0.0002244073257315904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.985974895998879
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0028591041736945044,
          "train_loss": 0.0030326612759381533,
          "validation_loss": 0.00211365451104939,
          "train_metric": 0.0003054385306313634,
          "validation_metric": 0.0002128795749740675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.098923420000574
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.004608736351656262,
          "train_loss": 0.00397402374073863,
          "validation_loss": 0.003508537309244275,
          "train_metric": 0.0004002491186838597,
          "validation_metric": 0.00035336706787347794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.1171512130003975
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.00286026913090609,
          "train_loss": 0.0014775455929338932,
          "validation_loss": 0.0010522636584937572,
          "train_metric": 0.00014881297829560935,
          "validation_metric": 0.00010598013614071533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.134073361999981
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0021280356049828697,
          "train_loss": 0.002708700718358159,
          "validation_loss": 0.002405296778306365,
          "train_metric": 0.000272810400929302,
          "validation_metric": 0.00024225268862210214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.151652368000214
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.001992033234273549,
          "train_loss": 0.0011474380735307932,
          "validation_loss": 0.0007747580530121922,
          "train_metric": 0.00011556575918802992,
          "validation_metric": 7.803079643053934e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.168812447000164
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0017130327360064257,
          "train_loss": 0.0012672528391703963,
          "validation_loss": 0.0007068800623528659,
          "train_metric": 0.00012763305858243257,
          "validation_metric": 7.119437213987112e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.187144297000486
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0014040459409443429,
          "train_loss": 0.0011916437651962042,
          "validation_loss": 0.0009949107188731432,
          "train_metric": 0.0001200179904117249,
          "validation_metric": 0.00010020376066677272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.205080861998795
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.001747392945617321,
          "train_loss": 0.006142228841781616,
          "validation_loss": 0.004849154036492109,
          "train_metric": 0.0006186228711158037,
          "validation_metric": 0.0004883890505880117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.22249540699886
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.005338571390893776,
          "train_loss": 0.002050586976110935,
          "validation_loss": 0.001345084747299552,
          "train_metric": 0.0002065276203211397,
          "validation_metric": 0.00013547198614105582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.239311555998938
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.003008897438121494,
          "train_loss": 0.001304400386288762,
          "validation_loss": 0.0010789069347083569,
          "train_metric": 0.00013137442874722183,
          "validation_metric": 0.00010866356024052948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.255879906999326
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.001527405282104155,
          "train_loss": 0.0015845762100070715,
          "validation_loss": 0.0014949182514101267,
          "train_metric": 0.0001595927169546485,
          "validation_metric": 0.00015056268603075296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.272634998000285
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0017504294264654163,
          "train_loss": 0.00201773876324296,
          "validation_loss": 0.0014195035910233855,
          "train_metric": 0.00020321928604971617,
          "validation_metric": 0.0001429672265658155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.303532488998826
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0016366948002541903,
          "train_loss": 0.0022696400992572308,
          "validation_loss": 0.0016380122397094965,
          "train_metric": 0.00022858985175844282,
          "validation_metric": 0.00016497458273079246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.346390998000061
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.002041146330157062,
          "train_loss": 0.0016334340907633305,
          "validation_loss": 0.0014241826720535755,
          "train_metric": 0.00016451349074486643,
          "validation_metric": 0.0001434384612366557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.364070944999185
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0030755292682442814,
          "train_loss": 0.004427006468176842,
          "validation_loss": 0.0034368832129985094,
          "train_metric": 0.00044587187585420907,
          "validation_metric": 0.0003461503074504435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.381325272999675
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0026588313339743763,
          "train_loss": 0.002534386469051242,
          "validation_loss": 0.002488184254616499,
          "train_metric": 0.0002552540972828865,
          "validation_metric": 0.0002506008022464812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.3981830649991025
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.002479716102243401,
          "train_loss": 0.004686662461608648,
          "validation_loss": 0.0033306381665170193,
          "train_metric": 0.00047202344285324216,
          "validation_metric": 0.00033544967300258577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4145092369999475
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0024963155192381237,
          "train_loss": 0.0010530151193961501,
          "validation_loss": 0.0006571036064997315,
          "train_metric": 0.00010605582792777568,
          "validation_metric": 6.618107727263123e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.432017900999199
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0028783707948605297,
          "train_loss": 0.0032023475505411625,
          "validation_loss": 0.00221449276432395,
          "train_metric": 0.0003225287364330143,
          "validation_metric": 0.00022303557489067316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.449056995999854
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.006648639478953555,
          "train_loss": 0.002668075729161501,
          "validation_loss": 0.002249521901831031,
          "train_metric": 0.00026871878071688116,
          "validation_metric": 0.00022656357032246888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4659158189988375
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.002443896486511221,
          "train_loss": 0.001777957659214735,
          "validation_loss": 0.0016573678003624082,
          "train_metric": 0.00017906937864609063,
          "validation_metric": 0.0001669240155024454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4824442509998335
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0015878250160312746,
          "train_loss": 0.0023226479534059763,
          "validation_loss": 0.001884183380752802,
          "train_metric": 0.0002339285856578499,
          "validation_metric": 0.0001897680340334773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.499498316999961
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0018268515541421948,
          "train_loss": 0.0028435485437512398,
          "validation_loss": 0.0026432392187416553,
          "train_metric": 0.0002863917616195977,
          "validation_metric": 0.0002662173646967858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.51623549299984
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0023720805875200313,
          "train_loss": 0.0012405995512381196,
          "validation_loss": 0.0010323423193767667,
          "train_metric": 0.0001249486522283405,
          "validation_metric": 0.00010397374717285857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.533735042999979
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0015942553000058979,
          "train_loss": 0.0015442351577803493,
          "validation_loss": 0.0012972073163837194,
          "train_metric": 0.00015552970580756664,
          "validation_metric": 0.0001306499761994928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.55038286799936
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0015883794549154118,
          "train_loss": 0.0011914421338588,
          "validation_loss": 0.0006690084701403975,
          "train_metric": 0.00011999769776593894,
          "validation_metric": 6.738008232787251e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.5667477549995965
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0015763302199047757,
          "train_loss": 0.0011714097345247865,
          "validation_loss": 0.0009343525744043291,
          "train_metric": 0.0001179800892714411,
          "validation_metric": 9.410457278136164e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.583886636999523
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0010614596321829595,
          "train_loss": 0.0009163783397525549,
          "validation_loss": 0.0006737568764947355,
          "train_metric": 9.22942635952495e-05,
          "validation_metric": 6.785832374589518e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.604411131000234
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.001610094856005162,
          "train_loss": 0.0015585293294861913,
          "validation_loss": 0.0012545366771519184,
          "train_metric": 0.00015696935588493943,
          "validation_metric": 0.000126352344523184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.6277112789994135
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.001347090812487295,
          "train_loss": 0.0036931971553713083,
          "validation_loss": 0.0034119775518774986,
          "train_metric": 0.00037196531775407493,
          "validation_metric": 0.00034364184830337763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.639270521000071
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.002149197709513828,
          "train_loss": 0.00475872028619051,
          "validation_loss": 0.003928786609321833,
          "train_metric": 0.0004792808904312551,
          "validation_metric": 0.00039569297223351896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.650807358999373
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.0022381906092050485,
          "train_loss": 0.0009477061103098094,
          "validation_loss": 0.0005806891131214797,
          "train_metric": 9.544949716655537e-05,
          "validation_metric": 5.848488217452541e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.663093027998912
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.001060031216184143,
          "train_loss": 0.000803382892627269,
          "validation_loss": 0.0005312809371389449,
          "train_metric": 8.091378549579531e-05,
          "validation_metric": 5.35086692252662e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.675528877000033
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.0012442783736332785,
          "train_loss": 0.001384223229251802,
          "validation_loss": 0.0010321554727852345,
          "train_metric": 0.0001394138962496072,
          "validation_metric": 0.00010395492427051067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.687374577999435
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0015790473189554177,
          "train_loss": 0.0017214352265000343,
          "validation_loss": 0.0015750599559396505,
          "train_metric": 0.00017337664030492306,
          "validation_metric": 0.00015863426961004734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.698982251000416
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.0014315489570435602,
          "train_loss": 0.002178517635911703,
          "validation_loss": 0.001599942217580974,
          "train_metric": 0.0002194122935179621,
          "validation_metric": 0.00016114032769110054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.710362218998853
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.001770471382769756,
          "train_loss": 0.0038569392636418343,
          "validation_loss": 0.003550994209945202,
          "train_metric": 0.0003884567995555699,
          "validation_metric": 0.0003576431190595031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.805894454999361
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.002259081698866794,
          "train_loss": 0.0024194365832954645,
          "validation_loss": 0.0020018205977976322,
          "train_metric": 0.00024367681180592626,
          "validation_metric": 0.00020161602878943086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.818405505000555
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0020777154677489307,
          "train_loss": 0.001362506183795631,
          "validation_loss": 0.0011317358585074544,
          "train_metric": 0.00013722662697546184,
          "validation_metric": 0.00011398427886888385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.829688616000567
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0011675081423163647,
          "train_loss": 0.001014362438581884,
          "validation_loss": 0.0008558441186323762,
          "train_metric": 0.00010216287046205252,
          "validation_metric": 8.619748405180871e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.841385685000205
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0011731334889191203,
          "train_loss": 0.0029615703970193863,
          "validation_loss": 0.002850678050890565,
          "train_metric": 0.00029827855178155005,
          "validation_metric": 0.0002871098695322871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.866125696999006
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.002123970443790313,
          "train_loss": 0.001368621364235878,
          "validation_loss": 0.0010012367274612188,
          "train_metric": 0.00013784253678750247,
          "validation_metric": 0.00010084088717121631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.914458518998799
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0014117876562522724,
          "train_loss": 0.0012317427899688482,
          "validation_loss": 0.0010901251807808876,
          "train_metric": 0.00012405661982484162,
          "validation_metric": 0.00010979340004269034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.961120329999176
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.001053848040101002,
          "train_loss": 0.0010594138875603676,
          "validation_loss": 0.0008060678956098855,
          "train_metric": 0.00010670028859749436,
          "validation_metric": 8.118420373648405e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.000713227998858
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0009520394050923642,
          "train_loss": 0.0008453712798655033,
          "validation_loss": 0.0005695449071936309,
          "train_metric": 8.514270302839577e-05,
          "validation_metric": 5.7362478401046246e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.044158274999063
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0017241679634025786,
          "train_loss": 0.002021695487201214,
          "validation_loss": 0.0018268442945554852,
          "train_metric": 0.00020361777569632977,
          "validation_metric": 0.00018399307737126946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.076648461999866
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0018784153526212322,
          "train_loss": 0.0022997374180704355,
          "validation_loss": 0.002072389703243971,
          "train_metric": 0.00023162116121966392,
          "validation_metric": 0.00020872350432910025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.09392527899945
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0031413399992743507,
          "train_loss": 0.0035017780028283596,
          "validation_loss": 0.0023084746208041906,
          "train_metric": 0.0003526863001752645,
          "validation_metric": 0.00023250111553352326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.111450860000332
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0021301168471836718,
          "train_loss": 0.0010058738989755511,
          "validation_loss": 0.0006636519683524966,
          "train_metric": 0.00010130793816642836,
          "validation_metric": 6.684059917461127e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.128959632998885
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.001339085036306642,
          "train_loss": 0.0027854128275066614,
          "validation_loss": 0.0021994607523083687,
          "train_metric": 0.0002805365948006511,
          "validation_metric": 0.0002215216081822291,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.14657562299908
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0017722308493830496,
          "train_loss": 0.001575938891619444,
          "validation_loss": 0.0012491346569731832,
          "train_metric": 0.0001587227889103815,
          "validation_metric": 0.00012580827751662582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.163984257000266
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0013540846475734725,
          "train_loss": 0.005353070329874754,
          "validation_loss": 0.00548142846673727,
          "train_metric": 0.000539141648914665,
          "validation_metric": 0.0005520693957805634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.181290168000487
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.004256073978467612,
          "train_loss": 0.0013005213113501668,
          "validation_loss": 0.0010776520939543843,
          "train_metric": 0.00013098375347908586,
          "validation_metric": 0.00010853716230485588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.198196866000217
        }
      ],
      "best": {
        "epoch": 193,
        "updates": 3088
      },
      "best_validation_metric": 5.35086692252662e-05,
      "config": {
        "label": "Tanh",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.007314962011197748
    },
    "14985df6b90f_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.23888491606339812,
          "train_loss": 0.08166317641735077,
          "validation_loss": 0.06555181741714478,
          "train_metric": 0.00822481606155634,
          "validation_metric": 0.0066021401435136795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.031406190999405226
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.061784587102010846,
          "train_loss": 0.054036371409893036,
          "validation_loss": 0.045189712196588516,
          "train_metric": 0.005442344583570957,
          "validation_metric": 0.004551341757178307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06317053299972031
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.04121296224184334,
          "train_loss": 0.03482135012745857,
          "validation_loss": 0.027817262336611748,
          "train_metric": 0.003507079090923071,
          "validation_metric": 0.0028016529977321625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0999594659988361
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.029260389623232186,
          "train_loss": 0.02433099038898945,
          "validation_loss": 0.01896427944302559,
          "train_metric": 0.002450528321787715,
          "validation_metric": 0.001910012448206544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11841342699881352
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.024199075298383832,
          "train_loss": 0.02122872695326805,
          "validation_loss": 0.021520046517252922,
          "train_metric": 0.0021380793768912554,
          "validation_metric": 0.0021674202289432287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1321630999991612
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.021335106168407947,
          "train_loss": 0.017451364547014236,
          "validation_loss": 0.01547304354608059,
          "train_metric": 0.0017576378304511309,
          "validation_metric": 0.0015583883505314589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14693344699844602
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.02976817957824096,
          "train_loss": 0.03284674137830734,
          "validation_loss": 0.034943778067827225,
          "train_metric": 0.0033082030713558197,
          "validation_metric": 0.0035194093361496925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1611930509989179
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.02878994884667918,
          "train_loss": 0.02708299085497856,
          "validation_loss": 0.019958991557359695,
          "train_metric": 0.002727699698880315,
          "validation_metric": 0.002010196214541793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1748685389993625
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.016897937923204154,
          "train_loss": 0.01903846114873886,
          "validation_loss": 0.01356436312198639,
          "train_metric": 0.0019174840999767184,
          "validation_metric": 0.0013661530101671815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1897934209991945
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.014107360038906336,
          "train_loss": 0.011092609725892544,
          "validation_loss": 0.009211700409650803,
          "train_metric": 0.001117207109928131,
          "validation_metric": 0.0009277687640860677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20798066399947857
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.011751949205063283,
          "train_loss": 0.008849341422319412,
          "validation_loss": 0.008775648660957813,
          "train_metric": 0.0008912733173929155,
          "validation_metric": 0.0008838512003421783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24179569399893808
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.010980227962136269,
          "train_loss": 0.025530699640512466,
          "validation_loss": 0.02131485566496849,
          "train_metric": 0.0025713585782796144,
          "validation_metric": 0.0021467541810125113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27204532899850165
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.02549975086003542,
          "train_loss": 0.01695833168923855,
          "validation_loss": 0.01574423722922802,
          "train_metric": 0.0017079809913411736,
          "validation_metric": 0.0015857020625844598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.302267243998358
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.019903491061995737,
          "train_loss": 0.013715945184230804,
          "validation_loss": 0.014348994940519333,
          "train_metric": 0.0013814197154715657,
          "validation_metric": 0.0014451781753450632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33606300999963423
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.013678477756911889,
          "train_loss": 0.019475849345326424,
          "validation_loss": 0.013384152203798294,
          "train_metric": 0.0019615362398326397,
          "validation_metric": 0.001348002697341144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36988598499920045
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.011750767356716096,
          "train_loss": 0.011588453315198421,
          "validation_loss": 0.008362028747797012,
          "train_metric": 0.0011671464890241623,
          "validation_metric": 0.0008421927923336625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3970105409989628
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.00787850258348044,
          "train_loss": 0.007302626967430115,
          "validation_loss": 0.0058487593196332455,
          "train_metric": 0.0007354938425123692,
          "validation_metric": 0.000589065661188215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41403330899993307
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.00889447663212195,
          "train_loss": 0.006209243554621935,
          "validation_loss": 0.004950473550707102,
          "train_metric": 0.0006253722822293639,
          "validation_metric": 0.0004985934938304126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43337194599916984
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.011325097439112142,
          "train_loss": 0.0055108158849179745,
          "validation_loss": 0.004341572523117065,
          "train_metric": 0.0005550291971303523,
          "validation_metric": 0.00043726732837967575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45545398099966405
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.007683465169975534,
          "train_loss": 0.004882046952843666,
          "validation_loss": 0.004007525276392698,
          "train_metric": 0.0004917018231935799,
          "validation_metric": 0.0004036233003716916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4700069929986057
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.008050136239035055,
          "train_loss": 0.005788047332316637,
          "validation_loss": 0.004216156434267759,
          "train_metric": 0.0005829509464092553,
          "validation_metric": 0.0004246358585078269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48382755700004054
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.01719522225903347,
          "train_loss": 0.023246873170137405,
          "validation_loss": 0.01762869581580162,
          "train_metric": 0.0023413393646478653,
          "validation_metric": 0.0017754975706338882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6722316579998733
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.02437688660575077,
          "train_loss": 0.009796607308089733,
          "validation_loss": 0.009899741038680077,
          "train_metric": 0.0009866784093901515,
          "validation_metric": 0.0009970655664801598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6871071199984726
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.01053882839914877,
          "train_loss": 0.006641744635999203,
          "validation_loss": 0.005747511517256498,
          "train_metric": 0.00066893215989694,
          "validation_metric": 0.0005788682610727847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7016537929994229
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.010213554996880703,
          "train_loss": 0.007118855603039265,
          "validation_loss": 0.007156596519052982,
          "train_metric": 0.0007169851451180875,
          "validation_metric": 0.0007207861635833979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7158108159983385
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.008563331473851576,
          "train_loss": 0.005285867024213076,
          "validation_loss": 0.004842318128794432,
          "train_metric": 0.0005323732038959861,
          "validation_metric": 0.0004877005412708968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7296438989997114
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.012568902275233995,
          "train_loss": 0.022204900160431862,
          "validation_loss": 0.018985990434885025,
          "train_metric": 0.0022363958414644003,
          "validation_metric": 0.0019121996592730284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7437764129990683
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01679351064376533,
          "train_loss": 0.006613657809793949,
          "validation_loss": 0.005368424113839865,
          "train_metric": 0.0006661033257842064,
          "validation_metric": 0.0005406880518421531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7577355159992294
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.007585612416733056,
          "train_loss": 0.007600780110806227,
          "validation_loss": 0.007061448413878679,
          "train_metric": 0.0007655228255316615,
          "validation_metric": 0.000711203261744231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7715362539984199
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.013993913438753225,
          "train_loss": 0.01605561003088951,
          "validation_loss": 0.010954132303595543,
          "train_metric": 0.001617062371224165,
          "validation_metric": 0.001103260088711977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7852807579984074
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008528002261300571,
          "train_loss": 0.006476330105215311,
          "validation_loss": 0.006094194017350674,
          "train_metric": 0.0006522722542285919,
          "validation_metric": 0.00061378488317132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7988350089999585
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.008905707880330738,
          "train_loss": 0.010734885931015015,
          "validation_loss": 0.008449658751487732,
          "train_metric": 0.0010811785468831658,
          "validation_metric": 0.0008510186453349888,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8122094459995424
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.007577283075079322,
          "train_loss": 0.003922881092876196,
          "validation_loss": 0.0034323823638260365,
          "train_metric": 0.00039509820635430515,
          "validation_metric": 0.0003456970152910799,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.826268748000075
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.00360892678872915,
          "train_loss": 0.0032230280339717865,
          "validation_loss": 0.0027567893266677856,
          "train_metric": 0.0003246115520596504,
          "validation_metric": 0.0002776537439785898,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8407349760000216
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.003924632663256489,
          "train_loss": 0.003139584558084607,
          "validation_loss": 0.002995756920427084,
          "train_metric": 0.00031620750087313354,
          "validation_metric": 0.00030172168044373393,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8543468769985338
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.004978348712029401,
          "train_loss": 0.00343205314129591,
          "validation_loss": 0.0031588103156536818,
          "train_metric": 0.00034566380782052875,
          "validation_metric": 0.00031814377871342003,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8679255749993899
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.003140195731248241,
          "train_loss": 0.0024505648761987686,
          "validation_loss": 0.00236724060960114,
          "train_metric": 0.00024681194918230176,
          "validation_metric": 0.00023841980146244168,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8820304359996953
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.002831874182447791,
          "train_loss": 0.0023875823244452477,
          "validation_loss": 0.0018939138390123844,
          "train_metric": 0.0002404685947112739,
          "validation_metric": 0.00019074806186836213,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8967350089988031
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.006535253283800557,
          "train_loss": 0.028644394129514694,
          "validation_loss": 0.018808677792549133,
          "train_metric": 0.0028849586378782988,
          "validation_metric": 0.0018943409668281674,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9106358269982593
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.016351179765479174,
          "train_loss": 0.01290273480117321,
          "validation_loss": 0.012090597301721573,
          "train_metric": 0.001299516181461513,
          "validation_metric": 0.0012177207972854376,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9244557699985307
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.008869607874657959,
          "train_loss": 0.007582868915051222,
          "validation_loss": 0.0061017367988824844,
          "train_metric": 0.0007637187372893095,
          "validation_metric": 0.0006145446677692235,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.938456692998443
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.0058883669480565,
          "train_loss": 0.005375618115067482,
          "validation_loss": 0.0037933329585939646,
          "train_metric": 0.000541412562597543,
          "validation_metric": 0.00038205055170692503,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9521059629987576
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.0042082878717337735,
          "train_loss": 0.0029756950680166483,
          "validation_loss": 0.0022573231253772974,
          "train_metric": 0.0002997011470142752,
          "validation_metric": 0.00022734931553713977,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9657960409986117
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.0032150754559552297,
          "train_loss": 0.0034050822723656893,
          "validation_loss": 0.002708876971155405,
          "train_metric": 0.00034294743090867996,
          "validation_metric": 0.0002728281542658806,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9796915729984903
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.003295619786513271,
          "train_loss": 0.003672040067613125,
          "validation_loss": 0.0038277199491858482,
          "train_metric": 0.0003698343934956938,
          "validation_metric": 0.0003855139366351068,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9933978669996577
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.003078979418205563,
          "train_loss": 0.0032184228766709566,
          "validation_loss": 0.0022972053848206997,
          "train_metric": 0.0003241478116251528,
          "validation_metric": 0.00023136608069762588,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.007410700998662
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.003616947193222586,
          "train_loss": 0.0023462269455194473,
          "validation_loss": 0.002109962049871683,
          "train_metric": 0.00023630340001545846,
          "validation_metric": 0.00021250764257274568,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0213481499995396
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0023250239974004216,
          "train_loss": 0.0038655574899166822,
          "validation_loss": 0.002962973900139332,
          "train_metric": 0.00038932482129894197,
          "validation_metric": 0.00029841987998224795,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0351115009998466
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.0029477911593858153,
          "train_loss": 0.0029394871089607477,
          "validation_loss": 0.0022885631769895554,
          "train_metric": 0.00029605437885038555,
          "validation_metric": 0.00023049568699207157,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0488922459990135
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.0034806174007826485,
          "train_loss": 0.00380074349232018,
          "validation_loss": 0.0030071388464421034,
          "train_metric": 0.0003827969776466489,
          "validation_metric": 0.0003028680512215942,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.0625507759996253
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.002605419354949845,
          "train_loss": 0.0018213038565590978,
          "validation_loss": 0.0016792693641036749,
          "train_metric": 0.00018343504052609205,
          "validation_metric": 0.00016912985302042216,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.0771361669994803
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.004080665821675211,
          "train_loss": 0.002199184149503708,
          "validation_loss": 0.0019179549999535084,
          "train_metric": 0.00022149377036839724,
          "validation_metric": 0.000193169413250871,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.0914020289983455
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.002720065040193731,
          "train_loss": 0.0019235775107517838,
          "validation_loss": 0.0016753959935158491,
          "train_metric": 0.00019373567192815244,
          "validation_metric": 0.00016873975982889533,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.105916384000011
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0023779427974659484,
          "train_loss": 0.0017264370108023286,
          "validation_loss": 0.0013534598983824253,
          "train_metric": 0.00017388039850629866,
          "validation_metric": 0.0001363155315630138,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.1203776949987514
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.0020156774007773492,
          "train_loss": 0.0016833902336657047,
          "validation_loss": 0.0012908440548926592,
          "train_metric": 0.00016954491729848087,
          "validation_metric": 0.0001300090953009203,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.1350246219990368
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0036628420784836635,
          "train_loss": 0.0045376671478152275,
          "validation_loss": 0.0038465620018541813,
          "train_metric": 0.0004570171586237848,
          "validation_metric": 0.0003874116519000381,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.148638922999453
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.002620893563289428,
          "train_loss": 0.0019287827890366316,
          "validation_loss": 0.00144688761793077,
          "train_metric": 0.00019425991922616959,
          "validation_metric": 0.00014572522195521742,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.162216234999505
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0022101903650764143,
          "train_loss": 0.002622372005134821,
          "validation_loss": 0.0017477951478213072,
          "train_metric": 0.0002641156897880137,
          "validation_metric": 0.00017603152082301676,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.2616030079989287
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.004260587873432087,
          "train_loss": 0.0030077865812927485,
          "validation_loss": 0.003109072567895055,
          "train_metric": 0.0003029332438018173,
          "validation_metric": 0.0003131343692075461,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.2755633999986458
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0022365374134096783,
          "train_loss": 0.002216682769358158,
          "validation_loss": 0.00154267274774611,
          "train_metric": 0.00022325613826978952,
          "validation_metric": 0.0001553723559482023,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.2892358819990477
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0018296962225576863,
          "train_loss": 0.0020384720992296934,
          "validation_loss": 0.001905615208670497,
          "train_metric": 0.00020530739857349545,
          "validation_metric": 0.00019192656327504665,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.3032582369996817
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0017216375963471364,
          "train_loss": 0.0014483369886875153,
          "validation_loss": 0.0011452959151938558,
          "train_metric": 0.00014587122132070363,
          "validation_metric": 0.00011535001249285415,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.3175910359987029
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.002812097238347633,
          "train_loss": 0.014162757433950901,
          "validation_loss": 0.010514027439057827,
          "train_metric": 0.0014264211058616638,
          "validation_metric": 0.0010589343728497624,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.331890139999814
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007845269508834463,
          "train_loss": 0.0029648931231349707,
          "validation_loss": 0.002757398411631584,
          "train_metric": 0.0002986131876241416,
          "validation_metric": 0.00027771503664553165,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.3458699279999564
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.004364003838418284,
          "train_loss": 0.008987737819552422,
          "validation_loss": 0.007578937336802483,
          "train_metric": 0.0009052121895365417,
          "validation_metric": 0.0007633228087797761,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.359642127999905
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.004419900164066348,
          "train_loss": 0.001808779314160347,
          "validation_loss": 0.0013286637840792537,
          "train_metric": 0.00018217363685835153,
          "validation_metric": 0.0001338181464234367,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.3734215599997697
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0031441365481441608,
          "train_loss": 0.0021038760896772146,
          "validation_loss": 0.001514333882369101,
          "train_metric": 0.00021189471590332687,
          "validation_metric": 0.00015251815784722567,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.3874631979997503
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.002240926074591698,
          "train_loss": 0.0018392369383946061,
          "validation_loss": 0.0014090447220951319,
          "train_metric": 0.00018524122424423695,
          "validation_metric": 0.00014191381342243403,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.4014958859988838
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.001856082963058725,
          "train_loss": 0.0015386184677481651,
          "validation_loss": 0.0014156489633023739,
          "train_metric": 0.0001549640146549791,
          "validation_metric": 0.0001425789960194379,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.4154267929989146
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0017993034671235364,
          "train_loss": 0.002309999894350767,
          "validation_loss": 0.0018170294351875782,
          "train_metric": 0.00023265468189492822,
          "validation_metric": 0.0001830045657698065,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.4293155589984963
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0013365714221436065,
          "train_loss": 0.0015817810781300068,
          "validation_loss": 0.0012222325894981623,
          "train_metric": 0.00015931121015455574,
          "validation_metric": 0.0001230988127645105,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.4425939689990628
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0014941570025257533,
          "train_loss": 0.0011872350005432963,
          "validation_loss": 0.0010791559470817447,
          "train_metric": 0.00011957396054640412,
          "validation_metric": 0.00010868861863855273,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.457169838999107
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0013925560051575303,
          "train_loss": 0.0013072044821456075,
          "validation_loss": 0.000965601415373385,
          "train_metric": 0.00013165685231797397,
          "validation_metric": 9.72518464550376e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.471976768998502
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0014475377829512581,
          "train_loss": 0.001517817610874772,
          "validation_loss": 0.0014482211554422975,
          "train_metric": 0.00015286901907529682,
          "validation_metric": 0.00014585955068469048,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.4857712849989184
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.001872633376478916,
          "train_loss": 0.0019470937550067902,
          "validation_loss": 0.0014618865679949522,
          "train_metric": 0.00019610414165072143,
          "validation_metric": 0.0001472358708269894,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.4995147849986097
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0054850954984431155,
          "train_loss": 0.0013190304161980748,
          "validation_loss": 0.001257438096217811,
          "train_metric": 0.0001328479265794158,
          "validation_metric": 0.00012664457608480006,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.5127765279994492
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.0064657146322133485,
          "train_loss": 0.0052958386950194836,
          "validation_loss": 0.004091070499271154,
          "train_metric": 0.000533377518877387,
          "validation_metric": 0.00041203771252185106,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.5262147779994848
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.004608973162248731,
          "train_loss": 0.004421045072376728,
          "validation_loss": 0.003347221063449979,
          "train_metric": 0.0004452714929357171,
          "validation_metric": 0.00033711985452100635,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.5399604689991975
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0037155583704588935,
          "train_loss": 0.001600156887434423,
          "validation_loss": 0.0012908277567476034,
          "train_metric": 0.00016116192273329943,
          "validation_metric": 0.0001300074509344995,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.5532996519996232
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0019795540592895122,
          "train_loss": 0.0014186494518071413,
          "validation_loss": 0.0011992478976026177,
          "train_metric": 0.00014288116653915495,
          "validation_metric": 0.00012078385043423623,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.5668167159983568
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0026520797437115107,
          "train_loss": 0.007639104966074228,
          "validation_loss": 0.004448641091585159,
          "train_metric": 0.0007693827501498163,
          "validation_metric": 0.00044805085053667426,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.5803137849998166
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.003698776454257313,
          "train_loss": 0.0014480710960924625,
          "validation_loss": 0.0010283075971528888,
          "train_metric": 0.0001458444312447682,
          "validation_metric": 0.00010356736311223358,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.5935007999996742
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0013876126940886024,
          "train_loss": 0.0011812495067715645,
          "validation_loss": 0.0009725693962536752,
          "train_metric": 0.0001189711329061538,
          "validation_metric": 9.79536198428832e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.6072934699986945
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0019841021130559966,
          "train_loss": 0.0021492906380444765,
          "validation_loss": 0.001738726976327598,
          "train_metric": 0.00021646871755365282,
          "validation_metric": 0.00017511819896753877,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.620646969999143
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0028180437220726162,
          "train_loss": 0.004712448921054602,
          "validation_loss": 0.0034331746865063906,
          "train_metric": 0.00047462055226787925,
          "validation_metric": 0.0003457767888903618,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.6342381869999372
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.002345712906389963,
          "train_loss": 0.0016680278349667788,
          "validation_loss": 0.001376044237986207,
          "train_metric": 0.0001679976558079943,
          "validation_metric": 0.00013859011232852936,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.6474477469982958
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0012541514261101838,
          "train_loss": 0.0010155471973121166,
          "validation_loss": 0.0007942775264382362,
          "train_metric": 0.00010228219616692513,
          "validation_metric": 7.999671652214602e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.66157821699926
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.001138269546572701,
          "train_loss": 0.0010479723569005728,
          "validation_loss": 0.0008517660899087787,
          "train_metric": 0.00010554794425843284,
          "validation_metric": 8.578676352044567e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.675336222999249
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0009965488243324216,
          "train_loss": 0.00098379363771528,
          "validation_loss": 0.0008038965752348304,
          "train_metric": 9.90840999293141e-05,
          "validation_metric": 8.096552483038977e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.6891607319994364
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0010749639877758455,
          "train_loss": 0.0010404292261227965,
          "validation_loss": 0.0009060397278517485,
          "train_metric": 0.00010478822514414787,
          "validation_metric": 9.125300857704133e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.7027903469988814
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.001460254143239581,
          "train_loss": 0.0019443311030045152,
          "validation_loss": 0.001366279786452651,
          "train_metric": 0.00019582589447963983,
          "validation_metric": 0.00013760670844931155,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.7165227299992694
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0012967810544068925,
          "train_loss": 0.001923888921737671,
          "validation_loss": 0.001730398740619421,
          "train_metric": 0.0001937670458573848,
          "validation_metric": 0.00017427944112569094,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.730172001000028
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0019328592552483315,
          "train_loss": 0.0038160188123583794,
          "validation_loss": 0.002962038852274418,
          "train_metric": 0.00038433546433225274,
          "validation_metric": 0.0002983256708830595,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.7435188999988895
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0017400491506123217,
          "train_loss": 0.0011966134188696742,
          "validation_loss": 0.0009719364461489022,
          "train_metric": 0.00012051851081196219,
          "validation_metric": 9.788988245418295e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.8408034749991202
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0017479341368016321,
          "train_loss": 0.003015235997736454,
          "validation_loss": 0.0029419991187751293,
          "train_metric": 0.00030368351144716144,
          "validation_metric": 0.0002963073493447155,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.854769399998986
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0018786116306728218,
          "train_loss": 0.0011003109393641353,
          "validation_loss": 0.0007198821404017508,
          "train_metric": 0.00011081926641054451,
          "validation_metric": 7.250389171531424e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.8690463919992908
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.001118400596169522,
          "train_loss": 0.0010641854023560882,
          "validation_loss": 0.000821062596514821,
          "train_metric": 0.00010718085832195356,
          "validation_metric": 8.26944233267568e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.8826925909997954
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0014151977466099197,
          "train_loss": 0.0010325670009478927,
          "validation_loss": 0.0008353047305718064,
          "train_metric": 0.00010399638267699629,
          "validation_metric": 8.412883471464738e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.8958520119995228
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0022217766472749645,
          "train_loss": 0.0025818648282438517,
          "validation_loss": 0.0016593376640230417,
          "train_metric": 0.0002600359730422497,
          "validation_metric": 0.00016712241631466895,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.9092533599996386
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0020864171729044756,
          "train_loss": 0.0021261502988636494,
          "validation_loss": 0.0019838493317365646,
          "train_metric": 0.00021413809736259282,
          "validation_metric": 0.00019980604702141136,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.9227650319990062
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0014630054174631368,
          "train_loss": 0.0009938398143276572,
          "validation_loss": 0.000675144896376878,
          "train_metric": 0.0001000959164230153,
          "validation_metric": 6.799811671953648e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.9369577759989625
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.001151106131146662,
          "train_loss": 0.0012100051390007138,
          "validation_loss": 0.0010119632352143526,
          "train_metric": 0.0001218672696268186,
          "validation_metric": 0.00010192122863372788,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.9505911999985983
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0018337278197577689,
          "train_loss": 0.002508733654394746,
          "validation_loss": 0.0015389523468911648,
          "train_metric": 0.0002526704629417509,
          "validation_metric": 0.00015499764413107187,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.9641109589992993
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0011815213028967264,
          "train_loss": 0.0011719854082912207,
          "validation_loss": 0.0011014487827196717,
          "train_metric": 0.00011803808592958376,
          "validation_metric": 0.00011093389184679836,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.9779247859987663
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0009116000474023167,
          "train_loss": 0.0008642240427434444,
          "validation_loss": 0.0006143442005850375,
          "train_metric": 8.704147330718115e-05,
          "validation_metric": 6.187448889249936e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.9926562619984907
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0008887615640560398,
          "train_loss": 0.001306250924244523,
          "validation_loss": 0.0010509275598451495,
          "train_metric": 0.00013156080967746675,
          "validation_metric": 0.00010584556730464101,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0061769309995725
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0014119302104518283,
          "train_loss": 0.0033885352313518524,
          "validation_loss": 0.0023170323111116886,
          "train_metric": 0.0003412808873690665,
          "validation_metric": 0.00023336302547249943,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0199814089992287
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.001290787953621475,
          "train_loss": 0.0008294096332974732,
          "validation_loss": 0.0006559633184224367,
          "train_metric": 8.353509474545717e-05,
          "validation_metric": 6.606622628169134e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0335890449987346
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0009402629475516733,
          "train_loss": 0.00081358797615394,
          "validation_loss": 0.0006201669457368553,
          "train_metric": 8.194160182029009e-05,
          "validation_metric": 6.246093835216016e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.046920249998948
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0016191280592465773,
          "train_loss": 0.0017546454910188913,
          "validation_loss": 0.0013112009037286043,
          "train_metric": 0.0001767214562278241,
          "validation_metric": 0.0001320593582931906,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.060268454999459
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.002896518899433431,
          "train_loss": 0.0031807313207536936,
          "validation_loss": 0.0036292378790676594,
          "train_metric": 0.0003203515661880374,
          "validation_metric": 0.00036552356323227286,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.073947019998741
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.002246349489723798,
          "train_loss": 0.0020763855427503586,
          "validation_loss": 0.0011937487870454788,
          "train_metric": 0.00020912595209665596,
          "validation_metric": 0.0001202300118166022,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0878989339998952
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.001525375646451721,
          "train_loss": 0.0011201828019693494,
          "validation_loss": 0.0008319831686094403,
          "train_metric": 0.00011282070772722363,
          "validation_metric": 8.379429345950484e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.101556551999238
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.001263603935512947,
          "train_loss": 0.0016336179105564952,
          "validation_loss": 0.0013926696265116334,
          "train_metric": 0.00016453202988486737,
          "validation_metric": 0.00014026457211002707,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.115093888998672
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.001392403897625627,
          "train_loss": 0.0013046482345089316,
          "validation_loss": 0.0009021891164593399,
          "train_metric": 0.00013139938528183848,
          "validation_metric": 9.086518548429012e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.128175411999109
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0015674877722631209,
          "train_loss": 0.0010566188720986247,
          "validation_loss": 0.0006733585614711046,
          "train_metric": 0.00010641875996952876,
          "validation_metric": 6.781821139156818e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.1417061309985
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0010538383594393963,
          "train_loss": 0.0007539812941104174,
          "validation_loss": 0.0005744522204622626,
          "train_metric": 7.593823102070019e-05,
          "validation_metric": 5.785672692582011e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.156172504999631
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0009630692547943909,
          "train_loss": 0.0007359401206485927,
          "validation_loss": 0.000599755730945617,
          "train_metric": 7.412119884975255e-05,
          "validation_metric": 6.0405207477742806e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.172340464998342
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0007728676646365784,
          "train_loss": 0.0007416592561639845,
          "validation_loss": 0.0006294507766142488,
          "train_metric": 7.469720003427938e-05,
          "validation_metric": 6.339597894111648e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.2044495199988887
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0007630875843460672,
          "train_loss": 0.0007877536118030548,
          "validation_loss": 0.0007380867609754205,
          "train_metric": 7.933966116979718e-05,
          "validation_metric": 7.433738937834278e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.2532232260000455
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0013088174546282971,
          "train_loss": 0.0016405705828219652,
          "validation_loss": 0.0012900199508294463,
          "train_metric": 0.00016523225349374115,
          "validation_metric": 0.00012992610572837293,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.2781098269988433
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0012000570604868699,
          "train_loss": 0.0007895866874605417,
          "validation_loss": 0.0007408070960082114,
          "train_metric": 7.952428131829947e-05,
          "validation_metric": 7.461137283826247e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.298704528999224
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0008329929696628824,
          "train_loss": 0.00073792296461761,
          "validation_loss": 0.0004965517437085509,
          "train_metric": 7.432090933434665e-05,
          "validation_metric": 5.001086901756935e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.3197464249988116
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0008067919116001576,
          "train_loss": 0.0007859061006456614,
          "validation_loss": 0.0005208409857004881,
          "train_metric": 7.915359310572967e-05,
          "validation_metric": 5.24571914866101e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.3403143449995696
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0008896638810256263,
          "train_loss": 0.0007983163231983781,
          "validation_loss": 0.0005726370727643371,
          "train_metric": 8.040350076043978e-05,
          "validation_metric": 5.7673903938848525e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.3602100419993803
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0009440040212211898,
          "train_loss": 0.0006803326541557908,
          "validation_loss": 0.000577577156946063,
          "train_metric": 6.852061051176861e-05,
          "validation_metric": 5.8171452110400423e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.3806548419997853
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0008375902725674678,
          "train_loss": 0.0006913222023285925,
          "validation_loss": 0.0006149025866761804,
          "train_metric": 6.962744373595342e-05,
          "validation_metric": 6.193073932081461e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.4023976099997526
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0007404214065900305,
          "train_loss": 0.0006538117304444313,
          "validation_loss": 0.0005017479415982962,
          "train_metric": 6.58495191601105e-05,
          "validation_metric": 5.053421773482114e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.42223770199962
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0007067936621751869,
          "train_loss": 0.0007418058812618256,
          "validation_loss": 0.0005111866048537195,
          "train_metric": 7.471197750419378e-05,
          "validation_metric": 5.148484342498705e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.4428683879996242
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0007714972107351059,
          "train_loss": 0.0007295311079360545,
          "validation_loss": 0.0005250779213383794,
          "train_metric": 7.347569771809503e-05,
          "validation_metric": 5.288392276270315e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.6012986469995667
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0007271570793818682,
          "train_loss": 0.0006390250637196004,
          "validation_loss": 0.0005283074569888413,
          "train_metric": 6.436026887968183e-05,
          "validation_metric": 5.3209190809866413e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.62374862499928
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0008120228048937861,
          "train_loss": 0.0006319767562672496,
          "validation_loss": 0.0004922047373838723,
          "train_metric": 6.36503827990964e-05,
          "validation_metric": 4.95730600960087e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.638202181999077
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0007135446003303514,
          "train_loss": 0.0006187251419760287,
          "validation_loss": 0.0005095007363706827,
          "train_metric": 6.23157320660539e-05,
          "validation_metric": 5.131504076416604e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.6523923219992867
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0006890664699312765,
          "train_loss": 0.000653604045510292,
          "validation_loss": 0.0005648205406032503,
          "train_metric": 6.582860078196973e-05,
          "validation_metric": 5.688665260095149e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.6658076509993407
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0006978870278544491,
          "train_loss": 0.000668891763780266,
          "validation_loss": 0.0004657748795580119,
          "train_metric": 6.7368324380368e-05,
          "validation_metric": 4.6911140088923275e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.6804906869983824
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0008951014151534764,
          "train_loss": 0.0009096621070057154,
          "validation_loss": 0.0005645695491693914,
          "train_metric": 9.16178323677741e-05,
          "validation_metric": 5.686137592419982e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.6945961869987514
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0010316480838810094,
          "train_loss": 0.0007710973732173443,
          "validation_loss": 0.0007840674370527267,
          "train_metric": 7.766210910631344e-05,
          "validation_metric": 7.896839088061824e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.708863616999224
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0012377157199807698,
          "train_loss": 0.0009587572421878576,
          "validation_loss": 0.0009327777661383152,
          "train_metric": 9.656252223066986e-05,
          "validation_metric": 9.394596418133005e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.7222148339988053
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.000721626689482946,
          "train_loss": 0.0007017616298981011,
          "validation_loss": 0.0004969692090526223,
          "train_metric": 7.067886326694861e-05,
          "validation_metric": 5.005292041460052e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.7364974140000413
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0007075288331179763,
          "train_loss": 0.0006003317539580166,
          "validation_loss": 0.00044900289503857493,
          "train_metric": 6.04632223257795e-05,
          "validation_metric": 4.5221924665383995e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.7512892819995614
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0009192513298330596,
          "train_loss": 0.0011007862631231546,
          "validation_loss": 0.0008914555073715746,
          "train_metric": 0.00011086717859143391,
          "validation_metric": 8.978414552984759e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.7656453089985007
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0007254140746226767,
          "train_loss": 0.0006168916006572545,
          "validation_loss": 0.0005967763718217611,
          "train_metric": 6.213106098584831e-05,
          "validation_metric": 6.010513607179746e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.7800972429995454
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0006394125721271848,
          "train_loss": 0.0005713716964237392,
          "validation_loss": 0.0004783312324434519,
          "train_metric": 5.7546461903257295e-05,
          "validation_metric": 4.817576700588688e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.813414673999432
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0005919034520047717,
          "train_loss": 0.0005738321342505515,
          "validation_loss": 0.0004173588240519166,
          "train_metric": 5.7794277381617576e-05,
          "validation_metric": 4.203485150355846e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.8450207549994957
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0005623415563604794,
          "train_loss": 0.0008316182065755129,
          "validation_loss": 0.0008234010310843587,
          "train_metric": 8.375753532163799e-05,
          "validation_metric": 8.29299315228127e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.877994829999807
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0007150471265049418,
          "train_loss": 0.0006346557638607919,
          "validation_loss": 0.0005659281741827726,
          "train_metric": 6.392020441126078e-05,
          "validation_metric": 5.699821485904977e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.9015838939994865
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0005973811012154329,
          "train_loss": 0.0005512955831363797,
          "validation_loss": 0.0004147751023992896,
          "train_metric": 5.552446600631811e-05,
          "validation_metric": 4.177462687948719e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.9165721499994106
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0005734397309424821,
          "train_loss": 0.0005500433035194874,
          "validation_loss": 0.0004248306795489043,
          "train_metric": 5.539834819501266e-05,
          "validation_metric": 4.278739288565703e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.930929868000021
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0005412529753812123,
          "train_loss": 0.000643218751065433,
          "validation_loss": 0.0006517860456369817,
          "train_metric": 6.478264549514279e-05,
          "validation_metric": 6.56455013086088e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.9452650289986195
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.001003230032438296,
          "train_loss": 0.0010066234972327948,
          "validation_loss": 0.0008854818297550082,
          "train_metric": 0.00010138342622667551,
          "validation_metric": 8.918248204281554e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.9590789750000113
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0008094745244306978,
          "train_loss": 0.0007328369538299739,
          "validation_loss": 0.0006760487449355423,
          "train_metric": 7.380865281447768e-05,
          "validation_metric": 6.808913894928992e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.972180816999753
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.000795601328718476,
          "train_loss": 0.0019932317081838846,
          "validation_loss": 0.001785092637874186,
          "train_metric": 0.00020075100474059582,
          "validation_metric": 0.00017978799587581307,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 2.9860863750000135
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.003982776901466423,
          "train_loss": 0.005400818772614002,
          "validation_loss": 0.0036137900315225124,
          "train_metric": 0.0005439506494440138,
          "validation_metric": 0.0003639676724560559,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.0004945209984726
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0022029654173820745,
          "train_loss": 0.001263036043383181,
          "validation_loss": 0.0009959045564755797,
          "train_metric": 0.0001272083754884079,
          "validation_metric": 0.00010030385601567104,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.014081024999541
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0013887511922803242,
          "train_loss": 0.0009063013130798936,
          "validation_loss": 0.0009328422020189464,
          "train_metric": 9.127935481956229e-05,
          "validation_metric": 9.395246161147952e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.0278713019997667
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0007176695453381399,
          "train_loss": 0.0006058876169845462,
          "validation_loss": 0.0004526523989625275,
          "train_metric": 6.102277620811947e-05,
          "validation_metric": 4.558949513011612e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.0419434609993914
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0006477558908954961,
          "train_loss": 0.0005650977254845202,
          "validation_loss": 0.0004101101076230407,
          "train_metric": 5.691457772627473e-05,
          "validation_metric": 4.130478191655129e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.056409161999909
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.000670060535412631,
          "train_loss": 0.0007781878812238574,
          "validation_loss": 0.0004984083352610469,
          "train_metric": 7.837622979423031e-05,
          "validation_metric": 5.0197857490275055e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.070767804998468
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0007500628398702247,
          "train_loss": 0.0005620337906293571,
          "validation_loss": 0.00041411624988541007,
          "train_metric": 5.660597889800556e-05,
          "validation_metric": 4.1708270146045834e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.084864727999957
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0005836155305587454,
          "train_loss": 0.0005472018383443356,
          "validation_loss": 0.00038942720857448876,
          "train_metric": 5.511216659215279e-05,
          "validation_metric": 3.9221678889589384e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.099298209999688
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.000623301673840615,
          "train_loss": 0.0005659356247633696,
          "validation_loss": 0.0005244302446953952,
          "train_metric": 5.6998957006726414e-05,
          "validation_metric": 5.28186974406708e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.1134798619987123
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0006820034814154496,
          "train_loss": 0.0006371367489919066,
          "validation_loss": 0.0006004131864756346,
          "train_metric": 6.417008262360469e-05,
          "validation_metric": 6.047141505405307e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.1273300149987335
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.00057037858368858,
          "train_loss": 0.0005716446903534234,
          "validation_loss": 0.0005395550979301333,
          "train_metric": 5.75739613850601e-05,
          "validation_metric": 5.434201375464909e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.1414869609998277
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0005781241870863596,
          "train_loss": 0.0005461816326715052,
          "validation_loss": 0.00045858975499868393,
          "train_metric": 5.500942643266171e-05,
          "validation_metric": 4.61874769825954e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.155625756999143
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.000547896164789563,
          "train_loss": 0.0013942391378805041,
          "validation_loss": 0.0011870099697262049,
          "train_metric": 0.00014042267866898328,
          "validation_metric": 0.00011955130321439356,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.1698217229986767
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0007566760159534169,
          "train_loss": 0.0005671782419085503,
          "validation_loss": 0.000506838783621788,
          "train_metric": 5.7124121667584404e-05,
          "validation_metric": 5.104693991597742e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.2735468779992516
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0005494132219610037,
          "train_loss": 0.0004830503021366894,
          "validation_loss": 0.0004159896052442491,
          "train_metric": 4.8651058023096994e-05,
          "validation_metric": 4.189694664091803e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.2897265029987466
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0005338625578588108,
          "train_loss": 0.0007496257312595844,
          "validation_loss": 0.0007941900985315442,
          "train_metric": 7.549955626018345e-05,
          "validation_metric": 7.998792716534808e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.304594356999587
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.000815714140117052,
          "train_loss": 0.0006566544179804623,
          "validation_loss": 0.0003837425902020186,
          "train_metric": 6.613582809222862e-05,
          "validation_metric": 3.8649151974823326e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3204720319990884
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0006869183707749471,
          "train_loss": 0.0006060826708562672,
          "validation_loss": 0.00037917433655820787,
          "train_metric": 6.104243220761418e-05,
          "validation_metric": 3.818904951913282e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3365524179989734
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0005823457850056002,
          "train_loss": 0.0005655809072777629,
          "validation_loss": 0.0004009166150353849,
          "train_metric": 5.696323569281958e-05,
          "validation_metric": 4.037885082652792e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3853017589990486
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0005157726445759181,
          "train_loss": 0.00046410958748310804,
          "validation_loss": 0.0004119148652534932,
          "train_metric": 4.674341471400112e-05,
          "validation_metric": 4.1486557165626436e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.421396988998822
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.0005552327647819766,
          "train_loss": 0.0006200886564329267,
          "validation_loss": 0.00037273025372996926,
          "train_metric": 6.2453058490064e-05,
          "validation_metric": 3.754002682399005e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.455450462999579
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0004995061026420444,
          "train_loss": 0.0004813708656001836,
          "validation_loss": 0.00045683892676606774,
          "train_metric": 4.848191383644007e-05,
          "validation_metric": 4.6011140511836857e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4882445159983035
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0006330435107884114,
          "train_loss": 0.0007435770821757615,
          "validation_loss": 0.0007447457755915821,
          "train_metric": 7.48903548810631e-05,
          "validation_metric": 7.500806532334536e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.520639496999138
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0006641903128183912,
          "train_loss": 0.0008348728297278285,
          "validation_loss": 0.000778575602453202,
          "train_metric": 8.408531721215695e-05,
          "validation_metric": 7.841528713470325e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.5517551789998834
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0006401800219464349,
          "train_loss": 0.00112800276838243,
          "validation_loss": 0.0009005216415971518,
          "train_metric": 0.0001136082864832133,
          "validation_metric": 9.069724183063954e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.586575041999822
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0008201971122616669,
          "train_loss": 0.000792064645793289,
          "validation_loss": 0.000698256422765553,
          "train_metric": 7.977385394042358e-05,
          "validation_metric": 7.032582652755082e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.618061723998835
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.001020786776280147,
          "train_loss": 0.0011857454665005207,
          "validation_loss": 0.0006720059318467975,
          "train_metric": 0.00011942394485231489,
          "validation_metric": 6.768197636120021e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.652930200998526
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0010134124404430622,
          "train_loss": 0.001211515860632062,
          "validation_loss": 0.000975983974058181,
          "train_metric": 0.00012201943900436163,
          "validation_metric": 9.82975252554752e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.683989571998609
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0007228523900266737,
          "train_loss": 0.0008792271837592125,
          "validation_loss": 0.0007564472034573555,
          "train_metric": 8.855254418449476e-05,
          "validation_metric": 7.618659583386034e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7184787269998196
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0006868507862236584,
          "train_loss": 0.0005417423672042787,
          "validation_loss": 0.00043250180897302926,
          "train_metric": 5.456230428535491e-05,
          "validation_metric": 4.355999772087671e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7497052369999437
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0005358506059565116,
          "train_loss": 0.000732689630240202,
          "validation_loss": 0.0004578224034048617,
          "train_metric": 7.379380986094475e-05,
          "validation_metric": 4.611019539879635e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.783899402998941
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.00058493761025602,
          "train_loss": 0.000660723599139601,
          "validation_loss": 0.0008176888804882765,
          "train_metric": 6.654565368080512e-05,
          "validation_metric": 8.235462155425921e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.8189988769991032
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0006076570989534957,
          "train_loss": 0.00046097167069092393,
          "validation_loss": 0.0004909292911179364,
          "train_metric": 4.642737621907145e-05,
          "validation_metric": 4.94446067023091e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.842392974998802
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0004607835016940953,
          "train_loss": 0.00043785443995147943,
          "validation_loss": 0.0003484237822704017,
          "train_metric": 4.40990952483844e-05,
          "validation_metric": 3.509196903905831e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.858681511999748
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.0004744569878312177,
          "train_loss": 0.0004554018669296056,
          "validation_loss": 0.0003306979197077453,
          "train_metric": 4.586640716297552e-05,
          "validation_metric": 3.330668550916016e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.874687224999434
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.00046554185973946005,
          "train_loss": 0.00045311416033655405,
          "validation_loss": 0.00035722393658943474,
          "train_metric": 4.563600305118598e-05,
          "validation_metric": 3.597828617785126e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8904253339987918
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0005746487495343899,
          "train_loss": 0.00047855067532509565,
          "validation_loss": 0.0003148105461150408,
          "train_metric": 4.819787136511877e-05,
          "validation_metric": 3.1706564186606556e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9063759169985133
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.0004498844546105829,
          "train_loss": 0.00041475496254861355,
          "validation_loss": 0.0003851274377666414,
          "train_metric": 4.1772600525291637e-05,
          "validation_metric": 3.87886211683508e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9210475539985055
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0005899745683564106,
          "train_loss": 0.0005434637423604727,
          "validation_loss": 0.0005797285703010857,
          "train_metric": 5.4735675803385675e-05,
          "validation_metric": 5.838813740410842e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.93479098699936
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.00046762269448663574,
          "train_loss": 0.0004214799264445901,
          "validation_loss": 0.00038602884160354733,
          "train_metric": 4.244991214363836e-05,
          "validation_metric": 3.887941056746058e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9490472149991547
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0005697526703443145,
          "train_loss": 0.00041759078158065677,
          "validation_loss": 0.00040157060720957816,
          "train_metric": 4.205821824143641e-05,
          "validation_metric": 4.044472370878793e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9632590309993248
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.00046158706845744746,
          "train_loss": 0.0004591844044625759,
          "validation_loss": 0.000461281044408679,
          "train_metric": 4.624736902769655e-05,
          "validation_metric": 4.6458539145532995e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.977548807999483
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0004429160389918252,
          "train_loss": 0.00040060980245471,
          "validation_loss": 0.0003516266297083348,
          "train_metric": 4.0347949834540486e-05,
          "validation_metric": 3.541454134392552e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9922542049989715
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.00041881856213876745,
          "train_loss": 0.0005992746446281672,
          "validation_loss": 0.0006487033679150045,
          "train_metric": 6.0356749600032344e-05,
          "validation_metric": 6.53350361972116e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.00658128699979
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0005972120961814653,
          "train_loss": 0.0006366533925756812,
          "validation_loss": 0.0006278879591263831,
          "train_metric": 6.412139191525057e-05,
          "validation_metric": 6.323857815004885e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.021068699999887
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0006097866862546653,
          "train_loss": 0.0005167157505638897,
          "validation_loss": 0.0005880665848962963,
          "train_metric": 5.204171611694619e-05,
          "validation_metric": 5.9227910242043436e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.035786991998975
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.00043761560846178327,
          "train_loss": 0.00038862760993652046,
          "validation_loss": 0.0003382504510227591,
          "train_metric": 3.91411485907156e-05,
          "validation_metric": 3.406735049793497e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.050379437998345
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.000437162455455109,
          "train_loss": 0.0003931705141440034,
          "validation_loss": 0.0003335249493829906,
          "train_metric": 3.959869354730472e-05,
          "validation_metric": 3.3591415558476e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.0667264129988325
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0003986622359661851,
          "train_loss": 0.0003838477423414588,
          "validation_loss": 0.000369026034604758,
          "train_metric": 3.865973849315196e-05,
          "validation_metric": 3.71669557353016e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.080954099999872
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0003895975423802156,
          "train_loss": 0.000380951096303761,
          "validation_loss": 0.0003599485207814723,
          "train_metric": 3.8368001696653664e-05,
          "validation_metric": 3.625269528129138e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.184892347999266
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0004010049442513264,
          "train_loss": 0.00038437574403360486,
          "validation_loss": 0.0003351276973262429,
          "train_metric": 3.871291482937522e-05,
          "validation_metric": 3.375283267814666e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.198876269998436
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.00041521532875776757,
          "train_loss": 0.000396302726585418,
          "validation_loss": 0.0004246995667926967,
          "train_metric": 3.991415724158287e-05,
          "validation_metric": 4.2774187022587284e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.2125378099990485
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0004157355169809307,
          "train_loss": 0.00045669457176700234,
          "validation_loss": 0.0003152732679154724,
          "train_metric": 4.599660314852372e-05,
          "validation_metric": 3.175317033310421e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.226441717999478
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0004196903528281837,
          "train_loss": 0.000386418221751228,
          "validation_loss": 0.00039148260839283466,
          "train_metric": 3.891862797900103e-05,
          "validation_metric": 3.942869443562813e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.239875152999957
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0004131152199988719,
          "train_loss": 0.000426965591032058,
          "validation_loss": 0.00047346146311610937,
          "train_metric": 4.3002411985071376e-05,
          "validation_metric": 4.7685309255030006e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.253601329999583
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.00040633550634083804,
          "train_loss": 0.00038776747533120215,
          "validation_loss": 0.0003243548271711916,
          "train_metric": 3.9054521039361134e-05,
          "validation_metric": 3.2667830964783207e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.267094339998948
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0003861074146698229,
          "train_loss": 0.00041193337528966367,
          "validation_loss": 0.00044810122926719487,
          "train_metric": 4.1488423448754475e-05,
          "validation_metric": 4.513111343840137e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.280883421999533
        }
      ],
      "best": {
        "epoch": 189,
        "updates": 3024
      },
      "best_validation_metric": 3.1706564186606556e-05,
      "config": {
        "label": "LayerNorm + silu + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.005630858210486795
    },
    "14985df6b90f_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.2918598894029856,
          "train_loss": 0.08864232152700424,
          "validation_loss": 0.07799669355154037,
          "train_metric": 0.008927728049457073,
          "validation_metric": 0.007855541072785854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05241342800036364
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07088851742446423,
          "train_loss": 0.06481601297855377,
          "validation_loss": 0.0761234313249588,
          "train_metric": 0.006528032012283802,
          "validation_metric": 0.007666872814297676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0741382190008153
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.042878184583969414,
          "train_loss": 0.037330325692892075,
          "validation_loss": 0.03774002939462662,
          "train_metric": 0.003759773913770914,
          "validation_metric": 0.0038010375574231148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1135377709997556
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.035872515523806214,
          "train_loss": 0.03026277758181095,
          "validation_loss": 0.029498307034373283,
          "train_metric": 0.0030479563865810633,
          "validation_metric": 0.0029709613882005215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13598972799991316
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.03364978532772511,
          "train_loss": 0.04825786501169205,
          "validation_loss": 0.0387478843331337,
          "train_metric": 0.004860354587435722,
          "validation_metric": 0.0039025451987981796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15667419000055816
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.021768706443253905,
          "train_loss": 0.01536557637155056,
          "validation_loss": 0.012449638918042183,
          "train_metric": 0.0015475646359845996,
          "validation_metric": 0.001253881840966642,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1786722169999848
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.016361093265004456,
          "train_loss": 0.03012162074446678,
          "validation_loss": 0.022713040933012962,
          "train_metric": 0.003033739048987627,
          "validation_metric": 0.0022875743452459574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19894348600064404
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.015725061297416687,
          "train_loss": 0.02052845060825348,
          "validation_loss": 0.0129659753292799,
          "train_metric": 0.002067550551146269,
          "validation_metric": 0.0013058854965493083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2188367199996719
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.03412104261224158,
          "train_loss": 0.0281328447163105,
          "validation_loss": 0.02257581613957882,
          "train_metric": 0.0028334371745586395,
          "validation_metric": 0.0022737535182386637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2389256900005421
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.0245996072771959,
          "train_loss": 0.014161459170281887,
          "validation_loss": 0.01212075911462307,
          "train_metric": 0.001426290487870574,
          "validation_metric": 0.001220758305862546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2603622210008325
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.01971524575492367,
          "train_loss": 0.013194844126701355,
          "validation_loss": 0.01390079502016306,
          "train_metric": 0.0013289365451782942,
          "validation_metric": 0.0014000372029840946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27976405899971724
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.016061629517935216,
          "train_loss": 0.009087281301617622,
          "validation_loss": 0.00776623236015439,
          "train_metric": 0.0009152375860139728,
          "validation_metric": 0.000782186514697969,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30087182900024345
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.016754603464505635,
          "train_loss": 0.011239910498261452,
          "validation_loss": 0.008611350320279598,
          "train_metric": 0.0011320423800498247,
          "validation_metric": 0.0008673035772517323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3323373059993173
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.014856431516818702,
          "train_loss": 0.011729576624929905,
          "validation_loss": 0.00934813916683197,
          "train_metric": 0.0011813601013273,
          "validation_metric": 0.0009415103122591972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3686320689994318
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.022869698019349016,
          "train_loss": 0.03948310762643814,
          "validation_loss": 0.036755409091711044,
          "train_metric": 0.003976593725383282,
          "validation_metric": 0.003701870096847415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39040597200073535
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.021364092215662822,
          "train_loss": 0.018701668828725815,
          "validation_loss": 0.017138896510004997,
          "train_metric": 0.001883563818410039,
          "validation_metric": 0.0017261671600863338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4121499989996664
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.011553974603884853,
          "train_loss": 0.008492782711982727,
          "validation_loss": 0.006388961803168058,
          "train_metric": 0.0008553619845770299,
          "validation_metric": 0.000643473002128303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4342916999994486
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.011279565369477496,
          "train_loss": 0.01520159374922514,
          "validation_loss": 0.010612648911774158,
          "train_metric": 0.0015310489106923342,
          "validation_metric": 0.0010688671609386802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4552436400008446
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.013181203394196928,
          "train_loss": 0.0082817068323493,
          "validation_loss": 0.009010737761855125,
          "train_metric": 0.0008341032662428916,
          "validation_metric": 0.0009075284469872713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47507078899980115
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.008930652780691162,
          "train_loss": 0.01041437592357397,
          "validation_loss": 0.012514843605458736,
          "train_metric": 0.0010488976258784533,
          "validation_metric": 0.0012604491785168648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4978744159998314
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.007346003723796457,
          "train_loss": 0.005623728036880493,
          "validation_loss": 0.004434825852513313,
          "train_metric": 0.0005664012278430164,
          "validation_metric": 0.00044665936729870737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5192945180006063
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.01089118086383678,
          "train_loss": 0.013976730406284332,
          "validation_loss": 0.010552308522164822,
          "train_metric": 0.0014076848747208714,
          "validation_metric": 0.0010627898154780269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5391502509992279
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01286614901619032,
          "train_loss": 0.012942020781338215,
          "validation_loss": 0.01053493283689022,
          "train_metric": 0.001303472905419767,
          "validation_metric": 0.0010610398603603244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5596752229994308
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.01680598038365133,
          "train_loss": 0.007113293278962374,
          "validation_loss": 0.005545926280319691,
          "train_metric": 0.0007164248963817954,
          "validation_metric": 0.0005585654289461672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5799424649994762
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.009716997010400519,
          "train_loss": 0.007082410156726837,
          "validation_loss": 0.0067711095325648785,
          "train_metric": 0.0007133144536055624,
          "validation_metric": 0.0006819613045081496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6002599999992526
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.007690569633268751,
          "train_loss": 0.005337899550795555,
          "validation_loss": 0.004222970921546221,
          "train_metric": 0.0005376136396080256,
          "validation_metric": 0.0004253221850376576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6226815070003795
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.015077051655680407,
          "train_loss": 0.024322792887687683,
          "validation_loss": 0.020289147272706032,
          "train_metric": 0.0024497027043253183,
          "validation_metric": 0.0020434486214071512,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6440793410001788
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01614713718299754,
          "train_loss": 0.010776746086776257,
          "validation_loss": 0.009896081872284412,
          "train_metric": 0.0010853944113478065,
          "validation_metric": 0.0009966972284018993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6642950240002392
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.009654254477936774,
          "train_loss": 0.01019070390611887,
          "validation_loss": 0.007867228239774704,
          "train_metric": 0.0010263704461976886,
          "validation_metric": 0.0007923583616502583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6873472450006375
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.009363326942548156,
          "train_loss": 0.007015187293291092,
          "validation_loss": 0.005033033899962902,
          "train_metric": 0.0007065440295264125,
          "validation_metric": 0.0005069086910225451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7068306050005049
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.01186647597933188,
          "train_loss": 0.036512408405542374,
          "validation_loss": 0.031734712421894073,
          "train_metric": 0.0036773961037397385,
          "validation_metric": 0.0031962040811777115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.727230258000418
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01852813793811947,
          "train_loss": 0.00820240005850792,
          "validation_loss": 0.007192524150013924,
          "train_metric": 0.0008261156617663801,
          "validation_metric": 0.0007244047010317445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7469982349994098
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.008491960179526359,
          "train_loss": 0.005819373764097691,
          "validation_loss": 0.004808956757187843,
          "train_metric": 0.0005861060344614089,
          "validation_metric": 0.00048434047494083643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7670767830004479
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.008550145721528679,
          "train_loss": 0.007180631626397371,
          "validation_loss": 0.00788583792746067,
          "train_metric": 0.0007232069037854671,
          "validation_metric": 0.000794232648331672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7874411869997857
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.009529343107715249,
          "train_loss": 0.013074944727122784,
          "validation_loss": 0.012273103930056095,
          "train_metric": 0.0013168605510145426,
          "validation_metric": 0.0012361020781099796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8075568690001091
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.026141693306271918,
          "train_loss": 0.037616487592458725,
          "validation_loss": 0.03295835480093956,
          "train_metric": 0.0037885946221649647,
          "validation_metric": 0.0033194448333233595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.828301644000021
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.01746085361810401,
          "train_loss": 0.013114810921251774,
          "validation_loss": 0.012586081400513649,
          "train_metric": 0.0013208757154643536,
          "validation_metric": 0.0012676240876317024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8485411220008245
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.010132260547834449,
          "train_loss": 0.007585165090858936,
          "validation_loss": 0.007952830754220486,
          "train_metric": 0.0007639501127414405,
          "validation_metric": 0.0008009800221771002,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8687823159998516
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.007146771589759737,
          "train_loss": 0.0056703523732721806,
          "validation_loss": 0.00408181780949235,
          "train_metric": 0.0005710971308872104,
          "validation_metric": 0.0004111057205591351,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.8903766259991244
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.006550184145453386,
          "train_loss": 0.007307811640202999,
          "validation_loss": 0.005809319671243429,
          "train_metric": 0.0007360160816460848,
          "validation_metric": 0.0005850933375768363,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9108130490003532
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.004882334527792409,
          "train_loss": 0.0038950606249272823,
          "validation_loss": 0.0029174494557082653,
          "train_metric": 0.00039229626418091357,
          "validation_metric": 0.0002938348043244332,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9321656259999145
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.00438209627463948,
          "train_loss": 0.003544508945196867,
          "validation_loss": 0.002924832981079817,
          "train_metric": 0.00035698997089639306,
          "validation_metric": 0.0002945784945040941,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9529277910005476
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.004237340333929751,
          "train_loss": 0.003243435872718692,
          "validation_loss": 0.0024500698782503605,
          "train_metric": 0.0003266669518779963,
          "validation_metric": 0.0002467620652168989,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9746227710002131
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.0033664343573036604,
          "train_loss": 0.0030133433174341917,
          "validation_loss": 0.0030486444011330605,
          "train_metric": 0.0003034929104615003,
          "validation_metric": 0.00030704832170158625,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9953961449991766
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.004683814971940592,
          "train_loss": 0.004118543118238449,
          "validation_loss": 0.002932007657364011,
          "train_metric": 0.0004148045554757118,
          "validation_metric": 0.00029530105530284345,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0159902089999377
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.002896019388572313,
          "train_loss": 0.0023314848076552153,
          "validation_loss": 0.0018630980048328638,
          "train_metric": 0.00023481860989704728,
          "validation_metric": 0.00018764441483654082,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0377998460007802
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0027481613724376075,
          "train_loss": 0.0026821109931916,
          "validation_loss": 0.0030792420729994774,
          "train_metric": 0.0002701324119698256,
          "validation_metric": 0.00031013000989332795,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0588523690003058
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.002649345431564143,
          "train_loss": 0.003925068769603968,
          "validation_loss": 0.004198695998638868,
          "train_metric": 0.0003953185223508626,
          "validation_metric": 0.0004228772595524788,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0791425900006288
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.003079888985666912,
          "train_loss": 0.002275755163282156,
          "validation_loss": 0.0015838653780519962,
          "train_metric": 0.000229205732466653,
          "validation_metric": 0.0001595211069798097,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0996248129995365
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.002326218116650125,
          "train_loss": 0.00522620091214776,
          "validation_loss": 0.003146749921143055,
          "train_metric": 0.0005263639031909406,
          "validation_metric": 0.00031692913034930825,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.1196330900002067
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.00480363740643952,
          "train_loss": 0.004983453545719385,
          "validation_loss": 0.0034358741249889135,
          "train_metric": 0.0005019151722081006,
          "validation_metric": 0.0003460486768744886,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.139922991000276
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0028441566828405485,
          "train_loss": 0.002820125548169017,
          "validation_loss": 0.0023110825568437576,
          "train_metric": 0.00028403272153809667,
          "validation_metric": 0.00023276376305148005,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.1609976689996984
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.001936301625391934,
          "train_loss": 0.0018195986049249768,
          "validation_loss": 0.0015617499593645334,
          "train_metric": 0.00018326332792639732,
          "validation_metric": 0.00015729373262729496,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.1829759609991015
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0017413092791684903,
          "train_loss": 0.0031077933963388205,
          "validation_loss": 0.003102565184235573,
          "train_metric": 0.0003130055556539446,
          "validation_metric": 0.00031247898004949093,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2032085070004541
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.004894323144981172,
          "train_loss": 0.004866339731961489,
          "validation_loss": 0.0039683375507593155,
          "train_metric": 0.0004901199135929346,
          "validation_metric": 0.000399676471715793,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2234061079998355
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0041966439530369826,
          "train_loss": 0.00346235278993845,
          "validation_loss": 0.0025353501550853252,
          "train_metric": 0.00034871548996306956,
          "validation_metric": 0.00025535112945362926,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.243466570000237
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.0034974187510670163,
          "train_loss": 0.0026655583642423153,
          "validation_loss": 0.0022552325390279293,
          "train_metric": 0.00026846525724977255,
          "validation_metric": 0.00022713873477187008,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2640142830005061
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0020061367249581963,
          "train_loss": 0.0017767900135368109,
          "validation_loss": 0.001756118144840002,
          "train_metric": 0.00017895179917104542,
          "validation_metric": 0.00017686979845166206,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.284276251000847
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.0017039553422364406,
          "train_loss": 0.0014474795898422599,
          "validation_loss": 0.0011961733689531684,
          "train_metric": 0.00014578485570382327,
          "validation_metric": 0.00012047417840221897,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3060070360006648
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0017270449061470572,
          "train_loss": 0.0019031697884202003,
          "validation_loss": 0.0018503154860809445,
          "train_metric": 0.00019168028666172177,
          "validation_metric": 0.00018635702144820243,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.326509680999152
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.002180475123168435,
          "train_loss": 0.002288784831762314,
          "validation_loss": 0.0016458157915621996,
          "train_metric": 0.0002305180241819471,
          "validation_metric": 0.00016576053167227656,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3469172170007369
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0015423729964822996,
          "train_loss": 0.0018165878718718886,
          "validation_loss": 0.001704301335848868,
          "train_metric": 0.00018296006601303816,
          "validation_metric": 0.00017165098688565195,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.366962562999106
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.0032234261998382863,
          "train_loss": 0.0023911548778414726,
          "validation_loss": 0.001842659548856318,
          "train_metric": 0.00024082837626338005,
          "validation_metric": 0.0001855859300121665,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.387246858999788
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.006601787816180149,
          "train_loss": 0.006433246657252312,
          "validation_loss": 0.006822941824793816,
          "train_metric": 0.0006479330477304757,
          "validation_metric": 0.000687181658577174,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4070261390006635
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0035965959759778343,
          "train_loss": 0.0028520552441477776,
          "validation_loss": 0.0024737175554037094,
          "train_metric": 0.0002872485783882439,
          "validation_metric": 0.0002491437771823257,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4266346450003766
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0028004393861920107,
          "train_loss": 0.004272925201803446,
          "validation_loss": 0.004459010437130928,
          "train_metric": 0.00043035339331254363,
          "validation_metric": 0.0004490951541811228,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4468529309997393
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0028591335758392233,
          "train_loss": 0.0016040296759456396,
          "validation_loss": 0.0011391775915399194,
          "train_metric": 0.00016155197226908058,
          "validation_metric": 0.00011473381164250895,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4679866839996976
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0040647276655363385,
          "train_loss": 0.003866623854264617,
          "validation_loss": 0.004257454536855221,
          "train_metric": 0.00038943218532949686,
          "validation_metric": 0.00042879529064521194,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4878029100000276
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0020060224014741834,
          "train_loss": 0.002261676359921694,
          "validation_loss": 0.0018569206586107612,
          "train_metric": 0.00022778772108722478,
          "validation_metric": 0.000187022247700952,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5094121580004867
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0018406348826829344,
          "train_loss": 0.0015012867515906692,
          "validation_loss": 0.0011148540070280433,
          "train_metric": 0.00015120411990210414,
          "validation_metric": 0.00011228400398977101,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5467224420008279
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.002162911870982498,
          "train_loss": 0.011232945136725903,
          "validation_loss": 0.00887205172330141,
          "train_metric": 0.0011313408613204956,
          "validation_metric": 0.0008935605874285102,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5741837700006727
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0084548872691812,
          "train_loss": 0.005071002524346113,
          "validation_loss": 0.004630617331713438,
          "train_metric": 0.0005107328179292381,
          "validation_metric": 0.0004663788713514805,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6043746859995736
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.005415471823653206,
          "train_loss": 0.003343538148328662,
          "validation_loss": 0.0032748656813055277,
          "train_metric": 0.0003367489262018353,
          "validation_metric": 0.0003298324591014534,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6318457569996099
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.003593408804590581,
          "train_loss": 0.004045362584292889,
          "validation_loss": 0.004012070596218109,
          "train_metric": 0.0004074340686202049,
          "validation_metric": 0.000404081103624776,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6568846259997372
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0027095369332528207,
          "train_loss": 0.0019463020144030452,
          "validation_loss": 0.0013752930099144578,
          "train_metric": 0.00019602442625910044,
          "validation_metric": 0.0001385144714731723,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6884235090001312
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.004001264758699108,
          "train_loss": 0.003718052990734577,
          "validation_loss": 0.0027752479072660208,
          "train_metric": 0.0003744686837308109,
          "validation_metric": 0.0002795127802528441,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7154928460004157
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.005240593753114808,
          "train_loss": 0.008636370301246643,
          "validation_loss": 0.009356452152132988,
          "train_metric": 0.00086982361972332,
          "validation_metric": 0.0009423475712537766,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7491432330007228
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006110473608714528,
          "train_loss": 0.006646636873483658,
          "validation_loss": 0.008010126650333405,
          "train_metric": 0.0006694248877465725,
          "validation_metric": 0.0008067506132647395,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7731447810001555
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.003902298129105475,
          "train_loss": 0.0018766896100714803,
          "validation_loss": 0.0012020838912576437,
          "train_metric": 0.00018901328439824283,
          "validation_metric": 0.00012106947542633861,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8029629249995196
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0020461318617890356,
          "train_loss": 0.0015792911872267723,
          "validation_loss": 0.0011573329102247953,
          "train_metric": 0.00015906042244751006,
          "validation_metric": 0.000116562339826487,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8340500270005577
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.002321494046555017,
          "train_loss": 0.006457476411014795,
          "validation_loss": 0.0064075966365635395,
          "train_metric": 0.0006503733457066119,
          "validation_metric": 0.0006453496171161532,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8708087609993527
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.003204850427209749,
          "train_loss": 0.002415082883089781,
          "validation_loss": 0.0018170818220824003,
          "train_metric": 0.00024323830439243466,
          "validation_metric": 0.00018300981901120394,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.89477811299912
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0031883777355687926,
          "train_loss": 0.001261584460735321,
          "validation_loss": 0.0008865423733368516,
          "train_metric": 0.00012706215784419328,
          "validation_metric": 8.928930765250698e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9261557800000446
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.002409536733466666,
          "train_loss": 0.001181971631012857,
          "validation_loss": 0.0009367477032355964,
          "train_metric": 0.00011904386337846518,
          "validation_metric": 9.434579988010228e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.953234741000415
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.001390066232488607,
          "train_loss": 0.0012464605970308185,
          "validation_loss": 0.0008673122501932085,
          "train_metric": 0.00012553896522149444,
          "validation_metric": 8.735251321922988e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9705352040000434
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.002419913191261003,
          "train_loss": 0.001348488382063806,
          "validation_loss": 0.001102409209124744,
          "train_metric": 0.0001358148001600057,
          "validation_metric": 0.00011103061842732131,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9891571629996179
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0017408568201062735,
          "train_loss": 0.0011814890895038843,
          "validation_loss": 0.0009288333239965141,
          "train_metric": 0.00011899525998160243,
          "validation_metric": 9.354869689559564e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0069719279999845
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0017324551154160872,
          "train_loss": 0.0011425481643527746,
          "validation_loss": 0.0009018828277476132,
          "train_metric": 0.00011507324961712584,
          "validation_metric": 9.083432814804837e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0218934130007256
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0012147173438279424,
          "train_loss": 0.0013103218516334891,
          "validation_loss": 0.0009822866413742304,
          "train_metric": 0.00013197080988902599,
          "validation_metric": 9.893231617752463e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0363948270005494
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0011784885446104454,
          "train_loss": 0.000960516685154289,
          "validation_loss": 0.0007663265569135547,
          "train_metric": 9.673971362644807e-05,
          "validation_metric": 7.718160486547276e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.052872180000122
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0010833909072971437,
          "train_loss": 0.0010734724346548319,
          "validation_loss": 0.000848552561365068,
          "train_metric": 0.0001081162117770873,
          "validation_metric": 8.546310709789395e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.067702543999985
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0012365112434054026,
          "train_loss": 0.0009756602230481803,
          "validation_loss": 0.0008622216410003603,
          "train_metric": 9.826493624132127e-05,
          "validation_metric": 8.683979831403121e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0824414390008315
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0025003928276419174,
          "train_loss": 0.0013205186696723104,
          "validation_loss": 0.001320013077929616,
          "train_metric": 0.00013299781130626798,
          "validation_metric": 0.0001329468796029687,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0971738940006617
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0016845814006956061,
          "train_loss": 0.0014334963634610176,
          "validation_loss": 0.0012157424353063107,
          "train_metric": 0.0001443765067961067,
          "validation_metric": 0.00012244514073245227,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1114199999992707
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0014987075119279325,
          "train_loss": 0.0009721279493533075,
          "validation_loss": 0.0006941115716472268,
          "train_metric": 9.790917101781815e-05,
          "validation_metric": 6.990836845943704e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.127833903999999
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0016023708831198746,
          "train_loss": 0.0011439751833677292,
          "validation_loss": 0.0009673596941865981,
          "train_metric": 0.00011521699343575165,
          "validation_metric": 9.742892871145159e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.146759798999483
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.001111484376451699,
          "train_loss": 0.0007552597671747208,
          "validation_loss": 0.0006622183136641979,
          "train_metric": 7.6067000918556e-05,
          "validation_metric": 6.669620051980019e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.162289096000677
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0009331957116955891,
          "train_loss": 0.0008972304640337825,
          "validation_loss": 0.0007863245555199683,
          "train_metric": 9.036575647769496e-05,
          "validation_metric": 7.919573545223102e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.176668026000698
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0009515590008959407,
          "train_loss": 0.0007746305782347918,
          "validation_loss": 0.0006479760631918907,
          "train_metric": 7.80179543653503e-05,
          "validation_metric": 6.526177457999438e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.191435372000342
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0010599801080388715,
          "train_loss": 0.0018618018366396427,
          "validation_loss": 0.0014919815585017204,
          "train_metric": 0.00018751384050119668,
          "validation_metric": 0.0001502669183537364,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2230334609994316
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0008684580097906291,
          "train_loss": 0.0007200469262897968,
          "validation_loss": 0.0005225190543569624,
          "train_metric": 7.252048817463219e-05,
          "validation_metric": 5.262620106805116e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2442516749997594
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0007888901418482419,
          "train_loss": 0.0010073018493130803,
          "validation_loss": 0.0008439203957095742,
          "train_metric": 0.00010145175474463031,
          "validation_metric": 8.499657269567251e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2586936089992378
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.000710979684299673,
          "train_loss": 0.0007370702805928886,
          "validation_loss": 0.0006512306863442063,
          "train_metric": 7.423501665471122e-05,
          "validation_metric": 6.558956374647096e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.273305139000513
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0008429464542132337,
          "train_loss": 0.0007028619293123484,
          "validation_loss": 0.0004955179756507277,
          "train_metric": 7.078967610141262e-05,
          "validation_metric": 4.990675370208919e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.296055377000812
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0010533402873988962,
          "train_loss": 0.0013825575588271022,
          "validation_loss": 0.0009572035050950944,
          "train_metric": 0.00013924615632276982,
          "validation_metric": 9.6406030934304e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.316392590999385
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.001499497830081964,
          "train_loss": 0.0007349012885242701,
          "validation_loss": 0.0006142314523458481,
          "train_metric": 7.401655602734536e-05,
          "validation_metric": 6.186313839862123e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3380269659992337
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0008827261553960852,
          "train_loss": 0.0007867153035476804,
          "validation_loss": 0.0006472402019426227,
          "train_metric": 7.923508383100852e-05,
          "validation_metric": 6.51876543997787e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3591530039993813
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0007599805849167751,
          "train_loss": 0.0008101004641503096,
          "validation_loss": 0.0006061627063900232,
          "train_metric": 8.159034769050777e-05,
          "validation_metric": 6.105047941673547e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.376309706000029
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0023320287582464516,
          "train_loss": 0.002897440455853939,
          "validation_loss": 0.002189994091168046,
          "train_metric": 0.00029181959689594805,
          "validation_metric": 0.0002205682103522122,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3916382209990843
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.001986946204851847,
          "train_loss": 0.004696398973464966,
          "validation_loss": 0.0042628576047718525,
          "train_metric": 0.00047300412552431226,
          "validation_metric": 0.00042933941585943103,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.40632012199967
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.002292122386279516,
          "train_loss": 0.00269507453776896,
          "validation_loss": 0.001467134919948876,
          "train_metric": 0.0002714380680117756,
          "validation_metric": 0.00014776448369957507,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.421204674999899
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.001585295156473876,
          "train_loss": 0.002505150157958269,
          "validation_loss": 0.002000041538849473,
          "train_metric": 0.0002523095754440874,
          "validation_metric": 0.00020143689471296966,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.43612972200026
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0019906311317754444,
          "train_loss": 0.0011033102637156844,
          "validation_loss": 0.0008188696810975671,
          "train_metric": 0.00011112137144664302,
          "validation_metric": 8.247356163337827e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4553184529995633
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0008147057196765672,
          "train_loss": 0.0007392544648610055,
          "validation_loss": 0.0005941683775745332,
          "train_metric": 7.445500523317605e-05,
          "validation_metric": 5.984247036394663e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4712879540002177
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0010260439630656037,
          "train_loss": 0.002080790465697646,
          "validation_loss": 0.0016173674957826734,
          "train_metric": 0.00020956959633622319,
          "validation_metric": 0.00016289533232338727,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.486828410999806
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0013931800640420988,
          "train_loss": 0.0025870027020573616,
          "validation_loss": 0.00205914955586195,
          "train_metric": 0.0002605534391477704,
          "validation_metric": 0.0002073900104733184,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.5015855569999985
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0011363845806044992,
          "train_loss": 0.0006742316763848066,
          "validation_loss": 0.0005440751556307077,
          "train_metric": 6.79061486152932e-05,
          "validation_metric": 5.479726314661093e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.5167037899991556
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.000714739719114732,
          "train_loss": 0.0006696152267977595,
          "validation_loss": 0.000547104689758271,
          "train_metric": 6.744118581991643e-05,
          "validation_metric": 5.510238770511933e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.5316200130000652
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.000726640264474554,
          "train_loss": 0.0005994715029373765,
          "validation_loss": 0.000443046708824113,
          "train_metric": 6.037657294655219e-05,
          "validation_metric": 4.4622043787967414e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.5473786120001023
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.000699035513207491,
          "train_loss": 0.0006095705321058631,
          "validation_loss": 0.0004978777142241597,
          "train_metric": 6.139372271718457e-05,
          "validation_metric": 5.014442285755649e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.5633345459991688
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0006894093639857601,
          "train_loss": 0.0005615675472654402,
          "validation_loss": 0.0004286050098016858,
          "train_metric": 5.6559027143521234e-05,
          "validation_metric": 4.316752892918885e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.578139507999367
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0009189539723593043,
          "train_loss": 0.0007903330842964351,
          "validation_loss": 0.0006748268497176468,
          "train_metric": 7.959945651236922e-05,
          "validation_metric": 6.796608795411885e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.59510666999995
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0015095192238732125,
          "train_loss": 0.0013989565195515752,
          "validation_loss": 0.0014995195670053363,
          "train_metric": 0.00014089775504544377,
          "validation_metric": 0.00015102613542694598,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.6102968779996445
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0016194925956369843,
          "train_loss": 0.0017759758047759533,
          "validation_loss": 0.0012059382861480117,
          "train_metric": 0.00017886979912873358,
          "validation_metric": 0.00012145769142080098,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.62464116099909
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0013028637367824558,
          "train_loss": 0.0005991756916046143,
          "validation_loss": 0.0004957999917678535,
          "train_metric": 6.034678153810091e-05,
          "validation_metric": 4.993515904061496e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.6391110080003273
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0008587142992837471,
          "train_loss": 0.0006913767429068685,
          "validation_loss": 0.0007160301320254803,
          "train_metric": 6.96329225320369e-05,
          "validation_metric": 7.211593037936836e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.6535512610007572
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0008113021776807727,
          "train_loss": 0.0007229626644402742,
          "validation_loss": 0.0005817274213768542,
          "train_metric": 7.281415309989825e-05,
          "validation_metric": 5.8589463151292875e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.667654744000174
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0005925541572651127,
          "train_loss": 0.0005365508841350675,
          "validation_loss": 0.00039055911474861205,
          "train_metric": 5.403943941928446e-05,
          "validation_metric": 3.9335685869446024e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.6827004400001897
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0006207470269146143,
          "train_loss": 0.0008947577443905175,
          "validation_loss": 0.0007931837462820113,
          "train_metric": 9.011672955239192e-05,
          "validation_metric": 7.988657307578251e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.6980402120007057
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0009856000460786163,
          "train_loss": 0.0006760033429600298,
          "validation_loss": 0.0006067940266802907,
          "train_metric": 6.808457692386582e-05,
          "validation_metric": 6.111407856224105e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.714138780000212
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.001077648053978919,
          "train_loss": 0.0011006571585312486,
          "validation_loss": 0.0006227483972907066,
          "train_metric": 0.00011085417645517737,
          "validation_metric": 6.272093742154539e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.728071948999059
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0010348358591727447,
          "train_loss": 0.001040260773152113,
          "validation_loss": 0.000762815005145967,
          "train_metric": 0.00010477126488694921,
          "validation_metric": 7.682792056584731e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.749890356999458
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0006579512391908793,
          "train_loss": 0.0005078170797787607,
          "validation_loss": 0.00042929136543534696,
          "train_metric": 5.1145478209946305e-05,
          "validation_metric": 4.3236657802481204e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7648220720002428
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.000606933069320803,
          "train_loss": 0.0006670717848464847,
          "validation_loss": 0.0004781965399160981,
          "train_metric": 6.71850357321091e-05,
          "validation_metric": 4.8162204620894045e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7822891420000815
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0005396958949859254,
          "train_loss": 0.0004736183909699321,
          "validation_loss": 0.00040575317689217627,
          "train_metric": 4.770111263496801e-05,
          "validation_metric": 4.0865976188797504e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7971723430000566
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0004927599802613258,
          "train_loss": 0.00046927310177125037,
          "validation_loss": 0.0003673181345220655,
          "train_metric": 4.7263471060432494e-05,
          "validation_metric": 3.699494118336588e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.81445780499962
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0005609894415101735,
          "train_loss": 0.0005332434084266424,
          "validation_loss": 0.00046150796697475016,
          "train_metric": 5.3706320613855496e-05,
          "validation_metric": 4.648139656637795e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8296175209998182
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0009738629332787241,
          "train_loss": 0.0008304921793751419,
          "validation_loss": 0.0005294853472150862,
          "train_metric": 8.364413952222094e-05,
          "validation_metric": 5.332782893674448e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.843820265999966
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0008640666528663132,
          "train_loss": 0.000529735698364675,
          "validation_loss": 0.0004919539205729961,
          "train_metric": 5.335303285392001e-05,
          "validation_metric": 4.954779433319345e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8577515529996163
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0007753005484119058,
          "train_loss": 0.0006802050047554076,
          "validation_loss": 0.0005046048318035901,
          "train_metric": 6.850775389466435e-05,
          "validation_metric": 5.082194911665283e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8720215230005124
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0009336260773125105,
          "train_loss": 0.00047524814726784825,
          "validation_loss": 0.0003975180152337998,
          "train_metric": 4.7865250962786376e-05,
          "validation_metric": 4.0036553400568664e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8865914569996676
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0005038759554736316,
          "train_loss": 0.00046782390563748777,
          "validation_loss": 0.00040473323315382004,
          "train_metric": 4.711750807473436e-05,
          "validation_metric": 4.0763250581221655e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9076978739994956
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.000945818953368871,
          "train_loss": 0.0008276710286736488,
          "validation_loss": 0.0005718684988096356,
          "train_metric": 8.335999154951423e-05,
          "validation_metric": 5.759651321568526e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9227222520003124
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0010849764021259034,
          "train_loss": 0.00048262818017974496,
          "validation_loss": 0.00037242466351017356,
          "train_metric": 4.860854824073613e-05,
          "validation_metric": 3.750924952328205e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9441873579999083
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0005612868144453387,
          "train_loss": 0.0004631209303624928,
          "validation_loss": 0.0003701538371387869,
          "train_metric": 4.664385051000863e-05,
          "validation_metric": 3.7280533433659e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.960169329000564
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0011360869084455771,
          "train_loss": 0.0013377949362620711,
          "validation_loss": 0.00128399848472327,
          "train_metric": 0.00013473779836203903,
          "validation_metric": 0.00012931961100548506,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9830055099992023
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0015311713959818007,
          "train_loss": 0.0012828262988477945,
          "validation_loss": 0.0009347639279440045,
          "train_metric": 0.00012920155131723732,
          "validation_metric": 9.41460020840168e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9984544519993506
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0007702884449827252,
          "train_loss": 0.0005837542121298611,
          "validation_loss": 0.0004150537424720824,
          "train_metric": 5.8793586504179984e-05,
          "validation_metric": 4.180269024800509e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.0135846890007087
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0005370584503907594,
          "train_loss": 0.0004953020252287388,
          "validation_loss": 0.0003903490141965449,
          "train_metric": 4.9885005864780396e-05,
          "validation_metric": 3.931452374672517e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.027941305999775
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0007483099707314977,
          "train_loss": 0.0010151524329558015,
          "validation_loss": 0.0006455963593907654,
          "train_metric": 0.000102242425782606,
          "validation_metric": 6.502209726022556e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.0422121249994234
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0006709594654239481,
          "train_loss": 0.0004956838674843311,
          "validation_loss": 0.0003564115904737264,
          "train_metric": 4.992346657672897e-05,
          "validation_metric": 3.5896471672458574e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.056406835999951
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0009004975090647349,
          "train_loss": 0.0004513422609306872,
          "validation_loss": 0.0003502455074340105,
          "train_metric": 4.545753836282529e-05,
          "validation_metric": 3.5275446862215176e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.0714464169996063
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0006481930049631046,
          "train_loss": 0.0008068275637924671,
          "validation_loss": 0.00048712673014961183,
          "train_metric": 8.126072498271242e-05,
          "validation_metric": 4.9061614845413715e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.085365269000249
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0005660286406055093,
          "train_loss": 0.0005851404275745153,
          "validation_loss": 0.0004207146994303912,
          "train_metric": 5.893320121685974e-05,
          "validation_metric": 4.237284520058893e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.0985904400004074
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0006385936303558992,
          "train_loss": 0.0006696440977975726,
          "validation_loss": 0.0005013167392462492,
          "train_metric": 6.744410347891971e-05,
          "validation_metric": 5.049079118180089e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.112231017999875
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0005787678583146771,
          "train_loss": 0.00042351576848886907,
          "validation_loss": 0.0003361297131050378,
          "train_metric": 4.265495590516366e-05,
          "validation_metric": 3.385375384823419e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.12642773799962
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0005533422490771045,
          "train_loss": 0.0006642754888162017,
          "validation_loss": 0.0004976690979674459,
          "train_metric": 6.690339796477929e-05,
          "validation_metric": 5.012340625398792e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.140261267000824
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.00048564630378677975,
          "train_loss": 0.0004120109078940004,
          "validation_loss": 0.0003079709713347256,
          "train_metric": 4.149623055127449e-05,
          "validation_metric": 3.101770926150493e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.1550610569993296
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.00043557440039876383,
          "train_loss": 0.0003972981357946992,
          "validation_loss": 0.0003044064505957067,
          "train_metric": 4.0014412661548704e-05,
          "validation_metric": 3.065870623686351e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.169532079999044
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.00042027176641568076,
          "train_loss": 0.0003937441506423056,
          "validation_loss": 0.0003121639310847968,
          "train_metric": 3.9656464650761336e-05,
          "validation_metric": 3.144000220345333e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.1910827230003633
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0004471573001865181,
          "train_loss": 0.0005128788761794567,
          "validation_loss": 0.00045719838817603886,
          "train_metric": 5.165527545614168e-05,
          "validation_metric": 4.6047356590861455e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.225730382999245
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0005911839343752945,
          "train_loss": 0.0008396205957978964,
          "validation_loss": 0.000600555504206568,
          "train_metric": 8.456351497443393e-05,
          "validation_metric": 6.0485755966510624e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.2554628210000374
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0007345873282247339,
          "train_loss": 0.000823371927253902,
          "validation_loss": 0.0006545680807903409,
          "train_metric": 8.29270138638094e-05,
          "validation_metric": 6.592571298824623e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.2810105589996965
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0006214027580426773,
          "train_loss": 0.00041622709250077605,
          "validation_loss": 0.0003520985774230212,
          "train_metric": 4.1920869989553466e-05,
          "validation_metric": 3.546207517501898e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3187622250006825
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0005430116298157373,
          "train_loss": 0.00045219185994938016,
          "validation_loss": 0.0003464340406935662,
          "train_metric": 4.5543107262346894e-05,
          "validation_metric": 3.489156370051205e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3388218250001955
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0004981085967301624,
          "train_loss": 0.0007995185442268848,
          "validation_loss": 0.0006866967887617648,
          "train_metric": 8.05245726951398e-05,
          "validation_metric": 6.91615859977901e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3588171520004835
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0005443296240628115,
          "train_loss": 0.0003774147480726242,
          "validation_loss": 0.0003044845652766526,
          "train_metric": 3.801183629548177e-05,
          "validation_metric": 3.066657154704444e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3791531549995852
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0004987593783880584,
          "train_loss": 0.0005878182710148394,
          "validation_loss": 0.0003850880602840334,
          "train_metric": 5.920289913774468e-05,
          "validation_metric": 3.8784663047408685e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.3998245040002075
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0004682969520217739,
          "train_loss": 0.0003782364074140787,
          "validation_loss": 0.0003154903242830187,
          "train_metric": 3.809458576142788e-05,
          "validation_metric": 3.177503094775602e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4198145939990354
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.00042609466800058726,
          "train_loss": 0.0004531600570771843,
          "validation_loss": 0.0003786420274991542,
          "train_metric": 4.564061600831337e-05,
          "validation_metric": 3.813544026343152e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4396688519991585
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.00043618015206448035,
          "train_loss": 0.0004061362415086478,
          "validation_loss": 0.00034454429987818,
          "train_metric": 4.090455695404671e-05,
          "validation_metric": 3.4701246477197856e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.4600559980008256
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.00038705534552718746,
          "train_loss": 0.0003860268625430763,
          "validation_loss": 0.0003176046593580395,
          "train_metric": 3.8879214116605e-05,
          "validation_metric": 3.198797639925033e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.480282726999576
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.00038788578149251407,
          "train_loss": 0.00036845181602984667,
          "validation_loss": 0.00029599666595458984,
          "train_metric": 3.710911187226884e-05,
          "validation_metric": 2.98117029160494e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.5033081300007325
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0003742994495041785,
          "train_loss": 0.0003607433463912457,
          "validation_loss": 0.00030182398040778935,
          "train_metric": 3.6332749004941434e-05,
          "validation_metric": 3.0398603485082276e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.5250956879990554
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0003831854255622602,
          "train_loss": 0.0004304691683501005,
          "validation_loss": 0.00034296681405976415,
          "train_metric": 4.3355281377444044e-05,
          "validation_metric": 3.4542357752798125e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.5468182120002893
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.00040377465575147653,
          "train_loss": 0.0003839514101855457,
          "validation_loss": 0.00032372953137382865,
          "train_metric": 3.867018313030712e-05,
          "validation_metric": 3.260485391365364e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.5678970930002833
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0003805406122410204,
          "train_loss": 0.00041513118776492774,
          "validation_loss": 0.0003306072030682117,
          "train_metric": 4.18104900745675e-05,
          "validation_metric": 3.329754690639675e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.6900733429993124
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.00035288317758386256,
          "train_loss": 0.00038229173514992,
          "validation_loss": 0.00031757738906890154,
          "train_metric": 3.8503021642100066e-05,
          "validation_metric": 3.198523336322978e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.7119363309993787
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0003871398585033603,
          "train_loss": 0.00039162556640803814,
          "validation_loss": 0.00033871171763166785,
          "train_metric": 3.944308991776779e-05,
          "validation_metric": 3.411380748730153e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.7317735010001343
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0003805838987318566,
          "train_loss": 0.00035200314596295357,
          "validation_loss": 0.00029398815240710974,
          "train_metric": 3.545246727298945e-05,
          "validation_metric": 2.9609409466502257e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.752402064999842
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.00036845328668277944,
          "train_loss": 0.0003470314550213516,
          "validation_loss": 0.0002906238369178027,
          "train_metric": 3.495173586998135e-05,
          "validation_metric": 2.9270569939399138e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.7732150079991698
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0003569741484170663,
          "train_loss": 0.00034633715404197574,
          "validation_loss": 0.0002855377388186753,
          "train_metric": 3.488181027933024e-05,
          "validation_metric": 2.8758317057508975e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.7938738459997694
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.0003653632238638238,
          "train_loss": 0.0003448373463470489,
          "validation_loss": 0.0002897440572269261,
          "train_metric": 3.4730754123302177e-05,
          "validation_metric": 2.9181963327573612e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.813976105999245
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0003668381232273532,
          "train_loss": 0.00038811040576547384,
          "validation_loss": 0.00030609112582169473,
          "train_metric": 3.908905273419805e-05,
          "validation_metric": 3.082837429246865e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.833729182000752
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0003705934750541928,
          "train_loss": 0.00034400628646835685,
          "validation_loss": 0.0002883311826735735,
          "train_metric": 3.464705514488742e-05,
          "validation_metric": 2.903966378653422e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8534624629992322
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.000358027868969657,
          "train_loss": 0.0003427918709348887,
          "validation_loss": 0.00028842175379395485,
          "train_metric": 3.4524746297392994e-05,
          "validation_metric": 2.9048780561424792e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8734644659998594
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.00035015714456676506,
          "train_loss": 0.00034095620503649116,
          "validation_loss": 0.0002925486187450588,
          "train_metric": 3.4339860576437786e-05,
          "validation_metric": 2.9464428735082038e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8927528679996612
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0003717328590937541,
          "train_loss": 0.0003448535571806133,
          "validation_loss": 0.00028974839369766414,
          "train_metric": 3.473238757578656e-05,
          "validation_metric": 2.9182399885030463e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.912935906000712
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.00037997526305844076,
          "train_loss": 0.00035468279384076595,
          "validation_loss": 0.0002983494778163731,
          "train_metric": 3.572235073079355e-05,
          "validation_metric": 3.0048668122617528e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9326605280002696
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.00036374804403749295,
          "train_loss": 0.0006160202901810408,
          "validation_loss": 0.0004455654416233301,
          "train_metric": 6.204330566106364e-05,
          "validation_metric": 4.487572005018592e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9519818400003714
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0004721450022771023,
          "train_loss": 0.00035751413088291883,
          "validation_loss": 0.0003169221745338291,
          "train_metric": 3.6007513699587435e-05,
          "validation_metric": 3.191924042766914e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.971552736000376
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.00035820324046653695,
          "train_loss": 0.0004252365615684539,
          "validation_loss": 0.00039024968282319605,
          "train_metric": 4.282827285351232e-05,
          "validation_metric": 3.930451930500567e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9911581050000677
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.00037599009101541014,
          "train_loss": 0.00033787806751206517,
          "validation_loss": 0.0002902658306993544,
          "train_metric": 3.402984293643385e-05,
          "validation_metric": 2.9234513931442052e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.010964016999424
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.00035646352625917643,
          "train_loss": 0.00034618392237462103,
          "validation_loss": 0.00028491095872595906,
          "train_metric": 3.486637433525175e-05,
          "validation_metric": 2.8695183573290706e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.032345773999623
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0003508990780574095,
          "train_loss": 0.00033606382203288376,
          "validation_loss": 0.00029277519206516445,
          "train_metric": 3.384711453691125e-05,
          "validation_metric": 2.948724795714952e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.052618775000155
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.000341857502462517,
          "train_loss": 0.00033858270035125315,
          "validation_loss": 0.0002810624719131738,
          "train_metric": 3.41008126270026e-05,
          "validation_metric": 2.8307582397246733e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.073669049999808
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.00035830794877256267,
          "train_loss": 0.00032971007749438286,
          "validation_loss": 0.00028371921507641673,
          "train_metric": 3.320719406474382e-05,
          "validation_metric": 2.857516119547654e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.09388322400082
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0003608262413763441,
          "train_loss": 0.00033694604644551873,
          "validation_loss": 0.00028370440122671425,
          "train_metric": 3.3935972169274464e-05,
          "validation_metric": 2.8573669624165632e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.1144967780001025
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0003329938381284592,
          "train_loss": 0.00033036223612725735,
          "validation_loss": 0.0002812645398080349,
          "train_metric": 3.3272866858169436e-05,
          "validation_metric": 2.83279314317042e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.134653842000262
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.00033863157841551583,
          "train_loss": 0.0003271506866440177,
          "validation_loss": 0.00027439850964583457,
          "train_metric": 3.294941416243091e-05,
          "validation_metric": 2.7636411687126383e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.155986841999038
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0003613564049373963,
          "train_loss": 0.0003503981570247561,
          "validation_loss": 0.0003054361732210964,
          "train_metric": 3.529081732267514e-05,
          "validation_metric": 3.076241409871727e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.176004107999688
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0003534037914505461,
          "train_loss": 0.0003285643470007926,
          "validation_loss": 0.00028213229961693287,
          "train_metric": 3.309179737698287e-05,
          "validation_metric": 2.8415330234565772e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.196337719999065
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.00034584171135065844,
          "train_loss": 0.00033882088609971106,
          "validation_loss": 0.00027868844335898757,
          "train_metric": 3.412479782127775e-05,
          "validation_metric": 2.8068479878129438e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.216241382999215
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0003491957268124679,
          "train_loss": 0.00034122608485631645,
          "validation_loss": 0.00028471730183809996,
          "train_metric": 3.436703991610557e-05,
          "validation_metric": 2.86756876448635e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.236549286999434
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0003388841496416717,
          "train_loss": 0.00032751221442595124,
          "validation_loss": 0.00028702145209535956,
          "train_metric": 3.298583396826871e-05,
          "validation_metric": 2.8907745218020864e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.256665159999102
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.00036009566429129336,
          "train_loss": 0.0003371155180502683,
          "validation_loss": 0.0002805476833600551,
          "train_metric": 3.395304156583734e-05,
          "validation_metric": 2.8255733923288062e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.2764622989998315
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.00033929487381101353,
          "train_loss": 0.00032389455009251833,
          "validation_loss": 0.00027900704299099743,
          "train_metric": 3.262147220084444e-05,
          "validation_metric": 2.8100565032218583e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.2966046999990795
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.0003414490684008342,
          "train_loss": 0.00034779153065755963,
          "validation_loss": 0.0003101510810665786,
          "train_metric": 3.502828985801898e-05,
          "validation_metric": 3.123728674836457e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.316733629000737
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.00037752490698039765,
          "train_loss": 0.0003540367470122874,
          "validation_loss": 0.0002878240484278649,
          "train_metric": 3.5657285479828715e-05,
          "validation_metric": 2.8988586564082652e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.336726441000792
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.00032694958008505637,
          "train_loss": 0.00032493629259988666,
          "validation_loss": 0.00028570802533067763,
          "train_metric": 3.272639514761977e-05,
          "validation_metric": 2.8775466489605606e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.3561149739998655
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.00034163308646384394,
          "train_loss": 0.00032465136609971523,
          "validation_loss": 0.00026979678659699857,
          "train_metric": 3.269769513281062e-05,
          "validation_metric": 2.7172945920028724e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.376387782000165
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0003464586839072581,
          "train_loss": 0.0003325994766782969,
          "validation_loss": 0.0002850258897524327,
          "train_metric": 3.34982032654807e-05,
          "validation_metric": 2.8706766897812486e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.396360332000768
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.00034477428926038556,
          "train_loss": 0.00033490030909888446,
          "validation_loss": 0.0002992725057993084,
          "train_metric": 3.3729935239534825e-05,
          "validation_metric": 3.0141631214064546e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.416023623000001
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.00036395677580003394,
          "train_loss": 0.0003955359279643744,
          "validation_loss": 0.00030799192609265447,
          "train_metric": 3.983693022746593e-05,
          "validation_metric": 3.1019819289213046e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.435977585000728
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0003637172485468909,
          "train_loss": 0.00033946134499274194,
          "validation_loss": 0.00029862605151720345,
          "train_metric": 3.418929918552749e-05,
          "validation_metric": 3.0076520488364622e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.456352429999242
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.0003453317731327843,
          "train_loss": 0.0003265762934461236,
          "validation_loss": 0.0002872067561838776,
          "train_metric": 3.2891570299398154e-05,
          "validation_metric": 2.8926409868290648e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.476720266000484
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0003766615209315205,
          "train_loss": 0.0003248974389862269,
          "validation_loss": 0.0002694205613806844,
          "train_metric": 3.272247704444453e-05,
          "validation_metric": 2.713505455176346e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.503738396999324
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.0003424898759476491,
          "train_loss": 0.00032872549491003156,
          "validation_loss": 0.00027292314916849136,
          "train_metric": 3.310803003842011e-05,
          "validation_metric": 2.748782208072953e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.532141835999937
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.0003638399357441813,
          "train_loss": 0.00031779910204932094,
          "validation_loss": 0.000267805386101827,
          "train_metric": 3.20075596391689e-05,
          "validation_metric": 2.6972378691425547e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.559078860000227
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0003240369351260597,
          "train_loss": 0.00032727583311498165,
          "validation_loss": 0.00029358992469497025,
          "train_metric": 3.2962019758997485e-05,
          "validation_metric": 2.9569302569143474e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.579685566999615
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.0003389974217498093,
          "train_loss": 0.0003199097700417042,
          "validation_loss": 0.0002637380675878376,
          "train_metric": 3.22201412927825e-05,
          "validation_metric": 2.65627313638106e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.602570541999739
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0003186224976161611,
          "train_loss": 0.0003257069911342114,
          "validation_loss": 0.00028936899616383016,
          "train_metric": 3.280401142546907e-05,
          "validation_metric": 2.9144188374630176e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.623134277000645
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.0003394426112208748,
          "train_loss": 0.0003246062551625073,
          "validation_loss": 0.00028636588831432164,
          "train_metric": 3.2693151297280565e-05,
          "validation_metric": 2.884172681660857e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.642963813999813
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0003623662250902271,
          "train_loss": 0.0003415491955820471,
          "validation_loss": 0.00028102510259486735,
          "train_metric": 3.439958891249262e-05,
          "validation_metric": 2.8303822546149604e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.663664874999085
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.0003272360008850228,
          "train_loss": 0.00031415163539350033,
          "validation_loss": 0.00027221255004405975,
          "train_metric": 3.164020017720759e-05,
          "validation_metric": 2.7416248485678807e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.683603449000657
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.0003188829805367277,
          "train_loss": 0.00031163834501057863,
          "validation_loss": 0.00026662443997338414,
          "train_metric": 3.1387073249788955e-05,
          "validation_metric": 2.6853436793317087e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.705194878000839
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.0003325373509142082,
          "train_loss": 0.0003209039568901062,
          "validation_loss": 0.0002712107670959085,
          "train_metric": 3.232027302146889e-05,
          "validation_metric": 2.7315356419421732e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.724902137999379
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.00033897991215781076,
          "train_loss": 0.0003217901394236833,
          "validation_loss": 0.0002717422612477094,
          "train_metric": 3.2409523555543274e-05,
          "validation_metric": 2.736888563958928e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.744687779000742
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.00031959877560439054,
          "train_loss": 0.00031945749651640654,
          "validation_loss": 0.00027865232550539076,
          "train_metric": 3.2174590160138905e-05,
          "validation_metric": 2.8064840080332942e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.764916820999133
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0003142996829410549,
          "train_loss": 0.00031227697036229074,
          "validation_loss": 0.0002636255230754614,
          "train_metric": 3.145138907711953e-05,
          "validation_metric": 2.65513954218477e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.785957554999186
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.00031671516717324266,
          "train_loss": 0.00031198651413433254,
          "validation_loss": 0.00027528300415724516,
          "train_metric": 3.1422136089531705e-05,
          "validation_metric": 2.772549123619683e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.805352359000608
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0003694421702675754,
          "train_loss": 0.0004324360634200275,
          "validation_loss": 0.0003207287809345871,
          "train_metric": 4.35533765994478e-05,
          "validation_metric": 3.23026288242545e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.825501438999709
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.00036274780995881883,
          "train_loss": 0.0003121584886685014,
          "validation_loss": 0.00025871407706290483,
          "train_metric": 3.1439460144611076e-05,
          "validation_metric": 2.6056735805468634e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.846600172999388
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.00031668919382354943,
          "train_loss": 0.00030838060774840415,
          "validation_loss": 0.00027612410485744476,
          "train_metric": 3.105896394117735e-05,
          "validation_metric": 2.7810208848677576e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.866874568999265
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.00031027929799165577,
          "train_loss": 0.0003066586796194315,
          "validation_loss": 0.000258448941167444,
          "train_metric": 3.0885537853464484e-05,
          "validation_metric": 2.6030029403045774e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.888308068999322
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.0003137228522973601,
          "train_loss": 0.0003081993490923196,
          "validation_loss": 0.00027267757104709744,
          "train_metric": 3.1040712201502174e-05,
          "validation_metric": 2.7463083824841306e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.908729991000655
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.0003109598001174163,
          "train_loss": 0.0003258029173593968,
          "validation_loss": 0.00026580135454423726,
          "train_metric": 3.281368117313832e-05,
          "validation_metric": 2.677053998922929e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.92887255200003
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.0003349289954712731,
          "train_loss": 0.0003251967136748135,
          "validation_loss": 0.00028535755700431764,
          "train_metric": 3.275262133684009e-05,
          "validation_metric": 2.8740170819219202e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.948320170000443
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.0002926884708358557,
          "train_loss": 0.0003303444245830178,
          "validation_loss": 0.0002721581840887666,
          "train_metric": 3.3271080610575154e-05,
          "validation_metric": 2.741077696555294e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.96804908500053
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.000341374528943561,
          "train_loss": 0.00030995564884506166,
          "validation_loss": 0.000260973465628922,
          "train_metric": 3.121759800706059e-05,
          "validation_metric": 2.628429137985222e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.987477778999164
        }
      ],
      "best": {
        "epoch": 235,
        "updates": 3760
      },
      "best_validation_metric": 2.6030029403045774e-05,
      "config": {
        "label": "LayerNorm + silu + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.005101963289072724
    },
    "92630a9c55d2_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 1,
          "samples_seen": 1024,
          "online_loss": 0.9984928965568542,
          "train_loss": 0.9801727533340454,
          "validation_loss": 0.707870364189148,
          "train_metric": 0.09871941059827805,
          "validation_metric": 0.07129411399364471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.007117434999599936
        },
        {
          "epoch": 2,
          "updates": 2,
          "samples_seen": 2048,
          "online_loss": 0.9801727533340454,
          "train_loss": 0.9603869915008545,
          "validation_loss": 0.6948423981666565,
          "train_metric": 0.0967266634106636,
          "validation_metric": 0.06998198479413986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.012902547998237424
        },
        {
          "epoch": 3,
          "updates": 3,
          "samples_seen": 3072,
          "online_loss": 0.9603869915008545,
          "train_loss": 0.9384909868240356,
          "validation_loss": 0.6804297566413879,
          "train_metric": 0.09452136605978012,
          "validation_metric": 0.06853038817644119,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01915436499984935
        },
        {
          "epoch": 4,
          "updates": 4,
          "samples_seen": 4096,
          "online_loss": 0.9384909272193909,
          "train_loss": 0.9138616919517517,
          "validation_loss": 0.6641983389854431,
          "train_metric": 0.09204079955816269,
          "validation_metric": 0.06689562648534775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10391816799892695
        },
        {
          "epoch": 5,
          "updates": 5,
          "samples_seen": 5120,
          "online_loss": 0.9138616919517517,
          "train_loss": 0.8859156370162964,
          "validation_loss": 0.6457259654998779,
          "train_metric": 0.08922617882490158,
          "validation_metric": 0.0650351494550705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1102512649995333
        },
        {
          "epoch": 6,
          "updates": 6,
          "samples_seen": 6144,
          "online_loss": 0.8859155774116516,
          "train_loss": 0.8541439175605774,
          "validation_loss": 0.6246242523193359,
          "train_metric": 0.08602624386548996,
          "validation_metric": 0.06290987133979797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11618843199903495
        },
        {
          "epoch": 7,
          "updates": 7,
          "samples_seen": 7168,
          "online_loss": 0.8541439771652222,
          "train_loss": 0.818126916885376,
          "validation_loss": 0.6005696654319763,
          "train_metric": 0.08239874243736267,
          "validation_metric": 0.06048717349767685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12323149999974703
        },
        {
          "epoch": 8,
          "updates": 8,
          "samples_seen": 8192,
          "online_loss": 0.8181267976760864,
          "train_loss": 0.7775458693504333,
          "validation_loss": 0.573311448097229,
          "train_metric": 0.07831156998872757,
          "validation_metric": 0.05774183198809624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12921053399986704
        },
        {
          "epoch": 9,
          "updates": 9,
          "samples_seen": 9216,
          "online_loss": 0.7775458097457886,
          "train_loss": 0.7321982979774475,
          "validation_loss": 0.5426748991012573,
          "train_metric": 0.07374434173107147,
          "validation_metric": 0.05465622991323471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1356187969995517
        },
        {
          "epoch": 10,
          "updates": 10,
          "samples_seen": 10240,
          "online_loss": 0.7321983575820923,
          "train_loss": 0.6820284724235535,
          "validation_loss": 0.5085822343826294,
          "train_metric": 0.06869141012430191,
          "validation_metric": 0.0512225367128849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14204525299828674
        },
        {
          "epoch": 11,
          "updates": 11,
          "samples_seen": 11264,
          "online_loss": 0.6820285320281982,
          "train_loss": 0.6271827816963196,
          "validation_loss": 0.4710938334465027,
          "train_metric": 0.06316754966974258,
          "validation_metric": 0.0474468432366848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14786413099864149
        },
        {
          "epoch": 12,
          "updates": 12,
          "samples_seen": 12288,
          "online_loss": 0.6271827220916748,
          "train_loss": 0.5680776834487915,
          "validation_loss": 0.4304536283016205,
          "train_metric": 0.05721469968557358,
          "validation_metric": 0.04335371404886246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15443323499857797
        },
        {
          "epoch": 13,
          "updates": 13,
          "samples_seen": 13312,
          "online_loss": 0.5680776238441467,
          "train_loss": 0.5054815411567688,
          "validation_loss": 0.38714104890823364,
          "train_metric": 0.05091025307774544,
          "validation_metric": 0.038991428911685944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1608947979984805
        },
        {
          "epoch": 14,
          "updates": 14,
          "samples_seen": 14336,
          "online_loss": 0.5054816007614136,
          "train_loss": 0.4406189024448395,
          "validation_loss": 0.3419446349143982,
          "train_metric": 0.04437752440571785,
          "validation_metric": 0.034439411014318466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1678276749989891
        },
        {
          "epoch": 15,
          "updates": 15,
          "samples_seen": 15360,
          "online_loss": 0.44061893224716187,
          "train_loss": 0.37529289722442627,
          "validation_loss": 0.2960549294948578,
          "train_metric": 0.03779812902212143,
          "validation_metric": 0.029817568138241768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1748394639998878
        },
        {
          "epoch": 16,
          "updates": 16,
          "samples_seen": 16384,
          "online_loss": 0.37529292702674866,
          "train_loss": 0.31199777126312256,
          "validation_loss": 0.25114503502845764,
          "train_metric": 0.03142327442765236,
          "validation_metric": 0.025294410064816475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18236121499830915
        },
        {
          "epoch": 17,
          "updates": 17,
          "samples_seen": 17408,
          "online_loss": 0.31199777126312256,
          "train_loss": 0.25395697355270386,
          "validation_loss": 0.2093973308801651,
          "train_metric": 0.025577614083886147,
          "validation_metric": 0.02108973078429699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18902684099884937
        },
        {
          "epoch": 18,
          "updates": 18,
          "samples_seen": 18432,
          "online_loss": 0.25395697355270386,
          "train_loss": 0.20501048862934113,
          "validation_loss": 0.17344781756401062,
          "train_metric": 0.02064790576696396,
          "validation_metric": 0.017469028010964394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.194948635999026
        },
        {
          "epoch": 19,
          "updates": 19,
          "samples_seen": 19456,
          "online_loss": 0.20501047372817993,
          "train_loss": 0.1692221313714981,
          "validation_loss": 0.1461627185344696,
          "train_metric": 0.017043432220816612,
          "validation_metric": 0.014720974490046501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2026164159997279
        },
        {
          "epoch": 20,
          "updates": 20,
          "samples_seen": 20480,
          "online_loss": 0.1692221313714981,
          "train_loss": 0.1499544233083725,
          "validation_loss": 0.1300608068704605,
          "train_metric": 0.01510285958647728,
          "validation_metric": 0.013099249452352524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20996661399840377
        },
        {
          "epoch": 21,
          "updates": 21,
          "samples_seen": 21504,
          "online_loss": 0.1499544233083725,
          "train_loss": 0.14820151031017303,
          "validation_loss": 0.12625624239444733,
          "train_metric": 0.014926312491297722,
          "validation_metric": 0.012716066092252731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.216010408999864
        },
        {
          "epoch": 22,
          "updates": 22,
          "samples_seen": 22528,
          "online_loss": 0.14820151031017303,
          "train_loss": 0.16058266162872314,
          "validation_loss": 0.13311734795570374,
          "train_metric": 0.01617329753935337,
          "validation_metric": 0.013407093472778797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22195766899858427
        },
        {
          "epoch": 23,
          "updates": 23,
          "samples_seen": 23552,
          "online_loss": 0.16058264672756195,
          "train_loss": 0.17885594069957733,
          "validation_loss": 0.14566545188426971,
          "train_metric": 0.018013713881373405,
          "validation_metric": 0.01467089168727398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2272505299988552
        },
        {
          "epoch": 24,
          "updates": 24,
          "samples_seen": 24576,
          "online_loss": 0.17885592579841614,
          "train_loss": 0.1931406855583191,
          "validation_loss": 0.15717023611068726,
          "train_metric": 0.019452422857284546,
          "validation_metric": 0.01582960970699787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2324099639990891
        },
        {
          "epoch": 25,
          "updates": 25,
          "samples_seen": 25600,
          "online_loss": 0.1931406855583191,
          "train_loss": 0.19712764024734497,
          "validation_loss": 0.16245920956134796,
          "train_metric": 0.019853973761200905,
          "validation_metric": 0.016362296417355537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2379667179993703
        },
        {
          "epoch": 26,
          "updates": 26,
          "samples_seen": 26624,
          "online_loss": 0.19712765514850616,
          "train_loss": 0.19032138586044312,
          "validation_loss": 0.1599966287612915,
          "train_metric": 0.019168473780155182,
          "validation_metric": 0.01611427403986454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24367068599894992
        },
        {
          "epoch": 27,
          "updates": 27,
          "samples_seen": 27648,
          "online_loss": 0.19032138586044312,
          "train_loss": 0.1762879490852356,
          "validation_loss": 0.15137168765068054,
          "train_metric": 0.01775507815182209,
          "validation_metric": 0.015245603397488594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24863076899964653
        },
        {
          "epoch": 28,
          "updates": 28,
          "samples_seen": 28672,
          "online_loss": 0.1762879341840744,
          "train_loss": 0.159781813621521,
          "validation_loss": 0.13958294689655304,
          "train_metric": 0.016092637553811073,
          "validation_metric": 0.014058283530175686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2539314829991781
        },
        {
          "epoch": 29,
          "updates": 29,
          "samples_seen": 29696,
          "online_loss": 0.1597817987203598,
          "train_loss": 0.14472904801368713,
          "validation_loss": 0.12753748893737793,
          "train_metric": 0.014576580375432968,
          "validation_metric": 0.012845107354223728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2593289469987212
        },
        {
          "epoch": 30,
          "updates": 30,
          "samples_seen": 30720,
          "online_loss": 0.14472904801368713,
          "train_loss": 0.13338318467140198,
          "validation_loss": 0.11724204570055008,
          "train_metric": 0.013433866202831268,
          "validation_metric": 0.011808189563453197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2650056729999051
        },
        {
          "epoch": 31,
          "updates": 31,
          "samples_seen": 31744,
          "online_loss": 0.13338318467140198,
          "train_loss": 0.12634509801864624,
          "validation_loss": 0.10961552709341049,
          "train_metric": 0.012725015170872211,
          "validation_metric": 0.011040073819458485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2712634879990219
        },
        {
          "epoch": 32,
          "updates": 32,
          "samples_seen": 32768,
          "online_loss": 0.12634509801864624,
          "train_loss": 0.12302938848733902,
          "validation_loss": 0.10467509180307388,
          "train_metric": 0.012391069903969765,
          "validation_metric": 0.0105424914509058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27707926599941857
        },
        {
          "epoch": 33,
          "updates": 33,
          "samples_seen": 33792,
          "online_loss": 0.12302938848733902,
          "train_loss": 0.12224037200212479,
          "validation_loss": 0.1018718034029007,
          "train_metric": 0.01231160294264555,
          "validation_metric": 0.010260154493153095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28284304799854
        },
        {
          "epoch": 34,
          "updates": 34,
          "samples_seen": 34816,
          "online_loss": 0.12224036455154419,
          "train_loss": 0.12265750020742416,
          "validation_loss": 0.10042364150285721,
          "train_metric": 0.012353615835309029,
          "validation_metric": 0.01011430099606514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2892764209991583
        },
        {
          "epoch": 35,
          "updates": 35,
          "samples_seen": 35840,
          "online_loss": 0.12265750765800476,
          "train_loss": 0.12315250188112259,
          "validation_loss": 0.09956672787666321,
          "train_metric": 0.012403469532728195,
          "validation_metric": 0.010027995333075523,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.295359401998212
        },
        {
          "epoch": 36,
          "updates": 36,
          "samples_seen": 36864,
          "online_loss": 0.12315249443054199,
          "train_loss": 0.12294331192970276,
          "validation_loss": 0.09870384633541107,
          "train_metric": 0.012382399290800095,
          "validation_metric": 0.009941089898347855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30100217599829193
        },
        {
          "epoch": 37,
          "updates": 37,
          "samples_seen": 37888,
          "online_loss": 0.12294329702854156,
          "train_loss": 0.12162825465202332,
          "validation_loss": 0.09746500849723816,
          "train_metric": 0.012249953113496304,
          "validation_metric": 0.00981631875038147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30835542699969665
        },
        {
          "epoch": 38,
          "updates": 38,
          "samples_seen": 38912,
          "online_loss": 0.12162826210260391,
          "train_loss": 0.11914759129285812,
          "validation_loss": 0.09570664912462234,
          "train_metric": 0.012000108137726784,
          "validation_metric": 0.009639223106205463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3140591349983879
        },
        {
          "epoch": 39,
          "updates": 39,
          "samples_seen": 39936,
          "online_loss": 0.11914759129285812,
          "train_loss": 0.11570756882429123,
          "validation_loss": 0.09347426891326904,
          "train_metric": 0.011653644032776356,
          "validation_metric": 0.009414385072886944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32042075599929376
        },
        {
          "epoch": 40,
          "updates": 40,
          "samples_seen": 40960,
          "online_loss": 0.11570758372545242,
          "train_loss": 0.11168835312128067,
          "validation_loss": 0.09094579517841339,
          "train_metric": 0.011248841881752014,
          "validation_metric": 0.009159727953374386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3265039589987282
        },
        {
          "epoch": 41,
          "updates": 41,
          "samples_seen": 41984,
          "online_loss": 0.11168834567070007,
          "train_loss": 0.10754545778036118,
          "validation_loss": 0.08836610615253448,
          "train_metric": 0.01083158515393734,
          "validation_metric": 0.008899910375475883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3321624039999733
        },
        {
          "epoch": 42,
          "updates": 42,
          "samples_seen": 43008,
          "online_loss": 0.10754545778036118,
          "train_loss": 0.10371240228414536,
          "validation_loss": 0.08597948402166367,
          "train_metric": 0.010445533320307732,
          "validation_metric": 0.008659538812935352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42578572999991593
        },
        {
          "epoch": 43,
          "updates": 43,
          "samples_seen": 44032,
          "online_loss": 0.10371240228414536,
          "train_loss": 0.10051248967647552,
          "validation_loss": 0.08396806567907333,
          "train_metric": 0.010123251006007195,
          "validation_metric": 0.008456956595182419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4316938959982508
        },
        {
          "epoch": 44,
          "updates": 44,
          "samples_seen": 45056,
          "online_loss": 0.10051250457763672,
          "train_loss": 0.0980939120054245,
          "validation_loss": 0.08240629732608795,
          "train_metric": 0.009879658930003643,
          "validation_metric": 0.008299660868942738,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4386692129992298
        },
        {
          "epoch": 45,
          "updates": 45,
          "samples_seen": 46080,
          "online_loss": 0.0980939045548439,
          "train_loss": 0.09640473127365112,
          "validation_loss": 0.0812433734536171,
          "train_metric": 0.009709532372653484,
          "validation_metric": 0.008182534947991371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4446024459994078
        },
        {
          "epoch": 46,
          "updates": 46,
          "samples_seen": 47104,
          "online_loss": 0.09640473127365112,
          "train_loss": 0.09522069990634918,
          "validation_loss": 0.08032234013080597,
          "train_metric": 0.009590279310941696,
          "validation_metric": 0.008089771494269371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4503752729997359
        },
        {
          "epoch": 47,
          "updates": 47,
          "samples_seen": 48128,
          "online_loss": 0.09522069990634918,
          "train_loss": 0.09422354400157928,
          "validation_loss": 0.07943321764469147,
          "train_metric": 0.00948985107243061,
          "validation_metric": 0.008000223897397518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45719089699923643
        },
        {
          "epoch": 48,
          "updates": 48,
          "samples_seen": 49152,
          "online_loss": 0.09422355890274048,
          "train_loss": 0.09310679882764816,
          "validation_loss": 0.07838327437639236,
          "train_metric": 0.0093773752450943,
          "validation_metric": 0.007894476875662804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46297487399897364
        },
        {
          "epoch": 49,
          "updates": 49,
          "samples_seen": 50176,
          "online_loss": 0.09310679882764816,
          "train_loss": 0.091671422123909,
          "validation_loss": 0.07705798745155334,
          "train_metric": 0.009232809767127037,
          "validation_metric": 0.007760997861623764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46932650599956105
        },
        {
          "epoch": 50,
          "updates": 50,
          "samples_seen": 51200,
          "online_loss": 0.0916714146733284,
          "train_loss": 0.08987609297037125,
          "validation_loss": 0.07545018196105957,
          "train_metric": 0.009051991626620293,
          "validation_metric": 0.007599066011607647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47585154799890006
        },
        {
          "epoch": 51,
          "updates": 51,
          "samples_seen": 52224,
          "online_loss": 0.08987608551979065,
          "train_loss": 0.08782747387886047,
          "validation_loss": 0.07364872843027115,
          "train_metric": 0.008845660835504532,
          "validation_metric": 0.0074176304042339325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48138570899936894
        },
        {
          "epoch": 52,
          "updates": 52,
          "samples_seen": 53248,
          "online_loss": 0.08782747387886047,
          "train_loss": 0.08572174608707428,
          "validation_loss": 0.071796715259552,
          "train_metric": 0.008633580058813095,
          "validation_metric": 0.007231102325022221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.487693462999232
        },
        {
          "epoch": 53,
          "updates": 53,
          "samples_seen": 54272,
          "online_loss": 0.08572175353765488,
          "train_loss": 0.08376579731702805,
          "validation_loss": 0.07003885507583618,
          "train_metric": 0.008436583913862705,
          "validation_metric": 0.007054056506603956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49356660899866256
        },
        {
          "epoch": 54,
          "updates": 54,
          "samples_seen": 55296,
          "online_loss": 0.08376579731702805,
          "train_loss": 0.08210919797420502,
          "validation_loss": 0.06847812980413437,
          "train_metric": 0.008269738405942917,
          "validation_metric": 0.0068968660198152065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49928310299947043
        },
        {
          "epoch": 55,
          "updates": 55,
          "samples_seen": 56320,
          "online_loss": 0.08210919797420502,
          "train_loss": 0.08080670237541199,
          "validation_loss": 0.06715337932109833,
          "train_metric": 0.00813855417072773,
          "validation_metric": 0.0067634424194693565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5062227609996626
        },
        {
          "epoch": 56,
          "updates": 56,
          "samples_seen": 57344,
          "online_loss": 0.08080669492483139,
          "train_loss": 0.07981666922569275,
          "validation_loss": 0.06604038178920746,
          "train_metric": 0.008038842119276524,
          "validation_metric": 0.006651346106082201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5120255539986829
        },
        {
          "epoch": 57,
          "updates": 57,
          "samples_seen": 58368,
          "online_loss": 0.07981666177511215,
          "train_loss": 0.07902746647596359,
          "validation_loss": 0.06507012993097305,
          "train_metric": 0.007959357462823391,
          "validation_metric": 0.00655362568795681,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5183061789994099
        },
        {
          "epoch": 58,
          "updates": 58,
          "samples_seen": 59392,
          "online_loss": 0.07902747392654419,
          "train_loss": 0.07829805463552475,
          "validation_loss": 0.06415574997663498,
          "train_metric": 0.007885892875492573,
          "validation_metric": 0.006461531855165958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5249021039999207
        },
        {
          "epoch": 59,
          "updates": 59,
          "samples_seen": 60416,
          "online_loss": 0.07829805463552475,
          "train_loss": 0.07749872654676437,
          "validation_loss": 0.06321925669908524,
          "train_metric": 0.007805387023836374,
          "validation_metric": 0.0063672116957604885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5307733289992029
        },
        {
          "epoch": 60,
          "updates": 60,
          "samples_seen": 61440,
          "online_loss": 0.07749871909618378,
          "train_loss": 0.07654216140508652,
          "validation_loss": 0.06221172958612442,
          "train_metric": 0.007709046360105276,
          "validation_metric": 0.006265738047659397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5372828319996188
        },
        {
          "epoch": 61,
          "updates": 61,
          "samples_seen": 62464,
          "online_loss": 0.07654216885566711,
          "train_loss": 0.07539970427751541,
          "validation_loss": 0.06112385541200638,
          "train_metric": 0.007593982852995396,
          "validation_metric": 0.006156171206384897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5436483019984735
        },
        {
          "epoch": 62,
          "updates": 62,
          "samples_seen": 63488,
          "online_loss": 0.07539971172809601,
          "train_loss": 0.07410182803869247,
          "validation_loss": 0.059985652565956116,
          "train_metric": 0.007463264279067516,
          "validation_metric": 0.006041535176336765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5493535559999145
        },
        {
          "epoch": 63,
          "updates": 63,
          "samples_seen": 64512,
          "online_loss": 0.07410182058811188,
          "train_loss": 0.07272505760192871,
          "validation_loss": 0.058856528252363205,
          "train_metric": 0.007324601523578167,
          "validation_metric": 0.00592781463637948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.555755904999387
        },
        {
          "epoch": 64,
          "updates": 64,
          "samples_seen": 65536,
          "online_loss": 0.07272505760192871,
          "train_loss": 0.07136928290128708,
          "validation_loss": 0.05780830234289169,
          "train_metric": 0.007188053335994482,
          "validation_metric": 0.005822240374982357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5615365909998218
        },
        {
          "epoch": 65,
          "updates": 65,
          "samples_seen": 66560,
          "online_loss": 0.07136929035186768,
          "train_loss": 0.07013064622879028,
          "validation_loss": 0.05690465122461319,
          "train_metric": 0.007063302211463451,
          "validation_metric": 0.005731228739023209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5671142209994287
        },
        {
          "epoch": 66,
          "updates": 66,
          "samples_seen": 67584,
          "online_loss": 0.07013064622879028,
          "train_loss": 0.06907621771097183,
          "validation_loss": 0.05618225410580635,
          "train_metric": 0.006957103032618761,
          "validation_metric": 0.005658470559865236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5740435829993658
        },
        {
          "epoch": 67,
          "updates": 67,
          "samples_seen": 68608,
          "online_loss": 0.06907621026039124,
          "train_loss": 0.06822699308395386,
          "validation_loss": 0.05563873052597046,
          "train_metric": 0.006871572230011225,
          "validation_metric": 0.005603729747235775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5797329229990282
        },
        {
          "epoch": 68,
          "updates": 68,
          "samples_seen": 69632,
          "online_loss": 0.06822699308395386,
          "train_loss": 0.06755425035953522,
          "validation_loss": 0.05523163080215454,
          "train_metric": 0.006803817581385374,
          "validation_metric": 0.005562727805227041,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5859508029989229
        },
        {
          "epoch": 69,
          "updates": 69,
          "samples_seen": 70656,
          "online_loss": 0.06755425035953522,
          "train_loss": 0.06699102371931076,
          "validation_loss": 0.054889969527721405,
          "train_metric": 0.006747090723365545,
          "validation_metric": 0.005528316367417574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5921607969994511
        },
        {
          "epoch": 70,
          "updates": 70,
          "samples_seen": 71680,
          "online_loss": 0.06699103116989136,
          "train_loss": 0.06645527482032776,
          "validation_loss": 0.054535090923309326,
          "train_metric": 0.006693131756037474,
          "validation_metric": 0.005492574535310268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5979108099982113
        },
        {
          "epoch": 71,
          "updates": 71,
          "samples_seen": 72704,
          "online_loss": 0.06645527482032776,
          "train_loss": 0.06587681919336319,
          "validation_loss": 0.054103780537843704,
          "train_metric": 0.0066348714753985405,
          "validation_metric": 0.005449134390801191,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6052746799996385
        },
        {
          "epoch": 72,
          "updates": 72,
          "samples_seen": 73728,
          "online_loss": 0.06587681919336319,
          "train_loss": 0.06521832942962646,
          "validation_loss": 0.053565338253974915,
          "train_metric": 0.00656855059787631,
          "validation_metric": 0.005394905339926481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6116530439994676
        },
        {
          "epoch": 73,
          "updates": 73,
          "samples_seen": 74752,
          "online_loss": 0.06521833688020706,
          "train_loss": 0.0644831731915474,
          "validation_loss": 0.05292689800262451,
          "train_metric": 0.006494509056210518,
          "validation_metric": 0.00533060310408473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6178409659987665
        },
        {
          "epoch": 74,
          "updates": 74,
          "samples_seen": 75776,
          "online_loss": 0.0644831657409668,
          "train_loss": 0.06370873004198074,
          "validation_loss": 0.05222616344690323,
          "train_metric": 0.006416509859263897,
          "validation_metric": 0.005260028410702944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6249401919994853
        },
        {
          "epoch": 75,
          "updates": 75,
          "samples_seen": 76800,
          "online_loss": 0.06370873004198074,
          "train_loss": 0.06294945627450943,
          "validation_loss": 0.05151605233550072,
          "train_metric": 0.00634003896266222,
          "validation_metric": 0.0051885079592466354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6306441219985572
        },
        {
          "epoch": 76,
          "updates": 76,
          "samples_seen": 77824,
          "online_loss": 0.06294944882392883,
          "train_loss": 0.06225680932402611,
          "validation_loss": 0.05084725841879845,
          "train_metric": 0.006270278245210648,
          "validation_metric": 0.005121150054037571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6368168389999482
        },
        {
          "epoch": 77,
          "updates": 77,
          "samples_seen": 78848,
          "online_loss": 0.06225680559873581,
          "train_loss": 0.06166328862309456,
          "validation_loss": 0.05025501921772957,
          "train_metric": 0.006210501305758953,
          "validation_metric": 0.005061501637101173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6430467809987022
        },
        {
          "epoch": 78,
          "updates": 78,
          "samples_seen": 79872,
          "online_loss": 0.06166329234838486,
          "train_loss": 0.06117480248212814,
          "validation_loss": 0.049752868711948395,
          "train_metric": 0.006161301862448454,
          "validation_metric": 0.0050109270960092545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6488110069985851
        },
        {
          "epoch": 79,
          "updates": 79,
          "samples_seen": 80896,
          "online_loss": 0.061174798756837845,
          "train_loss": 0.06077233701944351,
          "validation_loss": 0.04933410882949829,
          "train_metric": 0.00612076697871089,
          "validation_metric": 0.004968751221895218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7434306789982656
        },
        {
          "epoch": 80,
          "updates": 80,
          "samples_seen": 81920,
          "online_loss": 0.06077233701944351,
          "train_loss": 0.06042107567191124,
          "validation_loss": 0.04897862672805786,
          "train_metric": 0.006085389759391546,
          "validation_metric": 0.004932948388159275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7491526599987992
        },
        {
          "epoch": 81,
          "updates": 81,
          "samples_seen": 82944,
          "online_loss": 0.06042107939720154,
          "train_loss": 0.06008296459913254,
          "validation_loss": 0.04866224154829979,
          "train_metric": 0.006051335949450731,
          "validation_metric": 0.004901082720607519,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7555724609992467
        },
        {
          "epoch": 82,
          "updates": 82,
          "samples_seen": 83968,
          "online_loss": 0.06008296459913254,
          "train_loss": 0.05972862243652344,
          "validation_loss": 0.04836529865860939,
          "train_metric": 0.006015648599714041,
          "validation_metric": 0.00487117562443018,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7614050609990954
        },
        {
          "epoch": 83,
          "updates": 83,
          "samples_seen": 84992,
          "online_loss": 0.05972862243652344,
          "train_loss": 0.059344932436943054,
          "validation_loss": 0.04807794466614723,
          "train_metric": 0.005977005232125521,
          "validation_metric": 0.004842234309762716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7671204769994802
        },
        {
          "epoch": 84,
          "updates": 84,
          "samples_seen": 86016,
          "online_loss": 0.059344932436943054,
          "train_loss": 0.05893638730049133,
          "validation_loss": 0.04780101031064987,
          "train_metric": 0.005935857072472572,
          "validation_metric": 0.004814343061298132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7738227529989672
        },
        {
          "epoch": 85,
          "updates": 85,
          "samples_seen": 87040,
          "online_loss": 0.05893639102578163,
          "train_loss": 0.05852045118808746,
          "validation_loss": 0.047542721033096313,
          "train_metric": 0.005893965717405081,
          "validation_metric": 0.004788328427821398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7793678039997758
        },
        {
          "epoch": 86,
          "updates": 86,
          "samples_seen": 88064,
          "online_loss": 0.05852045118808746,
          "train_loss": 0.05811929702758789,
          "validation_loss": 0.04731298238039017,
          "train_metric": 0.0058535621501505375,
          "validation_metric": 0.004765191115438938,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7853525239988812
        },
        {
          "epoch": 87,
          "updates": 87,
          "samples_seen": 89088,
          "online_loss": 0.05811929330229759,
          "train_loss": 0.057751115411520004,
          "validation_loss": 0.04711766168475151,
          "train_metric": 0.005816480610519648,
          "validation_metric": 0.004745518788695335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7926601529998152
        },
        {
          "epoch": 88,
          "updates": 88,
          "samples_seen": 90112,
          "online_loss": 0.0577511191368103,
          "train_loss": 0.057424191385507584,
          "validation_loss": 0.046954862773418427,
          "train_metric": 0.005783554166555405,
          "validation_metric": 0.004729121923446655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7984049529986805
        },
        {
          "epoch": 89,
          "updates": 89,
          "samples_seen": 91136,
          "online_loss": 0.057424187660217285,
          "train_loss": 0.05713528394699097,
          "validation_loss": 0.04681447893381119,
          "train_metric": 0.005754456855356693,
          "validation_metric": 0.00471498304978013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8050796029983758
        },
        {
          "epoch": 90,
          "updates": 90,
          "samples_seen": 92160,
          "online_loss": 0.05713528394699097,
          "train_loss": 0.05687222257256508,
          "validation_loss": 0.04668070375919342,
          "train_metric": 0.005727962125092745,
          "validation_metric": 0.004701509140431881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8110715969996818
        },
        {
          "epoch": 91,
          "updates": 91,
          "samples_seen": 93184,
          "online_loss": 0.05687222629785538,
          "train_loss": 0.05661909282207489,
          "validation_loss": 0.04653644934296608,
          "train_metric": 0.005702468100935221,
          "validation_metric": 0.0046869805082678795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.816740962998665
        },
        {
          "epoch": 92,
          "updates": 92,
          "samples_seen": 94208,
          "online_loss": 0.05661910027265549,
          "train_loss": 0.05636191368103027,
          "validation_loss": 0.0463680773973465,
          "train_metric": 0.00567656522616744,
          "validation_metric": 0.0046700225211679935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.823138123998433
        },
        {
          "epoch": 93,
          "updates": 93,
          "samples_seen": 95232,
          "online_loss": 0.056361909955739975,
          "train_loss": 0.05609273910522461,
          "validation_loss": 0.04616893082857132,
          "train_metric": 0.0056494553573429585,
          "validation_metric": 0.004649965558201075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8289512459996331
        },
        {
          "epoch": 94,
          "updates": 94,
          "samples_seen": 96256,
          "online_loss": 0.05609273910522461,
          "train_loss": 0.05581117421388626,
          "validation_loss": 0.0459405891597271,
          "train_metric": 0.0056210970506072044,
          "validation_metric": 0.004626967944204807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8347926399983407
        },
        {
          "epoch": 95,
          "updates": 95,
          "samples_seen": 97280,
          "online_loss": 0.05581117421388626,
          "train_loss": 0.0555228628218174,
          "validation_loss": 0.045691512525081635,
          "train_metric": 0.005592059344053268,
          "validation_metric": 0.004601881839334965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8416377809990081
        },
        {
          "epoch": 96,
          "updates": 96,
          "samples_seen": 98304,
          "online_loss": 0.0555228628218174,
          "train_loss": 0.05523599311709404,
          "validation_loss": 0.04543362185359001,
          "train_metric": 0.0055631669238209724,
          "validation_metric": 0.004575907252728939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.847430722998979
        },
        {
          "epoch": 97,
          "updates": 97,
          "samples_seen": 99328,
          "online_loss": 0.05523598939180374,
          "train_loss": 0.0549573078751564,
          "validation_loss": 0.04517822712659836,
          "train_metric": 0.005535098724067211,
          "validation_metric": 0.004550185985863209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8547367629998917
        },
        {
          "epoch": 98,
          "updates": 98,
          "samples_seen": 100352,
          "online_loss": 0.0549573078751564,
          "train_loss": 0.054689329117536545,
          "validation_loss": 0.04493291303515434,
          "train_metric": 0.005508108995854855,
          "validation_metric": 0.004525478929281235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.860799906999091
        },
        {
          "epoch": 99,
          "updates": 99,
          "samples_seen": 101376,
          "online_loss": 0.05468933284282684,
          "train_loss": 0.05442975461483002,
          "validation_loss": 0.04470039904117584,
          "train_metric": 0.0054819658398628235,
          "validation_metric": 0.004502060357481241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8662936099990475
        },
        {
          "epoch": 100,
          "updates": 100,
          "samples_seen": 102400,
          "online_loss": 0.054429762065410614,
          "train_loss": 0.054173048585653305,
          "validation_loss": 0.044479478150606155,
          "train_metric": 0.005456110928207636,
          "validation_metric": 0.0044798096641898155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8730959919994348
        },
        {
          "epoch": 101,
          "updates": 101,
          "samples_seen": 103424,
          "online_loss": 0.054173048585653305,
          "train_loss": 0.05391303449869156,
          "validation_loss": 0.04426708444952965,
          "train_metric": 0.005429922603070736,
          "validation_metric": 0.00445841858163476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8788400949997595
        },
        {
          "epoch": 102,
          "updates": 102,
          "samples_seen": 104448,
          "online_loss": 0.05391303449869156,
          "train_loss": 0.053645357489585876,
          "validation_loss": 0.04406028985977173,
          "train_metric": 0.005402964539825916,
          "validation_metric": 0.004437590949237347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8849181509995105
        },
        {
          "epoch": 103,
          "updates": 103,
          "samples_seen": 105472,
          "online_loss": 0.05364536494016647,
          "train_loss": 0.05336873233318329,
          "validation_loss": 0.04385726898908615,
          "train_metric": 0.005375103559345007,
          "validation_metric": 0.004417143762111664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8910876079989976
        },
        {
          "epoch": 104,
          "updates": 104,
          "samples_seen": 106496,
          "online_loss": 0.05336873233318329,
          "train_loss": 0.05308463051915169,
          "validation_loss": 0.04365719109773636,
          "train_metric": 0.005346489138901234,
          "validation_metric": 0.004396992269903421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8968063709999115
        },
        {
          "epoch": 105,
          "updates": 105,
          "samples_seen": 107520,
          "online_loss": 0.053084634244441986,
          "train_loss": 0.05279609560966492,
          "validation_loss": 0.04345939680933952,
          "train_metric": 0.0053174300119280815,
          "validation_metric": 0.004377070814371109,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9031365019982331
        },
        {
          "epoch": 106,
          "updates": 106,
          "samples_seen": 108544,
          "online_loss": 0.052796099334955215,
          "train_loss": 0.052506159991025925,
          "validation_loss": 0.04326242208480835,
          "train_metric": 0.005288228392601013,
          "validation_metric": 0.004357232712209225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.909156402998633
        },
        {
          "epoch": 107,
          "updates": 107,
          "samples_seen": 109568,
          "online_loss": 0.05250616744160652,
          "train_loss": 0.052216529846191406,
          "validation_loss": 0.04306358844041824,
          "train_metric": 0.0052590565755963326,
          "validation_metric": 0.004337206482887268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9149157219999324
        },
        {
          "epoch": 108,
          "updates": 108,
          "samples_seen": 110592,
          "online_loss": 0.05221652239561081,
          "train_loss": 0.051927030086517334,
          "validation_loss": 0.0428590327501297,
          "train_metric": 0.005229900125414133,
          "validation_metric": 0.004316604696214199,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9268669559987757
        },
        {
          "epoch": 109,
          "updates": 109,
          "samples_seen": 111616,
          "online_loss": 0.051927026361227036,
          "train_loss": 0.05163603648543358,
          "validation_loss": 0.04264454543590546,
          "train_metric": 0.005200591869652271,
          "validation_metric": 0.004295002203434706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9326678599991283
        },
        {
          "epoch": 110,
          "updates": 110,
          "samples_seen": 112640,
          "online_loss": 0.05163602530956268,
          "train_loss": 0.05134138837456703,
          "validation_loss": 0.042416688054800034,
          "train_metric": 0.005170917138457298,
          "validation_metric": 0.004272053483873606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9407396619990323
        },
        {
          "epoch": 111,
          "updates": 111,
          "samples_seen": 113664,
          "online_loss": 0.05134139209985733,
          "train_loss": 0.051041629165410995,
          "validation_loss": 0.042173851281404495,
          "train_metric": 0.005140725988894701,
          "validation_metric": 0.004247595556080341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9465436519985815
        },
        {
          "epoch": 112,
          "updates": 112,
          "samples_seen": 114688,
          "online_loss": 0.051041629165410995,
          "train_loss": 0.05073654651641846,
          "validation_loss": 0.041916802525520325,
          "train_metric": 0.0051099988631904125,
          "validation_metric": 0.004221706185489893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9526418189998367
        },
        {
          "epoch": 113,
          "updates": 113,
          "samples_seen": 115712,
          "online_loss": 0.050736550241708755,
          "train_loss": 0.0504271574318409,
          "validation_loss": 0.04164842143654823,
          "train_metric": 0.005078837741166353,
          "validation_metric": 0.004194676410406828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9588347259996226
        },
        {
          "epoch": 114,
          "updates": 114,
          "samples_seen": 116736,
          "online_loss": 0.0504271537065506,
          "train_loss": 0.05011497065424919,
          "validation_loss": 0.0413728766143322,
          "train_metric": 0.005047395825386047,
          "validation_metric": 0.004166924394667149,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9645231619997503
        },
        {
          "epoch": 115,
          "updates": 115,
          "samples_seen": 117760,
          "online_loss": 0.05011496692895889,
          "train_loss": 0.049801211804151535,
          "validation_loss": 0.04109450802206993,
          "train_metric": 0.005015795584768057,
          "validation_metric": 0.004138888791203499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9710932999987563
        },
        {
          "epoch": 116,
          "updates": 116,
          "samples_seen": 118784,
          "online_loss": 0.049801211804151535,
          "train_loss": 0.04948633164167404,
          "validation_loss": 0.040816936641931534,
          "train_metric": 0.004984081722795963,
          "validation_metric": 0.004110931884497404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9770263809987227
        },
        {
          "epoch": 117,
          "updates": 117,
          "samples_seen": 119808,
          "online_loss": 0.04948632791638374,
          "train_loss": 0.04916992783546448,
          "validation_loss": 0.04054240509867668,
          "train_metric": 0.0049522146582603455,
          "validation_metric": 0.00408328277990222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0635907809992204
        },
        {
          "epoch": 118,
          "updates": 118,
          "samples_seen": 120832,
          "online_loss": 0.049169931560754776,
          "train_loss": 0.048851124942302704,
          "validation_loss": 0.040271807461977005,
          "train_metric": 0.004920106381177902,
          "validation_metric": 0.004056029487401247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.070527771998968
        },
        {
          "epoch": 119,
          "updates": 119,
          "samples_seen": 121856,
          "online_loss": 0.048851124942302704,
          "train_loss": 0.04852908477187157,
          "validation_loss": 0.04000496491789818,
          "train_metric": 0.004887671675533056,
          "validation_metric": 0.004029153846204281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0765575529985654
        },
        {
          "epoch": 120,
          "updates": 120,
          "samples_seen": 122880,
          "online_loss": 0.04852908477187157,
          "train_loss": 0.04820341616868973,
          "validation_loss": 0.03974107280373573,
          "train_metric": 0.004854870960116386,
          "validation_metric": 0.004002575296908617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0824708659984026
        },
        {
          "epoch": 121,
          "updates": 121,
          "samples_seen": 123904,
          "online_loss": 0.04820341244339943,
          "train_loss": 0.047874368727207184,
          "validation_loss": 0.039479225873947144,
          "train_metric": 0.0048217312432825565,
          "validation_metric": 0.003976202569901943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0888717149991862
        },
        {
          "epoch": 122,
          "updates": 122,
          "samples_seen": 124928,
          "online_loss": 0.047874368727207184,
          "train_loss": 0.04754269868135452,
          "validation_loss": 0.03921851888298988,
          "train_metric": 0.0047883265651762486,
          "validation_metric": 0.003949945792555809,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0947556089995487
        },
        {
          "epoch": 123,
          "updates": 123,
          "samples_seen": 125952,
          "online_loss": 0.047542694956064224,
          "train_loss": 0.047209229320287704,
          "validation_loss": 0.038958027958869934,
          "train_metric": 0.0047547402791678905,
          "validation_metric": 0.003923709969967604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1004338299990195
        },
        {
          "epoch": 124,
          "updates": 124,
          "samples_seen": 126976,
          "online_loss": 0.047209229320287704,
          "train_loss": 0.04687444865703583,
          "validation_loss": 0.03869657963514328,
          "train_metric": 0.004721023142337799,
          "validation_metric": 0.0038973777554929256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1074890939999023
        },
        {
          "epoch": 125,
          "updates": 125,
          "samples_seen": 128000,
          "online_loss": 0.04687444493174553,
          "train_loss": 0.04653835669159889,
          "validation_loss": 0.03843270614743233,
          "train_metric": 0.004687172826379538,
          "validation_metric": 0.0038708013016730547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.113191853999524
        },
        {
          "epoch": 126,
          "updates": 126,
          "samples_seen": 129024,
          "online_loss": 0.04653835669159889,
          "train_loss": 0.04620053991675377,
          "validation_loss": 0.03816484659910202,
          "train_metric": 0.00465314881876111,
          "validation_metric": 0.0038438239134848118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1211407789996883
        },
        {
          "epoch": 127,
          "updates": 127,
          "samples_seen": 130048,
          "online_loss": 0.04620054364204407,
          "train_loss": 0.045860469341278076,
          "validation_loss": 0.0378916971385479,
          "train_metric": 0.004618898965418339,
          "validation_metric": 0.003816313110291958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1281884759991954
        },
        {
          "epoch": 128,
          "updates": 128,
          "samples_seen": 131072,
          "online_loss": 0.045860469341278076,
          "train_loss": 0.04551783576607704,
          "validation_loss": 0.037612538784742355,
          "train_metric": 0.004584390204399824,
          "validation_metric": 0.003788196947425604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1338821099998313
        },
        {
          "epoch": 129,
          "updates": 129,
          "samples_seen": 132096,
          "online_loss": 0.04551784321665764,
          "train_loss": 0.04517269879579544,
          "validation_loss": 0.03732748329639435,
          "train_metric": 0.004549629054963589,
          "validation_metric": 0.003759487299248576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1408214179991774
        },
        {
          "epoch": 130,
          "updates": 130,
          "samples_seen": 133120,
          "online_loss": 0.04517269879579544,
          "train_loss": 0.04482542723417282,
          "validation_loss": 0.03703742474317551,
          "train_metric": 0.004514652770012617,
          "validation_metric": 0.003730273339897394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1464787059994705
        },
        {
          "epoch": 131,
          "updates": 131,
          "samples_seen": 134144,
          "online_loss": 0.04482542350888252,
          "train_loss": 0.044476527720689774,
          "validation_loss": 0.03674387186765671,
          "train_metric": 0.004479512572288513,
          "validation_metric": 0.003700708970427513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1529721789993346
        },
        {
          "epoch": 132,
          "updates": 132,
          "samples_seen": 135168,
          "online_loss": 0.04447651654481888,
          "train_loss": 0.04412638768553734,
          "validation_loss": 0.03644854202866554,
          "train_metric": 0.004444248043000698,
          "validation_metric": 0.0036709634587168694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1589235199990071
        },
        {
          "epoch": 133,
          "updates": 133,
          "samples_seen": 136192,
          "online_loss": 0.04412638768553734,
          "train_loss": 0.04377516731619835,
          "validation_loss": 0.036152999848127365,
          "train_metric": 0.004408874083310366,
          "validation_metric": 0.0036411979235708714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.164313127999776
        },
        {
          "epoch": 134,
          "updates": 134,
          "samples_seen": 137216,
          "online_loss": 0.04377516731619835,
          "train_loss": 0.04342272877693176,
          "validation_loss": 0.03585836663842201,
          "train_metric": 0.004373378586024046,
          "validation_metric": 0.0036115229595452547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.172465851999732
        },
        {
          "epoch": 135,
          "updates": 135,
          "samples_seen": 138240,
          "online_loss": 0.04342272877693176,
          "train_loss": 0.04306885227560997,
          "validation_loss": 0.03556516021490097,
          "train_metric": 0.004337737336754799,
          "validation_metric": 0.0035819928161799908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1782506149993424
        },
        {
          "epoch": 136,
          "updates": 136,
          "samples_seen": 139264,
          "online_loss": 0.04306885600090027,
          "train_loss": 0.04271339625120163,
          "validation_loss": 0.03527336195111275,
          "train_metric": 0.004301936831325293,
          "validation_metric": 0.0035526035353541374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.18383853999876
        },
        {
          "epoch": 137,
          "updates": 137,
          "samples_seen": 140288,
          "online_loss": 0.04271339252591133,
          "train_loss": 0.04235638305544853,
          "validation_loss": 0.034982506185770035,
          "train_metric": 0.0042659807950258255,
          "validation_metric": 0.003523309947922826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.191822320999563
        },
        {
          "epoch": 138,
          "updates": 138,
          "samples_seen": 141312,
          "online_loss": 0.04235638678073883,
          "train_loss": 0.04199808090925217,
          "validation_loss": 0.034691981971263885,
          "train_metric": 0.00422989297658205,
          "validation_metric": 0.0034940489567816257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1977253719996952
        },
        {
          "epoch": 139,
          "updates": 139,
          "samples_seen": 142336,
          "online_loss": 0.04199808090925217,
          "train_loss": 0.04163879528641701,
          "validation_loss": 0.03440107777714729,
          "train_metric": 0.004193706903606653,
          "validation_metric": 0.0034647504799067974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2111147469986463
        },
        {
          "epoch": 140,
          "updates": 140,
          "samples_seen": 143360,
          "online_loss": 0.04163879528641701,
          "train_loss": 0.04127883166074753,
          "validation_loss": 0.034109245985746384,
          "train_metric": 0.004157452378422022,
          "validation_metric": 0.0034353581722825766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2176888629983296
        },
        {
          "epoch": 141,
          "updates": 141,
          "samples_seen": 144384,
          "online_loss": 0.04127883166074753,
          "train_loss": 0.040918368846178055,
          "validation_loss": 0.03381602466106415,
          "train_metric": 0.004121148493140936,
          "validation_metric": 0.0034058259334415197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2330464229999052
        },
        {
          "epoch": 142,
          "updates": 142,
          "samples_seen": 145408,
          "online_loss": 0.04091837257146835,
          "train_loss": 0.04055747389793396,
          "validation_loss": 0.03352108597755432,
          "train_metric": 0.004084800835698843,
          "validation_metric": 0.0033761209342628717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2471594899998308
        },
        {
          "epoch": 143,
          "updates": 143,
          "samples_seen": 146432,
          "online_loss": 0.04055747762322426,
          "train_loss": 0.04019613564014435,
          "validation_loss": 0.03322424367070198,
          "train_metric": 0.004048408009111881,
          "validation_metric": 0.003346224082633853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2634263959989767
        },
        {
          "epoch": 144,
          "updates": 144,
          "samples_seen": 147456,
          "online_loss": 0.040196143090724945,
          "train_loss": 0.03983437269926071,
          "validation_loss": 0.03292541950941086,
          "train_metric": 0.004011972341686487,
          "validation_metric": 0.003316127462312579,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2780602949987951
        },
        {
          "epoch": 145,
          "updates": 145,
          "samples_seen": 148480,
          "online_loss": 0.03983437269926071,
          "train_loss": 0.03947228565812111,
          "validation_loss": 0.03262472152709961,
          "train_metric": 0.003975504543632269,
          "validation_metric": 0.0032858429476618767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2867817829992418
        },
        {
          "epoch": 146,
          "updates": 146,
          "samples_seen": 149504,
          "online_loss": 0.03947228565812111,
          "train_loss": 0.03911004960536957,
          "validation_loss": 0.03232242166996002,
          "train_metric": 0.003939020447432995,
          "validation_metric": 0.0032553961500525475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2995200519999344
        },
        {
          "epoch": 147,
          "updates": 147,
          "samples_seen": 150528,
          "online_loss": 0.03911004588007927,
          "train_loss": 0.03874783217906952,
          "validation_loss": 0.03201889991760254,
          "train_metric": 0.0039025400765240192,
          "validation_metric": 0.0032248261850327253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.306475296998542
        },
        {
          "epoch": 148,
          "updates": 148,
          "samples_seen": 151552,
          "online_loss": 0.03874783590435982,
          "train_loss": 0.03838580846786499,
          "validation_loss": 0.03171461820602417,
          "train_metric": 0.003866077633574605,
          "validation_metric": 0.0031941800843924284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3123382939993462
        },
        {
          "epoch": 149,
          "updates": 149,
          "samples_seen": 152576,
          "online_loss": 0.03838580846786499,
          "train_loss": 0.03802404925227165,
          "validation_loss": 0.03141006454825401,
          "train_metric": 0.003829642664641142,
          "validation_metric": 0.0031635062769055367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3194717889982712
        },
        {
          "epoch": 150,
          "updates": 150,
          "samples_seen": 153600,
          "online_loss": 0.038024045526981354,
          "train_loss": 0.03766259551048279,
          "validation_loss": 0.031105604022741318,
          "train_metric": 0.0037932386621832848,
          "validation_metric": 0.003132842481136322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.325856658999328
        },
        {
          "epoch": 151,
          "updates": 151,
          "samples_seen": 154624,
          "online_loss": 0.03766259178519249,
          "train_loss": 0.03730147331953049,
          "validation_loss": 0.03080148622393608,
          "train_metric": 0.0037568677216768265,
          "validation_metric": 0.003102213144302368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3315841559997352
        },
        {
          "epoch": 152,
          "updates": 152,
          "samples_seen": 155648,
          "online_loss": 0.03730147331953049,
          "train_loss": 0.036940742284059525,
          "validation_loss": 0.030497752130031586,
          "train_metric": 0.0037205363623797894,
          "validation_metric": 0.0030716219916939735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3384344599999167
        },
        {
          "epoch": 153,
          "updates": 153,
          "samples_seen": 156672,
          "online_loss": 0.036940742284059525,
          "train_loss": 0.036580491811037064,
          "validation_loss": 0.030194329097867012,
          "train_metric": 0.0036842534318566322,
          "validation_metric": 0.003041062271222472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.344643107999218
        },
        {
          "epoch": 154,
          "updates": 154,
          "samples_seen": 157696,
          "online_loss": 0.03658049553632736,
          "train_loss": 0.036220844835042953,
          "validation_loss": 0.02989107184112072,
          "train_metric": 0.003648030571639538,
          "validation_metric": 0.003010519314557314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4430422159985028
        },
        {
          "epoch": 155,
          "updates": 155,
          "samples_seen": 158720,
          "online_loss": 0.036220844835042953,
          "train_loss": 0.03586187958717346,
          "validation_loss": 0.02958785928785801,
          "train_metric": 0.003611877327784896,
          "validation_metric": 0.002979980781674385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4488536729986663
        },
        {
          "epoch": 156,
          "updates": 156,
          "samples_seen": 159744,
          "online_loss": 0.03586187958717346,
          "train_loss": 0.035503678023815155,
          "validation_loss": 0.029284661635756493,
          "train_metric": 0.0035758004523813725,
          "validation_metric": 0.002949443878605962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4553412789991853
        },
        {
          "epoch": 157,
          "updates": 157,
          "samples_seen": 160768,
          "online_loss": 0.035503678023815155,
          "train_loss": 0.03514626622200012,
          "validation_loss": 0.028981536626815796,
          "train_metric": 0.0035398032050579786,
          "validation_metric": 0.002918914193287492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4610670459987887
        },
        {
          "epoch": 158,
          "updates": 158,
          "samples_seen": 161792,
          "online_loss": 0.03514626622200012,
          "train_loss": 0.03478968143463135,
          "validation_loss": 0.02867860347032547,
          "train_metric": 0.003503889311105013,
          "validation_metric": 0.0028884040657430887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4664828659988416
        },
        {
          "epoch": 159,
          "updates": 159,
          "samples_seen": 162816,
          "online_loss": 0.03478968143463135,
          "train_loss": 0.034433964639902115,
          "validation_loss": 0.02837599813938141,
          "train_metric": 0.003468062961474061,
          "validation_metric": 0.002857926534488797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.473132960998555
        },
        {
          "epoch": 160,
          "updates": 160,
          "samples_seen": 163840,
          "online_loss": 0.034433964639902115,
          "train_loss": 0.0340791717171669,
          "validation_loss": 0.028073862195014954,
          "train_metric": 0.0034323290456086397,
          "validation_metric": 0.0028274967335164547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4787891199994192
        },
        {
          "epoch": 161,
          "updates": 161,
          "samples_seen": 164864,
          "online_loss": 0.0340791717171669,
          "train_loss": 0.03372538834810257,
          "validation_loss": 0.02777230367064476,
          "train_metric": 0.003396698273718357,
          "validation_metric": 0.002797124907374382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4842668119999871
        },
        {
          "epoch": 162,
          "updates": 162,
          "samples_seen": 165888,
          "online_loss": 0.03372538834810257,
          "train_loss": 0.03337269648909569,
          "validation_loss": 0.027471434324979782,
          "train_metric": 0.0033611757680773735,
          "validation_metric": 0.002766821999102831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4911750149985892
        },
        {
          "epoch": 163,
          "updates": 163,
          "samples_seen": 166912,
          "online_loss": 0.03337269276380539,
          "train_loss": 0.03302113711833954,
          "validation_loss": 0.027171334251761436,
          "train_metric": 0.0033257680479437113,
          "validation_metric": 0.0027365973219275475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.496783365999363
        },
        {
          "epoch": 164,
          "updates": 164,
          "samples_seen": 167936,
          "online_loss": 0.03302113711833954,
          "train_loss": 0.032670751214027405,
          "validation_loss": 0.02687207981944084,
          "train_metric": 0.0032904786057770252,
          "validation_metric": 0.002706457395106554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.50283290599873
        },
        {
          "epoch": 165,
          "updates": 165,
          "samples_seen": 168960,
          "online_loss": 0.032670751214027405,
          "train_loss": 0.03232157975435257,
          "validation_loss": 0.02657373994588852,
          "train_metric": 0.003255311166867614,
          "validation_metric": 0.002676410134881735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5088753719992383
        },
        {
          "epoch": 166,
          "updates": 166,
          "samples_seen": 169984,
          "online_loss": 0.03232157975435257,
          "train_loss": 0.03197367489337921,
          "validation_loss": 0.02627638727426529,
          "train_metric": 0.003220271784812212,
          "validation_metric": 0.0026464613620191813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5146073829982925
        },
        {
          "epoch": 167,
          "updates": 167,
          "samples_seen": 171008,
          "online_loss": 0.03197367116808891,
          "train_loss": 0.031627099961042404,
          "validation_loss": 0.025980109348893166,
          "train_metric": 0.00318536558188498,
          "validation_metric": 0.0026166215538978577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5209894749987143
        },
        {
          "epoch": 168,
          "updates": 168,
          "samples_seen": 172032,
          "online_loss": 0.0316271036863327,
          "train_loss": 0.03128194436430931,
          "validation_loss": 0.025685034692287445,
          "train_metric": 0.003150603035464883,
          "validation_metric": 0.002586902119219303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.526986411998223
        },
        {
          "epoch": 169,
          "updates": 169,
          "samples_seen": 173056,
          "online_loss": 0.03128194436430931,
          "train_loss": 0.030938295647501945,
          "validation_loss": 0.025391343981027603,
          "train_metric": 0.003115992061793804,
          "validation_metric": 0.002557323081418872,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.532534512998609
        },
        {
          "epoch": 170,
          "updates": 170,
          "samples_seen": 174080,
          "online_loss": 0.030938293784856796,
          "train_loss": 0.03059622459113598,
          "validation_loss": 0.02509925700724125,
          "train_metric": 0.00308153941296041,
          "validation_metric": 0.002527905162423849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5390984189998562
        },
        {
          "epoch": 171,
          "updates": 171,
          "samples_seen": 175104,
          "online_loss": 0.03059622459113598,
          "train_loss": 0.030255800113081932,
          "validation_loss": 0.02480899728834629,
          "train_metric": 0.0030472532380372286,
          "validation_metric": 0.002498671179637313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5449706569997943
        },
        {
          "epoch": 172,
          "updates": 172,
          "samples_seen": 176128,
          "online_loss": 0.03025580197572708,
          "train_loss": 0.02991708554327488,
          "validation_loss": 0.024520769715309143,
          "train_metric": 0.003013139357790351,
          "validation_metric": 0.002469642087817192,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5507304839993594
        },
        {
          "epoch": 173,
          "updates": 173,
          "samples_seen": 177152,
          "online_loss": 0.02991708368062973,
          "train_loss": 0.02958015352487564,
          "validation_loss": 0.024234741926193237,
          "train_metric": 0.0029792047571390867,
          "validation_metric": 0.0024408346507698298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5598097829988546
        },
        {
          "epoch": 174,
          "updates": 174,
          "samples_seen": 178176,
          "online_loss": 0.02958015352487564,
          "train_loss": 0.029245084151625633,
          "validation_loss": 0.02395101636648178,
          "train_metric": 0.0029454578179866076,
          "validation_metric": 0.0024122591130435467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.568040962998566
        },
        {
          "epoch": 175,
          "updates": 175,
          "samples_seen": 179200,
          "online_loss": 0.029245084151625633,
          "train_loss": 0.028911957517266273,
          "validation_loss": 0.023669691756367683,
          "train_metric": 0.0029119066894054413,
          "validation_metric": 0.0023839245550334454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.577674483998635
        },
        {
          "epoch": 176,
          "updates": 176,
          "samples_seen": 180224,
          "online_loss": 0.028911957517266273,
          "train_loss": 0.028580863028764725,
          "validation_loss": 0.023390810936689377,
          "train_metric": 0.0028785597532987595,
          "validation_metric": 0.002355836797505617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5834917509982915
        },
        {
          "epoch": 177,
          "updates": 177,
          "samples_seen": 181248,
          "online_loss": 0.028580859303474426,
          "train_loss": 0.028251860290765762,
          "validation_loss": 0.023114414885640144,
          "train_metric": 0.0028454242274165154,
          "validation_metric": 0.002327999332919717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5907198329987295
        },
        {
          "epoch": 178,
          "updates": 178,
          "samples_seen": 182272,
          "online_loss": 0.02825186215341091,
          "train_loss": 0.027925048023462296,
          "validation_loss": 0.02284049242734909,
          "train_metric": 0.002812508726492524,
          "validation_metric": 0.0023004107642918825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.597618567999234
        },
        {
          "epoch": 179,
          "updates": 179,
          "samples_seen": 183296,
          "online_loss": 0.027925049886107445,
          "train_loss": 0.027600480243563652,
          "validation_loss": 0.022568993270397186,
          "train_metric": 0.002779819071292877,
          "validation_metric": 0.0022730662021785975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6057317259983392
        },
        {
          "epoch": 180,
          "updates": 180,
          "samples_seen": 184320,
          "online_loss": 0.027600478380918503,
          "train_loss": 0.027278238907456398,
          "validation_loss": 0.022299926728010178,
          "train_metric": 0.0027473641093820333,
          "validation_metric": 0.0022459670435637236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6140120019990718
        },
        {
          "epoch": 181,
          "updates": 181,
          "samples_seen": 185344,
          "online_loss": 0.02727823704481125,
          "train_loss": 0.0269584059715271,
          "validation_loss": 0.022033311426639557,
          "train_metric": 0.0027151519898325205,
          "validation_metric": 0.002219114452600479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6224197449992062
        },
        {
          "epoch": 182,
          "updates": 182,
          "samples_seen": 186368,
          "online_loss": 0.0269584059715271,
          "train_loss": 0.026641061529517174,
          "validation_loss": 0.021769290789961815,
          "train_metric": 0.0026831901632249355,
          "validation_metric": 0.002192523330450058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.628869415999361
        },
        {
          "epoch": 183,
          "updates": 183,
          "samples_seen": 187392,
          "online_loss": 0.026641061529517174,
          "train_loss": 0.026326287537813187,
          "validation_loss": 0.021508030593395233,
          "train_metric": 0.0026514872442930937,
          "validation_metric": 0.002166209975257516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6344887019986345
        },
        {
          "epoch": 184,
          "updates": 184,
          "samples_seen": 188416,
          "online_loss": 0.026326289400458336,
          "train_loss": 0.026014156639575958,
          "validation_loss": 0.02124970033764839,
          "train_metric": 0.0026200509164482355,
          "validation_metric": 0.0021401920821517706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6413379299992812
        },
        {
          "epoch": 185,
          "updates": 185,
          "samples_seen": 189440,
          "online_loss": 0.026014156639575958,
          "train_loss": 0.025704752653837204,
          "validation_loss": 0.02099444344639778,
          "train_metric": 0.0025888881646096706,
          "validation_metric": 0.002114483155310154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6470048279988987
        },
        {
          "epoch": 186,
          "updates": 186,
          "samples_seen": 190464,
          "online_loss": 0.025704750791192055,
          "train_loss": 0.025398142635822296,
          "validation_loss": 0.020742319524288177,
          "train_metric": 0.002558007836341858,
          "validation_metric": 0.002089090645313263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6533028689991625
        },
        {
          "epoch": 187,
          "updates": 187,
          "samples_seen": 191488,
          "online_loss": 0.025398142635822296,
          "train_loss": 0.025094393640756607,
          "validation_loss": 0.020493362098932266,
          "train_metric": 0.002527415519580245,
          "validation_metric": 0.0020640164148062468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.65938095599995
        },
        {
          "epoch": 188,
          "updates": 188,
          "samples_seen": 192512,
          "online_loss": 0.025094393640756607,
          "train_loss": 0.02479359321296215,
          "validation_loss": 0.020247574895620346,
          "train_metric": 0.002497119829058647,
          "validation_metric": 0.0020392616279423237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6652209019994189
        },
        {
          "epoch": 189,
          "updates": 189,
          "samples_seen": 193536,
          "online_loss": 0.0247935950756073,
          "train_loss": 0.024495810270309448,
          "validation_loss": 0.020004991441965103,
          "train_metric": 0.0024671282153576612,
          "validation_metric": 0.0020148293115198612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6717846339997777
        },
        {
          "epoch": 190,
          "updates": 190,
          "samples_seen": 194560,
          "online_loss": 0.024495810270309448,
          "train_loss": 0.02420111373066902,
          "validation_loss": 0.019765622913837433,
          "train_metric": 0.0024374471977353096,
          "validation_metric": 0.001990721095353365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6820373449991166
        },
        {
          "epoch": 191,
          "updates": 191,
          "samples_seen": 195584,
          "online_loss": 0.02420111373066902,
          "train_loss": 0.023909570649266243,
          "validation_loss": 0.019529428333044052,
          "train_metric": 0.002408084226772189,
          "validation_metric": 0.0019669325556606054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.691419661998225
        },
        {
          "epoch": 192,
          "updates": 192,
          "samples_seen": 196608,
          "online_loss": 0.023909566923975945,
          "train_loss": 0.023621244356036186,
          "validation_loss": 0.019296351820230484,
          "train_metric": 0.002379045356065035,
          "validation_metric": 0.0019434577552601695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6976473119993898
        },
        {
          "epoch": 193,
          "updates": 193,
          "samples_seen": 197632,
          "online_loss": 0.023621242493391037,
          "train_loss": 0.023336203768849373,
          "validation_loss": 0.019066333770751953,
          "train_metric": 0.0023503368720412254,
          "validation_metric": 0.0019202912226319313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.707103795999501
        },
        {
          "epoch": 194,
          "updates": 194,
          "samples_seen": 198656,
          "online_loss": 0.023336203768849373,
          "train_loss": 0.02305450849235058,
          "validation_loss": 0.018839357420802116,
          "train_metric": 0.002321965526789427,
          "validation_metric": 0.001897431444376707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8271613789984258
        },
        {
          "epoch": 195,
          "updates": 195,
          "samples_seen": 199680,
          "online_loss": 0.02305450662970543,
          "train_loss": 0.022776229307055473,
          "validation_loss": 0.018615521490573883,
          "train_metric": 0.0022939383052289486,
          "validation_metric": 0.0018748871516436338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8330486159993598
        },
        {
          "epoch": 196,
          "updates": 196,
          "samples_seen": 200704,
          "online_loss": 0.022776229307055473,
          "train_loss": 0.022501425817608833,
          "validation_loss": 0.01839490979909897,
          "train_metric": 0.002266261028125882,
          "validation_metric": 0.0018526678904891014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8397196259993507
        },
        {
          "epoch": 197,
          "updates": 197,
          "samples_seen": 201728,
          "online_loss": 0.022501423954963684,
          "train_loss": 0.02223014086484909,
          "validation_loss": 0.01817762851715088,
          "train_metric": 0.002238938584923744,
          "validation_metric": 0.0018307841382920742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8458642689984117
        },
        {
          "epoch": 198,
          "updates": 198,
          "samples_seen": 202752,
          "online_loss": 0.02223013900220394,
          "train_loss": 0.021962448954582214,
          "validation_loss": 0.0179637148976326,
          "train_metric": 0.0022119772620499134,
          "validation_metric": 0.0018092396203428507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8536617989993829
        },
        {
          "epoch": 199,
          "updates": 199,
          "samples_seen": 203776,
          "online_loss": 0.021962447091937065,
          "train_loss": 0.021698372438549995,
          "validation_loss": 0.017753133550286293,
          "train_metric": 0.0021853807847946882,
          "validation_metric": 0.0017880306113511324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8661225359992386
        },
        {
          "epoch": 200,
          "updates": 200,
          "samples_seen": 204800,
          "online_loss": 0.021698370575904846,
          "train_loss": 0.021437974646687508,
          "validation_loss": 0.017545830458402634,
          "train_metric": 0.002159154275432229,
          "validation_metric": 0.0017671518726274371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8797399019986187
        },
        {
          "epoch": 201,
          "updates": 201,
          "samples_seen": 205824,
          "online_loss": 0.02143797278404236,
          "train_loss": 0.02118128538131714,
          "validation_loss": 0.01734180375933647,
          "train_metric": 0.0021333012264221907,
          "validation_metric": 0.0017466029385104775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8879057149988512
        },
        {
          "epoch": 202,
          "updates": 202,
          "samples_seen": 206848,
          "online_loss": 0.02118128351867199,
          "train_loss": 0.02092834748327732,
          "validation_loss": 0.01714104227721691,
          "train_metric": 0.0021078267600387335,
          "validation_metric": 0.0017263829940930009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8969785239987687
        },
        {
          "epoch": 203,
          "updates": 203,
          "samples_seen": 207872,
          "online_loss": 0.02092834934592247,
          "train_loss": 0.020679185166954994,
          "validation_loss": 0.01694352924823761,
          "train_metric": 0.002082731807604432,
          "validation_metric": 0.001706490060314536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9055610989998968
        },
        {
          "epoch": 204,
          "updates": 204,
          "samples_seen": 208896,
          "online_loss": 0.020679183304309845,
          "train_loss": 0.020433826372027397,
          "validation_loss": 0.016749193891882896,
          "train_metric": 0.0020580203272402287,
          "validation_metric": 0.001686917501501739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9146697879987187
        },
        {
          "epoch": 205,
          "updates": 205,
          "samples_seen": 209920,
          "online_loss": 0.020433828234672546,
          "train_loss": 0.020192300900816917,
          "validation_loss": 0.016557956114411354,
          "train_metric": 0.0020336946472525597,
          "validation_metric": 0.0016676567029207945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9232605949982826
        },
        {
          "epoch": 206,
          "updates": 206,
          "samples_seen": 210944,
          "online_loss": 0.020192302763462067,
          "train_loss": 0.019954625517129898,
          "validation_loss": 0.016369763761758804,
          "train_metric": 0.002009756863117218,
          "validation_metric": 0.0016487027751281857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9324795109987463
        },
        {
          "epoch": 207,
          "updates": 207,
          "samples_seen": 211968,
          "online_loss": 0.019954625517129898,
          "train_loss": 0.01972080022096634,
          "validation_loss": 0.016184642910957336,
          "train_metric": 0.00198620674200356,
          "validation_metric": 0.0016300580464303493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9410507459997461
        },
        {
          "epoch": 208,
          "updates": 208,
          "samples_seen": 212992,
          "online_loss": 0.01972080022096634,
          "train_loss": 0.019490854814648628,
          "validation_loss": 0.01600264571607113,
          "train_metric": 0.001963047543540597,
          "validation_metric": 0.0016117278719320893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.950050564999401
        },
        {
          "epoch": 209,
          "updates": 209,
          "samples_seen": 214016,
          "online_loss": 0.019490854814648628,
          "train_loss": 0.019264785572886467,
          "validation_loss": 0.015823811292648315,
          "train_metric": 0.0019402786856517196,
          "validation_metric": 0.0015937162097543478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9585209029992257
        },
        {
          "epoch": 210,
          "updates": 210,
          "samples_seen": 215040,
          "online_loss": 0.019264785572886467,
          "train_loss": 0.019042592495679855,
          "validation_loss": 0.015648137778043747,
          "train_metric": 0.001917900051921606,
          "validation_metric": 0.0015760231763124466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.967673441999068
        },
        {
          "epoch": 211,
          "updates": 211,
          "samples_seen": 216064,
          "online_loss": 0.019042592495679855,
          "train_loss": 0.018824277445673943,
          "validation_loss": 0.015475531108677387,
          "train_metric": 0.0018959123408421874,
          "validation_metric": 0.0015586388763040304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.976502191999316
        },
        {
          "epoch": 212,
          "updates": 212,
          "samples_seen": 217088,
          "online_loss": 0.018824279308319092,
          "train_loss": 0.018609829246997833,
          "validation_loss": 0.01530594751238823,
          "train_metric": 0.0018743140390142798,
          "validation_metric": 0.0015415591187775135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9853793289985333
        },
        {
          "epoch": 213,
          "updates": 213,
          "samples_seen": 218112,
          "online_loss": 0.018609829246997833,
          "train_loss": 0.018399236723780632,
          "validation_loss": 0.015139342285692692,
          "train_metric": 0.0018531037494540215,
          "validation_metric": 0.001524779130704701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9945137169997906
        },
        {
          "epoch": 214,
          "updates": 214,
          "samples_seen": 219136,
          "online_loss": 0.01839923858642578,
          "train_loss": 0.018192484974861145,
          "validation_loss": 0.014975660480558872,
          "train_metric": 0.0018322805408388376,
          "validation_metric": 0.0015082936733961105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0033868449991132
        },
        {
          "epoch": 215,
          "updates": 215,
          "samples_seen": 220160,
          "online_loss": 0.018192484974861145,
          "train_loss": 0.01798955351114273,
          "validation_loss": 0.014814858324825764,
          "train_metric": 0.0018118417356163263,
          "validation_metric": 0.0014920983230695128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0124963449998177
        },
        {
          "epoch": 216,
          "updates": 216,
          "samples_seen": 221184,
          "online_loss": 0.01798955351114273,
          "train_loss": 0.01779041439294815,
          "validation_loss": 0.014656833373010159,
          "train_metric": 0.0017917855875566602,
          "validation_metric": 0.0014761826023459435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0209333809998498
        },
        {
          "epoch": 217,
          "updates": 217,
          "samples_seen": 222208,
          "online_loss": 0.0177904162555933,
          "train_loss": 0.01759505085647106,
          "validation_loss": 0.014501494355499744,
          "train_metric": 0.0017721091862767935,
          "validation_metric": 0.0014605375472456217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.030517596998834
        },
        {
          "epoch": 218,
          "updates": 218,
          "samples_seen": 223232,
          "online_loss": 0.01759505271911621,
          "train_loss": 0.01740342564880848,
          "validation_loss": 0.014348805882036686,
          "train_metric": 0.001752809272147715,
          "validation_metric": 0.0014451593160629272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0398566289986775
        },
        {
          "epoch": 219,
          "updates": 219,
          "samples_seen": 224256,
          "online_loss": 0.01740342564880848,
          "train_loss": 0.01721549779176712,
          "validation_loss": 0.014198750257492065,
          "train_metric": 0.001733881887048483,
          "validation_metric": 0.0014300461625680327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.049097940998763
        },
        {
          "epoch": 220,
          "updates": 220,
          "samples_seen": 225280,
          "online_loss": 0.01721549779176712,
          "train_loss": 0.0170312337577343,
          "validation_loss": 0.014051360078155994,
          "train_metric": 0.0017153233056887984,
          "validation_metric": 0.001415201579220593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.057775770999797
        },
        {
          "epoch": 221,
          "updates": 221,
          "samples_seen": 226304,
          "online_loss": 0.01703123189508915,
          "train_loss": 0.01685059256851673,
          "validation_loss": 0.013906624168157578,
          "train_metric": 0.001697129919193685,
          "validation_metric": 0.0014006244018673897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.066683822999039
        },
        {
          "epoch": 222,
          "updates": 222,
          "samples_seen": 227328,
          "online_loss": 0.016850590705871582,
          "train_loss": 0.016673535108566284,
          "validation_loss": 0.013764486648142338,
          "train_metric": 0.0016792971873655915,
          "validation_metric": 0.0013863085769116879,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.074971898999138
        },
        {
          "epoch": 223,
          "updates": 223,
          "samples_seen": 228352,
          "online_loss": 0.016673533245921135,
          "train_loss": 0.016500014811754227,
          "validation_loss": 0.013624902814626694,
          "train_metric": 0.001661820919252932,
          "validation_metric": 0.001372250379063189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0830153099996096
        },
        {
          "epoch": 224,
          "updates": 224,
          "samples_seen": 229376,
          "online_loss": 0.016500011086463928,
          "train_loss": 0.016329960897564888,
          "validation_loss": 0.01348782517015934,
          "train_metric": 0.0016446937806904316,
          "validation_metric": 0.0013584443368017673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0915857959989808
        },
        {
          "epoch": 225,
          "updates": 225,
          "samples_seen": 230400,
          "online_loss": 0.016329960897564888,
          "train_loss": 0.01616334542632103,
          "validation_loss": 0.013353202491998672,
          "train_metric": 0.0016279129777103662,
          "validation_metric": 0.001344885560683906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.09990596499847
        },
        {
          "epoch": 226,
          "updates": 226,
          "samples_seen": 231424,
          "online_loss": 0.01616334542632103,
          "train_loss": 0.01600009761750698,
          "validation_loss": 0.013221010565757751,
          "train_metric": 0.0016114714089781046,
          "validation_metric": 0.0013315719552338123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1083103259989002
        },
        {
          "epoch": 227,
          "updates": 227,
          "samples_seen": 232448,
          "online_loss": 0.01600009761750698,
          "train_loss": 0.01584017090499401,
          "validation_loss": 0.013091150671243668,
          "train_metric": 0.001595364068634808,
          "validation_metric": 0.0013184929266571999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1168798949984193
        },
        {
          "epoch": 228,
          "updates": 228,
          "samples_seen": 233472,
          "online_loss": 0.01584016904234886,
          "train_loss": 0.015683487057685852,
          "validation_loss": 0.012963546440005302,
          "train_metric": 0.001579583389684558,
          "validation_metric": 0.0013056410243734717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1255331809989
        },
        {
          "epoch": 229,
          "updates": 229,
          "samples_seen": 234496,
          "online_loss": 0.015683488920331,
          "train_loss": 0.015530006028711796,
          "validation_loss": 0.012838119640946388,
          "train_metric": 0.0015641250647604465,
          "validation_metric": 0.001293008215725422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.133928442999604
        },
        {
          "epoch": 230,
          "updates": 230,
          "samples_seen": 235520,
          "online_loss": 0.015530005097389221,
          "train_loss": 0.015379653312265873,
          "validation_loss": 0.012714877724647522,
          "train_metric": 0.0015489824581891298,
          "validation_metric": 0.0012805958976969123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1425537379982416
        },
        {
          "epoch": 231,
          "updates": 231,
          "samples_seen": 236544,
          "online_loss": 0.015379655174911022,
          "train_loss": 0.015232358127832413,
          "validation_loss": 0.01259379182010889,
          "train_metric": 0.001534147304482758,
          "validation_metric": 0.001268400577828288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2477507629992033
        },
        {
          "epoch": 232,
          "updates": 232,
          "samples_seen": 237568,
          "online_loss": 0.015232358127832413,
          "train_loss": 0.015088064596056938,
          "validation_loss": 0.012474863789975643,
          "train_metric": 0.0015196145977824926,
          "validation_metric": 0.0012564227217808366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.256613423998715
        },
        {
          "epoch": 233,
          "updates": 233,
          "samples_seen": 238592,
          "online_loss": 0.015088065527379513,
          "train_loss": 0.014946703799068928,
          "validation_loss": 0.012358047068119049,
          "train_metric": 0.0015053772367537022,
          "validation_metric": 0.0012446572072803974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.264917002999937
        },
        {
          "epoch": 234,
          "updates": 234,
          "samples_seen": 239616,
          "online_loss": 0.014946704730391502,
          "train_loss": 0.01480820495635271,
          "validation_loss": 0.012243317440152168,
          "train_metric": 0.0014914283528923988,
          "validation_metric": 0.0012331019388511777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.273238213998411
        },
        {
          "epoch": 235,
          "updates": 235,
          "samples_seen": 240640,
          "online_loss": 0.014808205887675285,
          "train_loss": 0.014672495424747467,
          "validation_loss": 0.012130625545978546,
          "train_metric": 0.0014777600299566984,
          "validation_metric": 0.0012217521434649825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.282449240999995
        },
        {
          "epoch": 236,
          "updates": 236,
          "samples_seen": 241664,
          "online_loss": 0.014672495424747467,
          "train_loss": 0.0145395053550601,
          "validation_loss": 0.012019959278404713,
          "train_metric": 0.0014643658651039004,
          "validation_metric": 0.0012106063077226281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.29189660999873
        },
        {
          "epoch": 237,
          "updates": 237,
          "samples_seen": 242688,
          "online_loss": 0.014539506286382675,
          "train_loss": 0.014409175142645836,
          "validation_loss": 0.011911289766430855,
          "train_metric": 0.0014512393390759826,
          "validation_metric": 0.001199661404825747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3004606399990735
        },
        {
          "epoch": 238,
          "updates": 238,
          "samples_seen": 243712,
          "online_loss": 0.014409175142645836,
          "train_loss": 0.014281423762440681,
          "validation_loss": 0.011804533191025257,
          "train_metric": 0.0014383726520463824,
          "validation_metric": 0.0011889094021171331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.309142970998437
        },
        {
          "epoch": 239,
          "updates": 239,
          "samples_seen": 244736,
          "online_loss": 0.014281424693763256,
          "train_loss": 0.014156187884509563,
          "validation_loss": 0.011699625290930271,
          "train_metric": 0.0014257594011723995,
          "validation_metric": 0.0011783433146774769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.317667639999854
        },
        {
          "epoch": 240,
          "updates": 240,
          "samples_seen": 245760,
          "online_loss": 0.014156188815832138,
          "train_loss": 0.01403339859098196,
          "validation_loss": 0.011596514843404293,
          "train_metric": 0.0014133926015347242,
          "validation_metric": 0.0011679584858939052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3268661929996597
        },
        {
          "epoch": 241,
          "updates": 241,
          "samples_seen": 246784,
          "online_loss": 0.014033399522304535,
          "train_loss": 0.013912971131503582,
          "validation_loss": 0.011495188809931278,
          "train_metric": 0.0014012636383995414,
          "validation_metric": 0.0011577532859519124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.335553079999954
        },
        {
          "epoch": 242,
          "updates": 242,
          "samples_seen": 247808,
          "online_loss": 0.013912972062826157,
          "train_loss": 0.013794847764074802,
          "validation_loss": 0.011395608074963093,
          "train_metric": 0.0013893668074160814,
          "validation_metric": 0.0011477239895612001,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3447539869994216
        },
        {
          "epoch": 243,
          "updates": 243,
          "samples_seen": 248832,
          "online_loss": 0.013794848695397377,
          "train_loss": 0.013678966090083122,
          "validation_loss": 0.011297738179564476,
          "train_metric": 0.001377695589326322,
          "validation_metric": 0.001137866755016148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.353675929000019
        },
        {
          "epoch": 244,
          "updates": 244,
          "samples_seen": 249856,
          "online_loss": 0.013678967952728271,
          "train_loss": 0.013565242290496826,
          "validation_loss": 0.011201547458767891,
          "train_metric": 0.00136624148581177,
          "validation_metric": 0.0011281786719337106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.36246038399986
        },
        {
          "epoch": 245,
          "updates": 245,
          "samples_seen": 250880,
          "online_loss": 0.013565242290496826,
          "train_loss": 0.013453617691993713,
          "validation_loss": 0.011107011698186398,
          "train_metric": 0.001354999141767621,
          "validation_metric": 0.0011186575284227729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.370914469998752
        },
        {
          "epoch": 246,
          "updates": 246,
          "samples_seen": 251904,
          "online_loss": 0.013453618623316288,
          "train_loss": 0.01334401685744524,
          "validation_loss": 0.01101410947740078,
          "train_metric": 0.001343960757367313,
          "validation_metric": 0.0011093007633462548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3796073399989837
        },
        {
          "epoch": 247,
          "updates": 247,
          "samples_seen": 252928,
          "online_loss": 0.01334401872009039,
          "train_loss": 0.013236390426754951,
          "validation_loss": 0.010922824963927269,
          "train_metric": 0.0013331207446753979,
          "validation_metric": 0.0011001069797202945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.38841894899997
        },
        {
          "epoch": 248,
          "updates": 248,
          "samples_seen": 253952,
          "online_loss": 0.013236389495432377,
          "train_loss": 0.01313064992427826,
          "validation_loss": 0.010833104141056538,
          "train_metric": 0.0013224710710346699,
          "validation_metric": 0.0010910704731941223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.397124671999336
        },
        {
          "epoch": 249,
          "updates": 249,
          "samples_seen": 254976,
          "online_loss": 0.013130651786923409,
          "train_loss": 0.013026751577854156,
          "validation_loss": 0.01074488926678896,
          "train_metric": 0.0013120067305862904,
          "validation_metric": 0.0010821858886629343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4055755749996024
        },
        {
          "epoch": 250,
          "updates": 250,
          "samples_seen": 256000,
          "online_loss": 0.013026750646531582,
          "train_loss": 0.012924624606966972,
          "validation_loss": 0.010658142156898975,
          "train_metric": 0.0013017207384109497,
          "validation_metric": 0.0010734490351751447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.414220861999638
        },
        {
          "epoch": 251,
          "updates": 251,
          "samples_seen": 257024,
          "online_loss": 0.012924623675644398,
          "train_loss": 0.012824207544326782,
          "validation_loss": 0.010572845116257668,
          "train_metric": 0.001291607040911913,
          "validation_metric": 0.0010648582829162478,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4231387709987757
        },
        {
          "epoch": 252,
          "updates": 252,
          "samples_seen": 258048,
          "online_loss": 0.012824206613004208,
          "train_loss": 0.012725435197353363,
          "validation_loss": 0.010488932020962238,
          "train_metric": 0.0012816591188311577,
          "validation_metric": 0.0010564068797975779,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.431762425998386
        },
        {
          "epoch": 253,
          "updates": 253,
          "samples_seen": 259072,
          "online_loss": 0.012725435197353363,
          "train_loss": 0.012628253549337387,
          "validation_loss": 0.010406396351754665,
          "train_metric": 0.0012718717334792018,
          "validation_metric": 0.001048094010911882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4403136569999333
        },
        {
          "epoch": 254,
          "updates": 254,
          "samples_seen": 260096,
          "online_loss": 0.012628254480659962,
          "train_loss": 0.012532608583569527,
          "validation_loss": 0.010325178503990173,
          "train_metric": 0.001262238365598023,
          "validation_metric": 0.0010399142047390342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4488675149987102
        },
        {
          "epoch": 255,
          "updates": 255,
          "samples_seen": 261120,
          "online_loss": 0.012532607652246952,
          "train_loss": 0.012438436038792133,
          "validation_loss": 0.010245261713862419,
          "train_metric": 0.001252753776498139,
          "validation_metric": 0.001031865132972598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.457125016999271
        },
        {
          "epoch": 256,
          "updates": 256,
          "samples_seen": 262144,
          "online_loss": 0.012438436038792133,
          "train_loss": 0.01234569400548935,
          "validation_loss": 0.010166634805500507,
          "train_metric": 0.001243413076736033,
          "validation_metric": 0.0010239462135359645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.465677711999888
        },
        {
          "epoch": 257,
          "updates": 257,
          "samples_seen": 263168,
          "online_loss": 0.012345693074166775,
          "train_loss": 0.012254317291080952,
          "validation_loss": 0.010089287534356117,
          "train_metric": 0.0012342099798843265,
          "validation_metric": 0.0010161561658605933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4742388279992156
        },
        {
          "epoch": 258,
          "updates": 258,
          "samples_seen": 264192,
          "online_loss": 0.012254317291080952,
          "train_loss": 0.012164264917373657,
          "validation_loss": 0.010013164021074772,
          "train_metric": 0.0012251401785761118,
          "validation_metric": 0.001008489285595715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.482630659998904
        },
        {
          "epoch": 259,
          "updates": 259,
          "samples_seen": 265216,
          "online_loss": 0.012164264917373657,
          "train_loss": 0.01207547914236784,
          "validation_loss": 0.009938233532011509,
          "train_metric": 0.0012161980848759413,
          "validation_metric": 0.0010009425459429622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.490982481998799
        },
        {
          "epoch": 260,
          "updates": 260,
          "samples_seen": 266240,
          "online_loss": 0.012075480073690414,
          "train_loss": 0.01198792178183794,
          "validation_loss": 0.009864467196166515,
          "train_metric": 0.0012073797406628728,
          "validation_metric": 0.0009935131529346108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4995238199990126
        },
        {
          "epoch": 261,
          "updates": 261,
          "samples_seen": 267264,
          "online_loss": 0.011987922713160515,
          "train_loss": 0.01190153881907463,
          "validation_loss": 0.00979185476899147,
          "train_metric": 0.0011986795580014586,
          "validation_metric": 0.000986199826002121,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5078093869997247
        },
        {
          "epoch": 262,
          "updates": 262,
          "samples_seen": 268288,
          "online_loss": 0.011901539750397205,
          "train_loss": 0.011816293001174927,
          "validation_loss": 0.009720349684357643,
          "train_metric": 0.0011900938116014004,
          "validation_metric": 0.0009789981413632631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.516413086999819
        },
        {
          "epoch": 263,
          "updates": 263,
          "samples_seen": 269312,
          "online_loss": 0.011816292069852352,
          "train_loss": 0.011732149869203568,
          "validation_loss": 0.009649911895394325,
          "train_metric": 0.0011816190090030432,
          "validation_metric": 0.0009719038498587906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.52504163499907
        },
        {
          "epoch": 264,
          "updates": 264,
          "samples_seen": 270336,
          "online_loss": 0.011732148006558418,
          "train_loss": 0.01164904423058033,
          "validation_loss": 0.009580502286553383,
          "train_metric": 0.0011732492130249739,
          "validation_metric": 0.0009649130515754223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.533694691999699
        },
        {
          "epoch": 265,
          "updates": 265,
          "samples_seen": 271360,
          "online_loss": 0.011649045161902905,
          "train_loss": 0.011566968634724617,
          "validation_loss": 0.009512129239737988,
          "train_metric": 0.001164982677437365,
          "validation_metric": 0.0009580268524587154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5425546239985124
        },
        {
          "epoch": 266,
          "updates": 266,
          "samples_seen": 272384,
          "online_loss": 0.011566968634724617,
          "train_loss": 0.011485865339636803,
          "validation_loss": 0.009444748982787132,
          "train_metric": 0.0011568142799660563,
          "validation_metric": 0.000951240595895797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.551377109999521
        },
        {
          "epoch": 267,
          "updates": 267,
          "samples_seen": 273408,
          "online_loss": 0.011485865339636803,
          "train_loss": 0.011405707336962223,
          "validation_loss": 0.009378346614539623,
          "train_metric": 0.001148741110228002,
          "validation_metric": 0.0009445527102798223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.560332519999065
        },
        {
          "epoch": 268,
          "updates": 268,
          "samples_seen": 274432,
          "online_loss": 0.011405707336962223,
          "train_loss": 0.011326462961733341,
          "validation_loss": 0.009312888607382774,
          "train_metric": 0.001140759908594191,
          "validation_metric": 0.0009379599941894412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5683977310000046
        },
        {
          "epoch": 269,
          "updates": 269,
          "samples_seen": 275456,
          "online_loss": 0.011326462961733341,
          "train_loss": 0.011248103342950344,
          "validation_loss": 0.009248367510735989,
          "train_metric": 0.0011328676482662559,
          "validation_metric": 0.0009314616327174008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6774814929995046
        },
        {
          "epoch": 270,
          "updates": 270,
          "samples_seen": 276480,
          "online_loss": 0.011248103342950344,
          "train_loss": 0.011170594021677971,
          "validation_loss": 0.009184745140373707,
          "train_metric": 0.001125061186030507,
          "validation_metric": 0.0009250539005734026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.686033015999783
        },
        {
          "epoch": 271,
          "updates": 271,
          "samples_seen": 277504,
          "online_loss": 0.011170594021677971,
          "train_loss": 0.011093907058238983,
          "validation_loss": 0.00912200566381216,
          "train_metric": 0.001117337727919221,
          "validation_metric": 0.0009187350515276194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.694800002998818
        },
        {
          "epoch": 272,
          "updates": 272,
          "samples_seen": 278528,
          "online_loss": 0.011093907989561558,
          "train_loss": 0.011018019169569016,
          "validation_loss": 0.009060116484761238,
          "train_metric": 0.0011096944799646735,
          "validation_metric": 0.0009125017677433789,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7028908759984915
        },
        {
          "epoch": 273,
          "updates": 273,
          "samples_seen": 279552,
          "online_loss": 0.011018018238246441,
          "train_loss": 0.010942908003926277,
          "validation_loss": 0.008999047800898552,
          "train_metric": 0.0011021295795217156,
          "validation_metric": 0.0009063511388376355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7110280289998627
        },
        {
          "epoch": 274,
          "updates": 274,
          "samples_seen": 280576,
          "online_loss": 0.010942908935248852,
          "train_loss": 0.010868547484278679,
          "validation_loss": 0.008938778191804886,
          "train_metric": 0.0010946402326226234,
          "validation_metric": 0.0009002810111269355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.719264151999596
        },
        {
          "epoch": 275,
          "updates": 275,
          "samples_seen": 281600,
          "online_loss": 0.01086854562163353,
          "train_loss": 0.010794912464916706,
          "validation_loss": 0.008879288099706173,
          "train_metric": 0.0010872241109609604,
          "validation_metric": 0.0008942894055508077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7278162789989437
        },
        {
          "epoch": 276,
          "updates": 276,
          "samples_seen": 282624,
          "online_loss": 0.01079491339623928,
          "train_loss": 0.010721991769969463,
          "validation_loss": 0.008820555172860622,
          "train_metric": 0.0010798797011375427,
          "validation_metric": 0.0008883739938028157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.736161735998394
        },
        {
          "epoch": 277,
          "updates": 277,
          "samples_seen": 283648,
          "online_loss": 0.010721991769969463,
          "train_loss": 0.010649763979017735,
          "validation_loss": 0.008762560784816742,
          "train_metric": 0.0010726051405072212,
          "validation_metric": 0.000882532971445471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7451339269991877
        },
        {
          "epoch": 278,
          "updates": 278,
          "samples_seen": 284672,
          "online_loss": 0.010649763979017735,
          "train_loss": 0.010578200221061707,
          "validation_loss": 0.008705265820026398,
          "train_metric": 0.001065397635102272,
          "validation_metric": 0.0008767624967731535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7535840639993694
        },
        {
          "epoch": 279,
          "updates": 279,
          "samples_seen": 285696,
          "online_loss": 0.010578199289739132,
          "train_loss": 0.010507294908165932,
          "validation_loss": 0.008648672141134739,
          "train_metric": 0.0010582561371847987,
          "validation_metric": 0.0008710625115782022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.762086911998267
        },
        {
          "epoch": 280,
          "updates": 280,
          "samples_seen": 286720,
          "online_loss": 0.010507293976843357,
          "train_loss": 0.010437027551233768,
          "validation_loss": 0.008592752739787102,
          "train_metric": 0.0010511792497709394,
          "validation_metric": 0.0008654305129311979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.770457399999941
        },
        {
          "epoch": 281,
          "updates": 281,
          "samples_seen": 287744,
          "online_loss": 0.010437028482556343,
          "train_loss": 0.010367385111749172,
          "validation_loss": 0.00853748805820942,
          "train_metric": 0.0010441651102155447,
          "validation_metric": 0.0008598645799793303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7794661189982435
        },
        {
          "epoch": 282,
          "updates": 282,
          "samples_seen": 288768,
          "online_loss": 0.010367385111749172,
          "train_loss": 0.0102983508259058,
          "validation_loss": 0.008482852950692177,
          "train_metric": 0.0010372123215347528,
          "validation_metric": 0.0008543618605472147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7881574899984116
        },
        {
          "epoch": 283,
          "updates": 283,
          "samples_seen": 289792,
          "online_loss": 0.010298351757228374,
          "train_loss": 0.010229918174445629,
          "validation_loss": 0.00842882227152586,
          "train_metric": 0.0010303199524059892,
          "validation_metric": 0.0008489202009513974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.796939226998802
        },
        {
          "epoch": 284,
          "updates": 284,
          "samples_seen": 290816,
          "online_loss": 0.010229918174445629,
          "train_loss": 0.010162069462239742,
          "validation_loss": 0.008375389501452446,
          "train_metric": 0.00102348648943007,
          "validation_metric": 0.0008435386116616428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8054215619995375
        },
        {
          "epoch": 285,
          "updates": 285,
          "samples_seen": 291840,
          "online_loss": 0.010162070393562317,
          "train_loss": 0.010094789788126945,
          "validation_loss": 0.008322522975504398,
          "train_metric": 0.0010167101863771677,
          "validation_metric": 0.0008382140658795834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.813841475999652
        },
        {
          "epoch": 286,
          "updates": 286,
          "samples_seen": 292864,
          "online_loss": 0.010094789788126945,
          "train_loss": 0.010028081014752388,
          "validation_loss": 0.0082702012732625,
          "train_metric": 0.0010099916253238916,
          "validation_metric": 0.0008329442935064435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.822247520000019
        },
        {
          "epoch": 287,
          "updates": 287,
          "samples_seen": 293888,
          "online_loss": 0.010028080083429813,
          "train_loss": 0.009961917996406555,
          "validation_loss": 0.008218408562242985,
          "train_metric": 0.0010033277794718742,
          "validation_metric": 0.0008277282468043268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8319882529995084
        },
        {
          "epoch": 288,
          "updates": 288,
          "samples_seen": 294912,
          "online_loss": 0.009961916133761406,
          "train_loss": 0.009896294213831425,
          "validation_loss": 0.008167136460542679,
          "train_metric": 0.0009967185324057937,
          "validation_metric": 0.0008225640631280839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8405007919991476
        },
        {
          "epoch": 289,
          "updates": 289,
          "samples_seen": 295936,
          "online_loss": 0.009896294213831425,
          "train_loss": 0.00983121246099472,
          "validation_loss": 0.008116361685097218,
          "train_metric": 0.000990163767710328,
          "validation_metric": 0.0008174502290785313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.84909007799979
        },
        {
          "epoch": 290,
          "updates": 290,
          "samples_seen": 296960,
          "online_loss": 0.00983121246099472,
          "train_loss": 0.009766650386154652,
          "validation_loss": 0.008066065609455109,
          "train_metric": 0.0009836612734943628,
          "validation_metric": 0.0008123844745568931,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.857788619998246
        },
        {
          "epoch": 291,
          "updates": 291,
          "samples_seen": 297984,
          "online_loss": 0.009766650386154652,
          "train_loss": 0.009702605195343494,
          "validation_loss": 0.008016236126422882,
          "train_metric": 0.0009772108169272542,
          "validation_metric": 0.0008073658682405949,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.866115422999428
        },
        {
          "epoch": 292,
          "updates": 292,
          "samples_seen": 299008,
          "online_loss": 0.00970260426402092,
          "train_loss": 0.009639074094593525,
          "validation_loss": 0.007966865785419941,
          "train_metric": 0.0009708123980090022,
          "validation_metric": 0.0008023935952223837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8747275239984447
        },
        {
          "epoch": 293,
          "updates": 293,
          "samples_seen": 300032,
          "online_loss": 0.0096390750259161,
          "train_loss": 0.00957604218274355,
          "validation_loss": 0.007917910814285278,
          "train_metric": 0.0009644638630561531,
          "validation_metric": 0.0007974631153047085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8830963619984686
        },
        {
          "epoch": 294,
          "updates": 294,
          "samples_seen": 301056,
          "online_loss": 0.009576041251420975,
          "train_loss": 0.009513500146567822,
          "validation_loss": 0.007869383320212364,
          "train_metric": 0.0009581649792380631,
          "validation_metric": 0.0007925754180178046,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.894000516998858
        },
        {
          "epoch": 295,
          "updates": 295,
          "samples_seen": 302080,
          "online_loss": 0.009513501077890396,
          "train_loss": 0.009451450780034065,
          "validation_loss": 0.007821260020136833,
          "train_metric": 0.0009519153973087668,
          "validation_metric": 0.0007877286989241838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.902556178998566
        },
        {
          "epoch": 296,
          "updates": 296,
          "samples_seen": 303104,
          "online_loss": 0.00945144984871149,
          "train_loss": 0.009389878250658512,
          "validation_loss": 0.007773526478558779,
          "train_metric": 0.0009457141859456897,
          "validation_metric": 0.0007829211535863578,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.917581292998875
        },
        {
          "epoch": 297,
          "updates": 297,
          "samples_seen": 304128,
          "online_loss": 0.009389877319335938,
          "train_loss": 0.009328783489763737,
          "validation_loss": 0.007726164534687996,
          "train_metric": 0.0009395609376952052,
          "validation_metric": 0.0007781509193591774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.933031744998516
        },
        {
          "epoch": 298,
          "updates": 298,
          "samples_seen": 305152,
          "online_loss": 0.009328783489763737,
          "train_loss": 0.009268156252801418,
          "validation_loss": 0.007679159287363291,
          "train_metric": 0.0009334547794423997,
          "validation_metric": 0.0007734167738817632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.946145550999063
        },
        {
          "epoch": 299,
          "updates": 299,
          "samples_seen": 306176,
          "online_loss": 0.009268156252801418,
          "train_loss": 0.009207993745803833,
          "validation_loss": 0.007632524240761995,
          "train_metric": 0.0009273954201489687,
          "validation_metric": 0.0007687198813073337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9576225629989494
        },
        {
          "epoch": 300,
          "updates": 300,
          "samples_seen": 307200,
          "online_loss": 0.009207993745803833,
          "train_loss": 0.009148284792900085,
          "validation_loss": 0.007586218416690826,
          "train_metric": 0.0009213816956616938,
          "validation_metric": 0.000764056108891964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9670046479986922
        },
        {
          "epoch": 301,
          "updates": 301,
          "samples_seen": 308224,
          "online_loss": 0.009148284792900085,
          "train_loss": 0.009089033119380474,
          "validation_loss": 0.0075402408838272095,
          "train_metric": 0.0009154140716418624,
          "validation_metric": 0.000759425456635654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9819274799992854
        },
        {
          "epoch": 302,
          "updates": 302,
          "samples_seen": 309248,
          "online_loss": 0.009089034050703049,
          "train_loss": 0.009030223824083805,
          "validation_loss": 0.007494592573493719,
          "train_metric": 0.0009094910928979516,
          "validation_metric": 0.0007548278663307428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.000606757999776
        },
        {
          "epoch": 303,
          "updates": 303,
          "samples_seen": 310272,
          "online_loss": 0.009030223824083805,
          "train_loss": 0.008971857838332653,
          "validation_loss": 0.00744926230981946,
          "train_metric": 0.0009036127012223005,
          "validation_metric": 0.0007502624066546559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0156659559997934
        },
        {
          "epoch": 304,
          "updates": 304,
          "samples_seen": 311296,
          "online_loss": 0.008971856907010078,
          "train_loss": 0.008913928642868996,
          "validation_loss": 0.007404227275401354,
          "train_metric": 0.0008977783145383,
          "validation_metric": 0.0007457266910932958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0262604009985807
        },
        {
          "epoch": 305,
          "updates": 305,
          "samples_seen": 312320,
          "online_loss": 0.008913928642868996,
          "train_loss": 0.00885642971843481,
          "validation_loss": 0.007359490264207125,
          "train_metric": 0.0008919871179386973,
          "validation_metric": 0.0007412208942696452,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.035477938999975
        },
        {
          "epoch": 306,
          "updates": 306,
          "samples_seen": 313344,
          "online_loss": 0.00885642971843481,
          "train_loss": 0.008799359202384949,
          "validation_loss": 0.0073150270618498325,
          "train_metric": 0.0008862392278388143,
          "validation_metric": 0.0007367426296696067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.142241727999135
        },
        {
          "epoch": 307,
          "updates": 307,
          "samples_seen": 314368,
          "online_loss": 0.008799359202384949,
          "train_loss": 0.008742716163396835,
          "validation_loss": 0.007270858157426119,
          "train_metric": 0.0008805343532003462,
          "validation_metric": 0.0007322942255996168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.15091640799983
        },
        {
          "epoch": 308,
          "updates": 308,
          "samples_seen": 315392,
          "online_loss": 0.00874271523207426,
          "train_loss": 0.008686481043696404,
          "validation_loss": 0.007226952351629734,
          "train_metric": 0.000874870631378144,
          "validation_metric": 0.0007278722478076816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1596408619989234
        },
        {
          "epoch": 309,
          "updates": 309,
          "samples_seen": 316416,
          "online_loss": 0.008686481043696404,
          "train_loss": 0.008630668744444847,
          "validation_loss": 0.0071833087131381035,
          "train_metric": 0.000869249168317765,
          "validation_metric": 0.0007234765798784792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.168256895000013
        },
        {
          "epoch": 310,
          "updates": 310,
          "samples_seen": 317440,
          "online_loss": 0.008630666881799698,
          "train_loss": 0.008575260639190674,
          "validation_loss": 0.007139925379306078,
          "train_metric": 0.0008636688580736518,
          "validation_metric": 0.0007191071636043489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.176924349998444
        },
        {
          "epoch": 311,
          "updates": 311,
          "samples_seen": 318464,
          "online_loss": 0.008575260639190674,
          "train_loss": 0.008520258590579033,
          "validation_loss": 0.007096801418811083,
          "train_metric": 0.0008581292349845171,
          "validation_metric": 0.0007147638825699687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1855057019984088
        },
        {
          "epoch": 312,
          "updates": 312,
          "samples_seen": 319488,
          "online_loss": 0.008520258590579033,
          "train_loss": 0.0084656598046422,
          "validation_loss": 0.007053927052766085,
          "train_metric": 0.0008526302408427,
          "validation_metric": 0.0007104457472451031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1943802759997197
        },
        {
          "epoch": 313,
          "updates": 313,
          "samples_seen": 320512,
          "online_loss": 0.0084656598046422,
          "train_loss": 0.008411458693444729,
          "validation_loss": 0.007011298555880785,
          "train_metric": 0.0008471713517792523,
          "validation_metric": 0.0007061522337608039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2027447369982838
        },
        {
          "epoch": 314,
          "updates": 314,
          "samples_seen": 321536,
          "online_loss": 0.008411458693444729,
          "train_loss": 0.008357652463018894,
          "validation_loss": 0.006968896370381117,
          "train_metric": 0.0008417520439252257,
          "validation_metric": 0.0007018817123025656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2111714929997106
        },
        {
          "epoch": 315,
          "updates": 315,
          "samples_seen": 322560,
          "online_loss": 0.008357652463018894,
          "train_loss": 0.008304235525429249,
          "validation_loss": 0.006926737725734711,
          "train_metric": 0.0008363722590729594,
          "validation_metric": 0.0006976356962695718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2199399819983228
        },
        {
          "epoch": 316,
          "updates": 316,
          "samples_seen": 323584,
          "online_loss": 0.008304236456751823,
          "train_loss": 0.008251198567450047,
          "validation_loss": 0.006884811911731958,
          "train_metric": 0.0008310304838232696,
          "validation_metric": 0.0006934130215086043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2286112189995038
        },
        {
          "epoch": 317,
          "updates": 317,
          "samples_seen": 324608,
          "online_loss": 0.008251199498772621,
          "train_loss": 0.008198549039661884,
          "validation_loss": 0.006843103561550379,
          "train_metric": 0.0008257278241217136,
          "validation_metric": 0.0006892122910358012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2390572459989926
        },
        {
          "epoch": 318,
          "updates": 318,
          "samples_seen": 325632,
          "online_loss": 0.008198549039661884,
          "train_loss": 0.008146271109580994,
          "validation_loss": 0.006801615934818983,
          "train_metric": 0.000820462591946125,
          "validation_metric": 0.000685033795889467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2478484629991726
        },
        {
          "epoch": 319,
          "updates": 319,
          "samples_seen": 326656,
          "online_loss": 0.008146270178258419,
          "train_loss": 0.008094376884400845,
          "validation_loss": 0.0067603494971990585,
          "train_metric": 0.000815235951449722,
          "validation_metric": 0.0006808776524849236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.256197823999173
        },
        {
          "epoch": 320,
          "updates": 320,
          "samples_seen": 327680,
          "online_loss": 0.008094376884400845,
          "train_loss": 0.008042850531637669,
          "validation_loss": 0.006719300057739019,
          "train_metric": 0.0008100465056486428,
          "validation_metric": 0.0006767433369532228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2651010199988377
        },
        {
          "epoch": 321,
          "updates": 321,
          "samples_seen": 328704,
          "online_loss": 0.008042850531637669,
          "train_loss": 0.00799169484525919,
          "validation_loss": 0.006678459700196981,
          "train_metric": 0.0008048943127505481,
          "validation_metric": 0.0006726299761794508,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2741786189999402
        },
        {
          "epoch": 322,
          "updates": 322,
          "samples_seen": 329728,
          "online_loss": 0.00799169484525919,
          "train_loss": 0.007940898649394512,
          "validation_loss": 0.006637827027589083,
          "train_metric": 0.0007997781503945589,
          "validation_metric": 0.0006685377447865903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2831069749991
        },
        {
          "epoch": 323,
          "updates": 323,
          "samples_seen": 330752,
          "online_loss": 0.007940899580717087,
          "train_loss": 0.007890474982559681,
          "validation_loss": 0.006597411353141069,
          "train_metric": 0.0007946997648105025,
          "validation_metric": 0.0006644671084359288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.292258221999873
        },
        {
          "epoch": 324,
          "updates": 324,
          "samples_seen": 331776,
          "online_loss": 0.007890474982559681,
          "train_loss": 0.007840398699045181,
          "validation_loss": 0.006557188928127289,
          "train_metric": 0.0007896562456153333,
          "validation_metric": 0.0006604160298593342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.301020246999542
        },
        {
          "epoch": 325,
          "updates": 325,
          "samples_seen": 332800,
          "online_loss": 0.007840399630367756,
          "train_loss": 0.007790687028318644,
          "validation_loss": 0.006517165806144476,
          "train_metric": 0.0007846495136618614,
          "validation_metric": 0.0006563850911334157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3099384409997583
        },
        {
          "epoch": 326,
          "updates": 326,
          "samples_seen": 333824,
          "online_loss": 0.007790687493979931,
          "train_loss": 0.007741323672235012,
          "validation_loss": 0.006477340590208769,
          "train_metric": 0.0007796776480972767,
          "validation_metric": 0.0006523739430122077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3181968289991346
        },
        {
          "epoch": 327,
          "updates": 327,
          "samples_seen": 334848,
          "online_loss": 0.007741323206573725,
          "train_loss": 0.007692311890423298,
          "validation_loss": 0.006437714211642742,
          "train_metric": 0.0007747415220364928,
          "validation_metric": 0.0006483830511569977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.326896996999494
        },
        {
          "epoch": 328,
          "updates": 328,
          "samples_seen": 335872,
          "online_loss": 0.007692311890423298,
          "train_loss": 0.007643647026270628,
          "validation_loss": 0.006398287136107683,
          "train_metric": 0.0007698400877416134,
          "validation_metric": 0.00064441206632182,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.335274071998356
        },
        {
          "epoch": 329,
          "updates": 329,
          "samples_seen": 336896,
          "online_loss": 0.007643646560609341,
          "train_loss": 0.007595326751470566,
          "validation_loss": 0.0063590435311198235,
          "train_metric": 0.000764973578043282,
          "validation_metric": 0.0006404595915228128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3441158199984784
        },
        {
          "epoch": 330,
          "updates": 330,
          "samples_seen": 337920,
          "online_loss": 0.007595326751470566,
          "train_loss": 0.007547348737716675,
          "validation_loss": 0.006319986190646887,
          "train_metric": 0.0007601413526572287,
          "validation_metric": 0.0006365259177982807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3526323179994506
        },
        {
          "epoch": 331,
          "updates": 331,
          "samples_seen": 338944,
          "online_loss": 0.007547348737716675,
          "train_loss": 0.007499709725379944,
          "validation_loss": 0.006281119305640459,
          "train_metric": 0.0007553433533757925,
          "validation_metric": 0.0006326113361865282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.361122157999489
        },
        {
          "epoch": 332,
          "updates": 332,
          "samples_seen": 339968,
          "online_loss": 0.0074997092597186565,
          "train_loss": 0.007452409714460373,
          "validation_loss": 0.006242451258003712,
          "train_metric": 0.0007505793473683298,
          "validation_metric": 0.0006287167780101299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.369464749999679
        },
        {
          "epoch": 333,
          "updates": 333,
          "samples_seen": 340992,
          "online_loss": 0.007452409714460373,
          "train_loss": 0.007405436597764492,
          "validation_loss": 0.006203946191817522,
          "train_metric": 0.0007458485197275877,
          "validation_metric": 0.0006248387508094311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3782206769992627
        },
        {
          "epoch": 334,
          "updates": 334,
          "samples_seen": 342016,
          "online_loss": 0.007405437063425779,
          "train_loss": 0.007358805742114782,
          "validation_loss": 0.006165639031678438,
          "train_metric": 0.0007411519181914628,
          "validation_metric": 0.0006209805724211037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3864594519982347
        },
        {
          "epoch": 335,
          "updates": 335,
          "samples_seen": 343040,
          "online_loss": 0.007358805742114782,
          "train_loss": 0.007312502712011337,
          "validation_loss": 0.006127506028860807,
          "train_metric": 0.0007364884950220585,
          "validation_metric": 0.0006171399727463722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3954268069992395
        },
        {
          "epoch": 336,
          "updates": 336,
          "samples_seen": 344064,
          "online_loss": 0.00731250224635005,
          "train_loss": 0.007266521919518709,
          "validation_loss": 0.006089556496590376,
          "train_metric": 0.0007318576099351048,
          "validation_metric": 0.0006133178249001503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4038110069996037
        },
        {
          "epoch": 337,
          "updates": 337,
          "samples_seen": 345088,
          "online_loss": 0.0072665223851799965,
          "train_loss": 0.00722087174654007,
          "validation_loss": 0.006051777862012386,
          "train_metric": 0.0007272597867995501,
          "validation_metric": 0.0006095129647292197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4123497839991614
        },
        {
          "epoch": 338,
          "updates": 338,
          "samples_seen": 346112,
          "online_loss": 0.007220872677862644,
          "train_loss": 0.007175542414188385,
          "validation_loss": 0.0060141924768686295,
          "train_metric": 0.0007226943271234632,
          "validation_metric": 0.0006057273712940514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.420567023998956
        },
        {
          "epoch": 339,
          "updates": 339,
          "samples_seen": 347136,
          "online_loss": 0.007175542879849672,
          "train_loss": 0.007130539510399103,
          "validation_loss": 0.0059767733328044415,
          "train_metric": 0.0007181618711911142,
          "validation_metric": 0.000601958658080548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.429512535998583
        },
        {
          "epoch": 340,
          "updates": 340,
          "samples_seen": 348160,
          "online_loss": 0.007130540907382965,
          "train_loss": 0.007085850462317467,
          "validation_loss": 0.005939533933997154,
          "train_metric": 0.0007136609638109803,
          "validation_metric": 0.0005982081056572497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.437864150999303
        },
        {
          "epoch": 341,
          "updates": 341,
          "samples_seen": 349184,
          "online_loss": 0.007085850927978754,
          "train_loss": 0.007041489705443382,
          "validation_loss": 0.005902469158172607,
          "train_metric": 0.0007091930019669235,
          "validation_metric": 0.0005944751319475472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4467057769998064
        },
        {
          "epoch": 342,
          "updates": 342,
          "samples_seen": 350208,
          "online_loss": 0.007041489705443382,
          "train_loss": 0.006997436285018921,
          "validation_loss": 0.005865579005330801,
          "train_metric": 0.0007047561812214553,
          "validation_metric": 0.0005907596787437797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.455439862998901
        },
        {
          "epoch": 343,
          "updates": 343,
          "samples_seen": 351232,
          "online_loss": 0.006997435819357634,
          "train_loss": 0.006953697185963392,
          "validation_loss": 0.005828861612826586,
          "train_metric": 0.0007003509090282023,
          "validation_metric": 0.0005870615714229643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.463960772998689
        },
        {
          "epoch": 344,
          "updates": 344,
          "samples_seen": 352256,
          "online_loss": 0.0069536976516246796,
          "train_loss": 0.006910279393196106,
          "validation_loss": 0.005792323034256697,
          "train_metric": 0.0006959781749173999,
          "validation_metric": 0.000583381624892354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.472604959999444
        },
        {
          "epoch": 345,
          "updates": 345,
          "samples_seen": 353280,
          "online_loss": 0.006910279393196106,
          "train_loss": 0.006867167539894581,
          "validation_loss": 0.005755954887717962,
          "train_metric": 0.000691635999828577,
          "validation_metric": 0.0005797187332063913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4812380939983996
        },
        {
          "epoch": 346,
          "updates": 346,
          "samples_seen": 354304,
          "online_loss": 0.006867167539894581,
          "train_loss": 0.006824370939284563,
          "validation_loss": 0.005719758104532957,
          "train_metric": 0.0006873256643302739,
          "validation_metric": 0.000576073070988059,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5898022669989587
        },
        {
          "epoch": 347,
          "updates": 347,
          "samples_seen": 355328,
          "online_loss": 0.006824371870607138,
          "train_loss": 0.006781881675124168,
          "validation_loss": 0.005683736875653267,
          "train_metric": 0.0006830462953075767,
          "validation_metric": 0.0005724452203139663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.600165431998903
        },
        {
          "epoch": 348,
          "updates": 348,
          "samples_seen": 356352,
          "online_loss": 0.006781882140785456,
          "train_loss": 0.006739698350429535,
          "validation_loss": 0.005647891666740179,
          "train_metric": 0.0006787978345528245,
          "validation_metric": 0.0005688349483534694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6090671559995826
        },
        {
          "epoch": 349,
          "updates": 349,
          "samples_seen": 357376,
          "online_loss": 0.006739698816090822,
          "train_loss": 0.006697824224829674,
          "validation_loss": 0.005612215492874384,
          "train_metric": 0.0006745802238583565,
          "validation_metric": 0.0005652418476529419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.618217409999488
        },
        {
          "epoch": 350,
          "updates": 350,
          "samples_seen": 358400,
          "online_loss": 0.006697824224829674,
          "train_loss": 0.006656260695308447,
          "validation_loss": 0.0055767083540558815,
          "train_metric": 0.0006703942781314254,
          "validation_metric": 0.0005616656853817403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6275849729991023
        },
        {
          "epoch": 351,
          "updates": 351,
          "samples_seen": 359424,
          "online_loss": 0.006656261160969734,
          "train_loss": 0.006614997051656246,
          "validation_loss": 0.005541387479752302,
          "train_metric": 0.0006662381929345429,
          "validation_metric": 0.0005581082077696919,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.636816911999631
        },
        {
          "epoch": 352,
          "updates": 352,
          "samples_seen": 360448,
          "online_loss": 0.006614996585994959,
          "train_loss": 0.006574041675776243,
          "validation_loss": 0.005506238434463739,
          "train_metric": 0.0006621134816668928,
          "validation_metric": 0.0005545681342482567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6468552389997058
        },
        {
          "epoch": 353,
          "updates": 353,
          "samples_seen": 361472,
          "online_loss": 0.0065740421414375305,
          "train_loss": 0.006533386651426554,
          "validation_loss": 0.005471252836287022,
          "train_metric": 0.000658018805552274,
          "validation_metric": 0.0005510445334948599,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6557794909986114
        },
        {
          "epoch": 354,
          "updates": 354,
          "samples_seen": 362496,
          "online_loss": 0.006533387117087841,
          "train_loss": 0.006493037100881338,
          "validation_loss": 0.0054364497773349285,
          "train_metric": 0.0006539548630826175,
          "validation_metric": 0.0005475392681546509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6649223059994256
        },
        {
          "epoch": 355,
          "updates": 355,
          "samples_seen": 363520,
          "online_loss": 0.006493037100881338,
          "train_loss": 0.006452986970543861,
          "validation_loss": 0.00540183437988162,
          "train_metric": 0.000649921246804297,
          "validation_metric": 0.0005440529785118997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.679025349998483
        },
        {
          "epoch": 356,
          "updates": 356,
          "samples_seen": 364544,
          "online_loss": 0.006452986970543861,
          "train_loss": 0.006413239054381847,
          "validation_loss": 0.0053673735819756985,
          "train_metric": 0.0006459180149249732,
          "validation_metric": 0.0005405822303146124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.69070943099905
        },
        {
          "epoch": 357,
          "updates": 357,
          "samples_seen": 365568,
          "online_loss": 0.006413239520043135,
          "train_loss": 0.006373794749379158,
          "validation_loss": 0.0053330883383750916,
          "train_metric": 0.0006419452838599682,
          "validation_metric": 0.0005371291190385818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6999624709987984
        },
        {
          "epoch": 358,
          "updates": 358,
          "samples_seen": 366592,
          "online_loss": 0.006373794749379158,
          "train_loss": 0.006334646139293909,
          "validation_loss": 0.005299003329128027,
          "train_metric": 0.0006380024133250117,
          "validation_metric": 0.0005336962640285492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.715109572998699
        },
        {
          "epoch": 359,
          "updates": 359,
          "samples_seen": 367616,
          "online_loss": 0.006334646139293909,
          "train_loss": 0.006295798812061548,
          "validation_loss": 0.005265074782073498,
          "train_metric": 0.0006340898107737303,
          "validation_metric": 0.0005302790086716413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7237793619988224
        },
        {
          "epoch": 360,
          "updates": 360,
          "samples_seen": 368640,
          "online_loss": 0.006295798812061548,
          "train_loss": 0.006257253233343363,
          "validation_loss": 0.005231328774243593,
          "train_metric": 0.0006302076508291066,
          "validation_metric": 0.000526880263350904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7323704149985133
        },
        {
          "epoch": 361,
          "updates": 361,
          "samples_seen": 369664,
          "online_loss": 0.006257253233343363,
          "train_loss": 0.006219008006155491,
          "validation_loss": 0.00519776763394475,
          "train_metric": 0.000626355642452836,
          "validation_metric": 0.0005235001444816589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.740980115999264
        },
        {
          "epoch": 362,
          "updates": 362,
          "samples_seen": 370688,
          "online_loss": 0.006219008006155491,
          "train_loss": 0.006181060336530209,
          "validation_loss": 0.00516439089551568,
          "train_metric": 0.0006225336692295969,
          "validation_metric": 0.0005201385938562453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.749773384999571
        },
        {
          "epoch": 363,
          "updates": 363,
          "samples_seen": 371712,
          "online_loss": 0.006181059870868921,
          "train_loss": 0.006143410224467516,
          "validation_loss": 0.005131188314408064,
          "train_metric": 0.000618741731159389,
          "validation_metric": 0.0005167945055291057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.758493797999108
        },
        {
          "epoch": 364,
          "updates": 364,
          "samples_seen": 372736,
          "online_loss": 0.006143410224467516,
          "train_loss": 0.006106054410338402,
          "validation_loss": 0.005098170600831509,
          "train_metric": 0.000614979478996247,
          "validation_metric": 0.0005134690436534584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7675644559985813
        },
        {
          "epoch": 365,
          "updates": 365,
          "samples_seen": 373760,
          "online_loss": 0.0061060539446771145,
          "train_loss": 0.006069005001336336,
          "validation_loss": 0.00506533682346344,
          "train_metric": 0.0006112479022704065,
          "validation_metric": 0.0005101621500216424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7765330909987824
        },
        {
          "epoch": 366,
          "updates": 366,
          "samples_seen": 374784,
          "online_loss": 0.006069004535675049,
          "train_loss": 0.006032249424606562,
          "validation_loss": 0.005032697692513466,
          "train_metric": 0.0006075460696592927,
          "validation_metric": 0.0005068748141638935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7850924029989983
        },
        {
          "epoch": 367,
          "updates": 367,
          "samples_seen": 375808,
          "online_loss": 0.006032249424606562,
          "train_loss": 0.005995792802423239,
          "validation_loss": 0.005000242497771978,
          "train_metric": 0.000603874446824193,
          "validation_metric": 0.0005036060465499759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7939713889991253
        },
        {
          "epoch": 368,
          "updates": 368,
          "samples_seen": 376832,
          "online_loss": 0.005995793268084526,
          "train_loss": 0.005959637463092804,
          "validation_loss": 0.004967961926013231,
          "train_metric": 0.0006002327427268028,
          "validation_metric": 0.000500354974064976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8027528329985216
        },
        {
          "epoch": 369,
          "updates": 369,
          "samples_seen": 377856,
          "online_loss": 0.005959636531770229,
          "train_loss": 0.005923774559050798,
          "validation_loss": 0.004935888573527336,
          "train_metric": 0.0005966208991594613,
          "validation_metric": 0.0004971246235072613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.811811534998924
        },
        {
          "epoch": 370,
          "updates": 370,
          "samples_seen": 378880,
          "online_loss": 0.005923775024712086,
          "train_loss": 0.005888213869184256,
          "validation_loss": 0.004904003348201513,
          "train_metric": 0.0005930393235757947,
          "validation_metric": 0.0004939133068546653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8204915699989215
        },
        {
          "epoch": 371,
          "updates": 371,
          "samples_seen": 379904,
          "online_loss": 0.005888213869184256,
          "train_loss": 0.005852950271219015,
          "validation_loss": 0.004872299265116453,
          "train_metric": 0.0005894877249374986,
          "validation_metric": 0.0004907200927846134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.829667634998259
        },
        {
          "epoch": 372,
          "updates": 372,
          "samples_seen": 380928,
          "online_loss": 0.005852949805557728,
          "train_loss": 0.005817978642880917,
          "validation_loss": 0.004840805195271969,
          "train_metric": 0.0005859654629603028,
          "validation_metric": 0.00048754815361462533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.851402940999833
        },
        {
          "epoch": 373,
          "updates": 373,
          "samples_seen": 381952,
          "online_loss": 0.005817979108542204,
          "train_loss": 0.005783315282315016,
          "validation_loss": 0.0048095183447003365,
          "train_metric": 0.000582474225666374,
          "validation_metric": 0.0004843970818910748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8691514160000224
        },
        {
          "epoch": 374,
          "updates": 374,
          "samples_seen": 382976,
          "online_loss": 0.0057833148166537285,
          "train_loss": 0.005748943891376257,
          "validation_loss": 0.004778396803885698,
          "train_metric": 0.0005790124996565282,
          "validation_metric": 0.00048126265755854547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8845569039986003
        },
        {
          "epoch": 375,
          "updates": 375,
          "samples_seen": 384000,
          "online_loss": 0.005748943891376257,
          "train_loss": 0.0057148695923388,
          "validation_loss": 0.004747489001601934,
          "train_metric": 0.0005755807505920529,
          "validation_metric": 0.0004781497409567237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9007431519985403
        },
        {
          "epoch": 376,
          "updates": 376,
          "samples_seen": 385024,
          "online_loss": 0.0057148695923388,
          "train_loss": 0.005681093782186508,
          "validation_loss": 0.004716797266155481,
          "train_metric": 0.0005721789202652872,
          "validation_metric": 0.00047505853581242263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9101163939994876
        },
        {
          "epoch": 377,
          "updates": 377,
          "samples_seen": 386048,
          "online_loss": 0.005681093782186508,
          "train_loss": 0.005647615529596806,
          "validation_loss": 0.004686308559030294,
          "train_metric": 0.0005688071832992136,
          "validation_metric": 0.00047198779066093266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9190384529993025
        },
        {
          "epoch": 378,
          "updates": 378,
          "samples_seen": 387072,
          "online_loss": 0.00564761646091938,
          "train_loss": 0.005614434834569693,
          "validation_loss": 0.004656006582081318,
          "train_metric": 0.0005654653068631887,
          "validation_metric": 0.0004689359338954091,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9284015339999314
        },
        {
          "epoch": 379,
          "updates": 379,
          "samples_seen": 388096,
          "online_loss": 0.005614434368908405,
          "train_loss": 0.0055815549567341805,
          "validation_loss": 0.0046259211376309395,
          "train_metric": 0.0005621536984108388,
          "validation_metric": 0.00046590587589889765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9368718749992695
        },
        {
          "epoch": 380,
          "updates": 380,
          "samples_seen": 389120,
          "online_loss": 0.005581554025411606,
          "train_loss": 0.005548967979848385,
          "validation_loss": 0.004596059210598469,
          "train_metric": 0.0005588717176578939,
          "validation_metric": 0.0004628981987480074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.945829671998581
        },
        {
          "epoch": 381,
          "updates": 381,
          "samples_seen": 390144,
          "online_loss": 0.0055489675141870975,
          "train_loss": 0.005516680423170328,
          "validation_loss": 0.0045663947239518166,
          "train_metric": 0.0005556198302656412,
          "validation_metric": 0.0004599105741363019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.954455828999926
        },
        {
          "epoch": 382,
          "updates": 382,
          "samples_seen": 391168,
          "online_loss": 0.005516679957509041,
          "train_loss": 0.005484689958393574,
          "validation_loss": 0.004536937922239304,
          "train_metric": 0.000552397919818759,
          "validation_metric": 0.0004569437587633729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9632195769991085
        },
        {
          "epoch": 383,
          "updates": 383,
          "samples_seen": 392192,
          "online_loss": 0.005484690424054861,
          "train_loss": 0.005452997051179409,
          "validation_loss": 0.004507716279476881,
          "train_metric": 0.0005492058699019253,
          "validation_metric": 0.00045400066301226616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.072768388999975
        },
        {
          "epoch": 384,
          "updates": 384,
          "samples_seen": 393216,
          "online_loss": 0.005452997051179409,
          "train_loss": 0.005421598441898823,
          "validation_loss": 0.004478716757148504,
          "train_metric": 0.0005460435058921576,
          "validation_metric": 0.0004510799190029502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.081315824998455
        },
        {
          "epoch": 385,
          "updates": 385,
          "samples_seen": 394240,
          "online_loss": 0.005421598441898823,
          "train_loss": 0.005390496458858252,
          "validation_loss": 0.0044499109499156475,
          "train_metric": 0.0005429111188277602,
          "validation_metric": 0.00044817873276770115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.08991918099855
        },
        {
          "epoch": 386,
          "updates": 386,
          "samples_seen": 395264,
          "online_loss": 0.005390496458858252,
          "train_loss": 0.005359691567718983,
          "validation_loss": 0.004421335645020008,
          "train_metric": 0.0005398085340857506,
          "validation_metric": 0.0004453007713891566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.098506251999424
        },
        {
          "epoch": 387,
          "updates": 387,
          "samples_seen": 396288,
          "online_loss": 0.005359692499041557,
          "train_loss": 0.005329187028110027,
          "validation_loss": 0.004393002949655056,
          "train_metric": 0.000536736159119755,
          "validation_metric": 0.0004424471699167043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.10731277999912
        },
        {
          "epoch": 388,
          "updates": 388,
          "samples_seen": 397312,
          "online_loss": 0.005329187493771315,
          "train_loss": 0.005298967007547617,
          "validation_loss": 0.00436487328261137,
          "train_metric": 0.0005336925387382507,
          "validation_metric": 0.0004396140284370631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.116010777999691
        },
        {
          "epoch": 389,
          "updates": 389,
          "samples_seen": 398336,
          "online_loss": 0.005298967007547617,
          "train_loss": 0.005269044078886509,
          "validation_loss": 0.004336976911872625,
          "train_metric": 0.000530678837094456,
          "validation_metric": 0.0004368044901639223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.124575371999526
        },
        {
          "epoch": 390,
          "updates": 390,
          "samples_seen": 399360,
          "online_loss": 0.005269043613225222,
          "train_loss": 0.0052394201047718525,
          "validation_loss": 0.004309314768761396,
          "train_metric": 0.0005276951706036925,
          "validation_metric": 0.00043401835137046874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.133172598998499
        },
        {
          "epoch": 391,
          "updates": 391,
          "samples_seen": 400384,
          "online_loss": 0.0052394201047718525,
          "train_loss": 0.005210084840655327,
          "validation_loss": 0.004281886387616396,
          "train_metric": 0.0005247407243587077,
          "validation_metric": 0.0004312558739911765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.141705666999769
        },
        {
          "epoch": 392,
          "updates": 392,
          "samples_seen": 401408,
          "online_loss": 0.005210084840655327,
          "train_loss": 0.005181041080504656,
          "validation_loss": 0.004254676401615143,
          "train_metric": 0.0005218154401518404,
          "validation_metric": 0.0004285154864192009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.150634634999733
        },
        {
          "epoch": 393,
          "updates": 393,
          "samples_seen": 402432,
          "online_loss": 0.005181041546165943,
          "train_loss": 0.005152284167706966,
          "validation_loss": 0.004227702040225267,
          "train_metric": 0.000518919201567769,
          "validation_metric": 0.0004257986438460648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.159900510998341
        },
        {
          "epoch": 394,
          "updates": 394,
          "samples_seen": 403456,
          "online_loss": 0.005152284167706966,
          "train_loss": 0.005123825743794441,
          "validation_loss": 0.004200973082333803,
          "train_metric": 0.000516052998136729,
          "validation_metric": 0.0004231066268403083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.169075191999582
        },
        {
          "epoch": 395,
          "updates": 395,
          "samples_seen": 404480,
          "online_loss": 0.005123825743794441,
          "train_loss": 0.0050956495106220245,
          "validation_loss": 0.00417448440566659,
          "train_metric": 0.0005132152000442147,
          "validation_metric": 0.0004204388242214918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.1779041519985185
        },
        {
          "epoch": 396,
          "updates": 396,
          "samples_seen": 405504,
          "online_loss": 0.005095649976283312,
          "train_loss": 0.0050677587278187275,
          "validation_loss": 0.00414821645244956,
          "train_metric": 0.0005104060983285308,
          "validation_metric": 0.00041779314051382244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.18630936199952
        },
        {
          "epoch": 397,
          "updates": 397,
          "samples_seen": 406528,
          "online_loss": 0.005067759193480015,
          "train_loss": 0.005040156189352274,
          "validation_loss": 0.004122194834053516,
          "train_metric": 0.0005076260422356427,
          "validation_metric": 0.00041517234058119357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.19527238199953
        },
        {
          "epoch": 398,
          "updates": 398,
          "samples_seen": 407552,
          "online_loss": 0.005040156189352274,
          "train_loss": 0.005012839566916227,
          "validation_loss": 0.004096421878784895,
          "train_metric": 0.0005048748571425676,
          "validation_metric": 0.00041257659904658794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.203477434999513
        },
        {
          "epoch": 399,
          "updates": 399,
          "samples_seen": 408576,
          "online_loss": 0.005012840032577515,
          "train_loss": 0.004985809326171875,
          "validation_loss": 0.004070883616805077,
          "train_metric": 0.0005021524266339839,
          "validation_metric": 0.0004100044898223132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.212354617999154
        },
        {
          "epoch": 400,
          "updates": 400,
          "samples_seen": 409600,
          "online_loss": 0.004985809326171875,
          "train_loss": 0.0049590542912483215,
          "validation_loss": 0.004045586101710796,
          "train_metric": 0.0004994577611796558,
          "validation_metric": 0.000407456565881148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.221114428999499
        },
        {
          "epoch": 401,
          "updates": 401,
          "samples_seen": 410624,
          "online_loss": 0.004959053825587034,
          "train_loss": 0.0049325767904520035,
          "validation_loss": 0.004020534455776215,
          "train_metric": 0.0004967910936102271,
          "validation_metric": 0.0004049335257150233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2295198519987025
        },
        {
          "epoch": 402,
          "updates": 402,
          "samples_seen": 411648,
          "online_loss": 0.004932577256113291,
          "train_loss": 0.004906381014734507,
          "validation_loss": 0.003995727282017469,
          "train_metric": 0.0004941527149640024,
          "validation_metric": 0.0004024350200779736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.238904215999355
        },
        {
          "epoch": 403,
          "updates": 403,
          "samples_seen": 412672,
          "online_loss": 0.004906380549073219,
          "train_loss": 0.004880458116531372,
          "validation_loss": 0.003971152473241091,
          "train_metric": 0.0004915418685413897,
          "validation_metric": 0.0003999598848167807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.247935200999564
        },
        {
          "epoch": 404,
          "updates": 404,
          "samples_seen": 413696,
          "online_loss": 0.004880458116531372,
          "train_loss": 0.004854806233197451,
          "validation_loss": 0.003946841694414616,
          "train_metric": 0.0004889583215117455,
          "validation_metric": 0.00039751140866428614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.257511316998716
        },
        {
          "epoch": 405,
          "updates": 405,
          "samples_seen": 414720,
          "online_loss": 0.004854806698858738,
          "train_loss": 0.00482943095266819,
          "validation_loss": 0.003922765143215656,
          "train_metric": 0.00048640265595167875,
          "validation_metric": 0.00039508650661446154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.266824882999572
        },
        {
          "epoch": 406,
          "updates": 406,
          "samples_seen": 415744,
          "online_loss": 0.00482943095266819,
          "train_loss": 0.004804321099072695,
          "validation_loss": 0.0038989235181361437,
          "train_metric": 0.0004838736495003104,
          "validation_metric": 0.0003926852950826287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.2759813479988225
        },
        {
          "epoch": 407,
          "updates": 407,
          "samples_seen": 416768,
          "online_loss": 0.0048043206334114075,
          "train_loss": 0.004779482260346413,
          "validation_loss": 0.0038753547705709934,
          "train_metric": 0.0004813719424419105,
          "validation_metric": 0.0003903115866705775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.284932360998937
        },
        {
          "epoch": 408,
          "updates": 408,
          "samples_seen": 417792,
          "online_loss": 0.0047794827260077,
          "train_loss": 0.004754903260618448,
          "validation_loss": 0.0038520172238349915,
          "train_metric": 0.00047889642883092165,
          "validation_metric": 0.00038796107401140034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.294188655998369
        },
        {
          "epoch": 409,
          "updates": 409,
          "samples_seen": 418816,
          "online_loss": 0.004754903260618448,
          "train_loss": 0.004730585962533951,
          "validation_loss": 0.0038289176300168037,
          "train_metric": 0.00047644731239415705,
          "validation_metric": 0.00038563457201234996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.3032211839999945
        },
        {
          "epoch": 410,
          "updates": 410,
          "samples_seen": 419840,
          "online_loss": 0.004730586428195238,
          "train_loss": 0.004706528037786484,
          "validation_loss": 0.003806076478213072,
          "train_metric": 0.00047402424388565123,
          "validation_metric": 0.00038333411794155836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.311887701998785
        },
        {
          "epoch": 411,
          "updates": 411,
          "samples_seen": 420864,
          "online_loss": 0.0047065275721251965,
          "train_loss": 0.004682730417698622,
          "validation_loss": 0.003783476073294878,
          "train_metric": 0.00047162751434370875,
          "validation_metric": 0.0003810577909462154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.321271981998507
        },
        {
          "epoch": 412,
          "updates": 412,
          "samples_seen": 421888,
          "online_loss": 0.004682730883359909,
          "train_loss": 0.004659177735447884,
          "validation_loss": 0.003761101746931672,
          "train_metric": 0.0004692553193308413,
          "validation_metric": 0.0003788044268731028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.330408453999553
        },
        {
          "epoch": 413,
          "updates": 413,
          "samples_seen": 422912,
          "online_loss": 0.004659177735447884,
          "train_loss": 0.004635884892195463,
          "validation_loss": 0.0037389928475022316,
          "train_metric": 0.0004669093759730458,
          "validation_metric": 0.00037657766370102763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.339705068998228
        },
        {
          "epoch": 414,
          "updates": 414,
          "samples_seen": 423936,
          "online_loss": 0.0046358853578567505,
          "train_loss": 0.004612834192812443,
          "validation_loss": 0.003717134241014719,
          "train_metric": 0.00046458779252134264,
          "validation_metric": 0.0003743761626537889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.348721976999514
        },
        {
          "epoch": 415,
          "updates": 415,
          "samples_seen": 424960,
          "online_loss": 0.00461283465847373,
          "train_loss": 0.004590027034282684,
          "validation_loss": 0.0036954875104129314,
          "train_metric": 0.00046229068539105356,
          "validation_metric": 0.00037219602381810546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.35749445699912
        },
        {
          "epoch": 416,
          "updates": 416,
          "samples_seen": 425984,
          "online_loss": 0.004590027034282684,
          "train_loss": 0.004567462019622326,
          "validation_loss": 0.0036740938667207956,
          "train_metric": 0.0004600180545821786,
          "validation_metric": 0.0003700412926264107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.36656259099982
        },
        {
          "epoch": 417,
          "updates": 417,
          "samples_seen": 427008,
          "online_loss": 0.004567461553961039,
          "train_loss": 0.004545136820524931,
          "validation_loss": 0.0036529579665511847,
          "train_metric": 0.00045776955084875226,
          "validation_metric": 0.00036791255115531385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.375365337999028
        },
        {
          "epoch": 418,
          "updates": 418,
          "samples_seen": 428032,
          "online_loss": 0.004545136820524931,
          "train_loss": 0.00452304445207119,
          "validation_loss": 0.0036320295184850693,
          "train_metric": 0.00045554444659501314,
          "validation_metric": 0.0003658047062344849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.3846702329992695
        },
        {
          "epoch": 419,
          "updates": 419,
          "samples_seen": 429056,
          "online_loss": 0.004523043986409903,
          "train_loss": 0.004501184914261103,
          "validation_loss": 0.003611351130530238,
          "train_metric": 0.00045334294554777443,
          "validation_metric": 0.00036372203612700105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.393657597998754
        },
        {
          "epoch": 420,
          "updates": 420,
          "samples_seen": 430080,
          "online_loss": 0.004501185845583677,
          "train_loss": 0.00447955122217536,
          "validation_loss": 0.0035909288562834263,
          "train_metric": 0.00045116402907297015,
          "validation_metric": 0.0003616652684286237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.402483228999699
        },
        {
          "epoch": 421,
          "updates": 421,
          "samples_seen": 431104,
          "online_loss": 0.004479550756514072,
          "train_loss": 0.0044581471011042595,
          "validation_loss": 0.00357069936580956,
          "train_metric": 0.000449008250143379,
          "validation_metric": 0.0003596277383621782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.510419737998745
        },
        {
          "epoch": 422,
          "updates": 422,
          "samples_seen": 432128,
          "online_loss": 0.0044581471011042595,
          "train_loss": 0.004436961840838194,
          "validation_loss": 0.0035507301799952984,
          "train_metric": 0.00044687450281344354,
          "validation_metric": 0.0003576165472622961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.519443329998467
        },
        {
          "epoch": 423,
          "updates": 423,
          "samples_seen": 433152,
          "online_loss": 0.004436961840838194,
          "train_loss": 0.004415992181748152,
          "validation_loss": 0.0035309968516230583,
          "train_metric": 0.0004447625542525202,
          "validation_metric": 0.0003556290757842362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.528785642998628
        },
        {
          "epoch": 424,
          "updates": 424,
          "samples_seen": 434176,
          "online_loss": 0.004415992181748152,
          "train_loss": 0.004395244177430868,
          "validation_loss": 0.0035114616621285677,
          "train_metric": 0.00044267287012189627,
          "validation_metric": 0.00035366154043003917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.537831612999071
        },
        {
          "epoch": 425,
          "updates": 425,
          "samples_seen": 435200,
          "online_loss": 0.004395244177430868,
          "train_loss": 0.00437470618635416,
          "validation_loss": 0.003492178162559867,
          "train_metric": 0.00044060443178750575,
          "validation_metric": 0.0003517193836160004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.5469534499989095
        },
        {
          "epoch": 426,
          "updates": 426,
          "samples_seen": 436224,
          "online_loss": 0.00437470618635416,
          "train_loss": 0.004354373551905155,
          "validation_loss": 0.0034731137566268444,
          "train_metric": 0.0004385565407574177,
          "validation_metric": 0.0003497992584016174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.556069635998938
        },
        {
          "epoch": 427,
          "updates": 427,
          "samples_seen": 437248,
          "online_loss": 0.004354373551905155,
          "train_loss": 0.004334247205406427,
          "validation_loss": 0.0034542533103376627,
          "train_metric": 0.00043652948806993663,
          "validation_metric": 0.00034789973869919777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.565079130999948
        },
        {
          "epoch": 428,
          "updates": 428,
          "samples_seen": 438272,
          "online_loss": 0.00433424673974514,
          "train_loss": 0.004314321558922529,
          "validation_loss": 0.0034356294199824333,
          "train_metric": 0.0004345226916484535,
          "validation_metric": 0.00034602402593009174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.574186599998939
        },
        {
          "epoch": 429,
          "updates": 429,
          "samples_seen": 439296,
          "online_loss": 0.004314321558922529,
          "train_loss": 0.004294591955840588,
          "validation_loss": 0.0034172351006418467,
          "train_metric": 0.00043253551120869815,
          "validation_metric": 0.0003441713924985379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.583393955999782
        },
        {
          "epoch": 430,
          "updates": 430,
          "samples_seen": 440320,
          "online_loss": 0.004294591955840588,
          "train_loss": 0.004275057464838028,
          "validation_loss": 0.0033990275114774704,
          "train_metric": 0.0004305681213736534,
          "validation_metric": 0.00034233758924528956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.592447121998703
        },
        {
          "epoch": 431,
          "updates": 431,
          "samples_seen": 441344,
          "online_loss": 0.004275057930499315,
          "train_loss": 0.004255714826285839,
          "validation_loss": 0.0033810390159487724,
          "train_metric": 0.00042861999827437103,
          "validation_metric": 0.0003405258757993579,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.601478265998594
        },
        {
          "epoch": 432,
          "updates": 432,
          "samples_seen": 442368,
          "online_loss": 0.0042557138949632645,
          "train_loss": 0.0042365603148937225,
          "validation_loss": 0.0033632752019912004,
          "train_metric": 0.00042669085087254643,
          "validation_metric": 0.0003387367760296911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.610896149999462
        },
        {
          "epoch": 433,
          "updates": 433,
          "samples_seen": 443392,
          "online_loss": 0.004236559849232435,
          "train_loss": 0.004217589274048805,
          "validation_loss": 0.0033457018435001373,
          "train_metric": 0.00042478012619540095,
          "validation_metric": 0.0003369668556842953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.620007848998284
        },
        {
          "epoch": 434,
          "updates": 434,
          "samples_seen": 444416,
          "online_loss": 0.004217589274048805,
          "train_loss": 0.004198799841105938,
          "validation_loss": 0.003328331746160984,
          "train_metric": 0.0004228877369314432,
          "validation_metric": 0.00033521742443554103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.629446163999091
        },
        {
          "epoch": 435,
          "updates": 435,
          "samples_seen": 445440,
          "online_loss": 0.004198799841105938,
          "train_loss": 0.004180187359452248,
          "validation_loss": 0.0033111870288848877,
          "train_metric": 0.00042101313010789454,
          "validation_metric": 0.0003334906359668821,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.638333906999833
        },
        {
          "epoch": 436,
          "updates": 436,
          "samples_seen": 446464,
          "online_loss": 0.00418018689379096,
          "train_loss": 0.004161747172474861,
          "validation_loss": 0.0032942069228738546,
          "train_metric": 0.0004191559855826199,
          "validation_metric": 0.00033178040757775307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.647542523998709
        },
        {
          "epoch": 437,
          "updates": 437,
          "samples_seen": 447488,
          "online_loss": 0.0041617476381361485,
          "train_loss": 0.0041434853337705135,
          "validation_loss": 0.003277441719546914,
          "train_metric": 0.0004173166526015848,
          "validation_metric": 0.00033009194885380566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.657253535999189
        },
        {
          "epoch": 438,
          "updates": 438,
          "samples_seen": 448512,
          "online_loss": 0.0041434853337705135,
          "train_loss": 0.004125385545194149,
          "validation_loss": 0.0032608804758638144,
          "train_metric": 0.0004154937050770968,
          "validation_metric": 0.0003284239792264998,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.666426433999732
        },
        {
          "epoch": 439,
          "updates": 439,
          "samples_seen": 449536,
          "online_loss": 0.004125385079532862,
          "train_loss": 0.004107448738068342,
          "validation_loss": 0.0032444859389215708,
          "train_metric": 0.0004136872012168169,
          "validation_metric": 0.00032677274430170655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.676051332999123
        },
        {
          "epoch": 440,
          "updates": 440,
          "samples_seen": 450560,
          "online_loss": 0.004107448738068342,
          "train_loss": 0.004089679103344679,
          "validation_loss": 0.0032282781321555376,
          "train_metric": 0.00041189754847437143,
          "validation_metric": 0.00032514031045138836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.685473378998722
        },
        {
          "epoch": 441,
          "updates": 441,
          "samples_seen": 451584,
          "online_loss": 0.004089678637683392,
          "train_loss": 0.004072064068168402,
          "validation_loss": 0.0032122887205332518,
          "train_metric": 0.00041012337896972895,
          "validation_metric": 0.0003235299664083868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.695253429999866
        },
        {
          "epoch": 442,
          "updates": 442,
          "samples_seen": 452608,
          "online_loss": 0.004072064068168402,
          "train_loss": 0.004054605960845947,
          "validation_loss": 0.0031964413356035948,
          "train_metric": 0.0004083650710526854,
          "validation_metric": 0.0003219338832423091,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.704342290999193
        },
        {
          "epoch": 443,
          "updates": 443,
          "samples_seen": 453632,
          "online_loss": 0.004054605960845947,
          "train_loss": 0.00403730059042573,
          "validation_loss": 0.003180786268785596,
          "train_metric": 0.0004066221881657839,
          "validation_metric": 0.00032035718322731555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.713451437999538
        },
        {
          "epoch": 444,
          "updates": 444,
          "samples_seen": 454656,
          "online_loss": 0.004037301056087017,
          "train_loss": 0.004020145162940025,
          "validation_loss": 0.0031653279438614845,
          "train_metric": 0.0004048942937515676,
          "validation_metric": 0.0003188002738170326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.722700025999075
        },
        {
          "epoch": 445,
          "updates": 445,
          "samples_seen": 455680,
          "online_loss": 0.004020145628601313,
          "train_loss": 0.004003140144050121,
          "validation_loss": 0.003150003030896187,
          "train_metric": 0.0004031815624330193,
          "validation_metric": 0.00031725678127259016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.731897454999853
        },
        {
          "epoch": 446,
          "updates": 446,
          "samples_seen": 456704,
          "online_loss": 0.004003139678388834,
          "train_loss": 0.0039862776175141335,
          "validation_loss": 0.0031348825432360172,
          "train_metric": 0.0004014832666143775,
          "validation_metric": 0.0003157338942401111,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.741441674999805
        },
        {
          "epoch": 447,
          "updates": 447,
          "samples_seen": 457728,
          "online_loss": 0.003986277151852846,
          "train_loss": 0.003969553392380476,
          "validation_loss": 0.003119916655123234,
          "train_metric": 0.0003997988533228636,
          "validation_metric": 0.0003142265777569264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.751072168999599
        },
        {
          "epoch": 448,
          "updates": 448,
          "samples_seen": 458752,
          "online_loss": 0.003969553392380476,
          "train_loss": 0.00395297072827816,
          "validation_loss": 0.0031050967518240213,
          "train_metric": 0.0003981287300121039,
          "validation_metric": 0.00031273395870812237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.760717844999817
        },
        {
          "epoch": 449,
          "updates": 449,
          "samples_seen": 459776,
          "online_loss": 0.003952971193939447,
          "train_loss": 0.0039365277625620365,
          "validation_loss": 0.003090477781370282,
          "train_metric": 0.0003964726347476244,
          "validation_metric": 0.0003112615959253162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.770030085999679
        },
        {
          "epoch": 450,
          "updates": 450,
          "samples_seen": 460800,
          "online_loss": 0.003936527296900749,
          "train_loss": 0.003920213785022497,
          "validation_loss": 0.0030759943183511496,
          "train_metric": 0.0003948295779991895,
          "validation_metric": 0.00030980288283899426,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.779201646999354
        },
        {
          "epoch": 451,
          "updates": 451,
          "samples_seen": 461824,
          "online_loss": 0.0039202142506837845,
          "train_loss": 0.003904037643224001,
          "validation_loss": 0.0030616542790085077,
          "train_metric": 0.0003932004328817129,
          "validation_metric": 0.0003083585761487484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.788565554999877
        },
        {
          "epoch": 452,
          "updates": 452,
          "samples_seen": 462848,
          "online_loss": 0.0039040378760546446,
          "train_loss": 0.003887985832989216,
          "validation_loss": 0.003047493053600192,
          "train_metric": 0.0003915836859960109,
          "validation_metric": 0.00030693234293721616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.798235021999062
        },
        {
          "epoch": 453,
          "updates": 453,
          "samples_seen": 463872,
          "online_loss": 0.003887986298650503,
          "train_loss": 0.0038720620796084404,
          "validation_loss": 0.0030334671027958393,
          "train_metric": 0.0003899799194186926,
          "validation_metric": 0.0003055197012145072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.807967124999777
        },
        {
          "epoch": 454,
          "updates": 454,
          "samples_seen": 464896,
          "online_loss": 0.003872062312439084,
          "train_loss": 0.003856260096654296,
          "validation_loss": 0.003019578754901886,
          "train_metric": 0.00038838834734633565,
          "validation_metric": 0.00030412088381126523,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.81712285499998
        },
        {
          "epoch": 455,
          "updates": 455,
          "samples_seen": 465920,
          "online_loss": 0.0038562598638236523,
          "train_loss": 0.003840590361505747,
          "validation_loss": 0.0030058666598051786,
          "train_metric": 0.0003868102212436497,
          "validation_metric": 0.00030273987795226276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.826216520999878
        },
        {
          "epoch": 456,
          "updates": 456,
          "samples_seen": 466944,
          "online_loss": 0.003840590361505747,
          "train_loss": 0.0038250323850661516,
          "validation_loss": 0.002992267720401287,
          "train_metric": 0.0003852432419080287,
          "validation_metric": 0.00030137025169096887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.835237869998309
        },
        {
          "epoch": 457,
          "updates": 457,
          "samples_seen": 467968,
          "online_loss": 0.003825032152235508,
          "train_loss": 0.0038095954805612564,
          "validation_loss": 0.002978814300149679,
          "train_metric": 0.00038368851528503,
          "validation_metric": 0.0003000152937602252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.84503407199918
        },
        {
          "epoch": 458,
          "updates": 458,
          "samples_seen": 468992,
          "online_loss": 0.0038095952477306128,
          "train_loss": 0.003794270334765315,
          "validation_loss": 0.0029655180405825377,
          "train_metric": 0.0003821450227405876,
          "validation_metric": 0.00029867611010558903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.970430214998487
        },
        {
          "epoch": 459,
          "updates": 459,
          "samples_seen": 470016,
          "online_loss": 0.003794271033257246,
          "train_loss": 0.0037790657952427864,
          "validation_loss": 0.002952328883111477,
          "train_metric": 0.0003806136955972761,
          "validation_metric": 0.00029734778217971325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.982591702999343
        },
        {
          "epoch": 460,
          "updates": 460,
          "samples_seen": 471040,
          "online_loss": 0.00377906602807343,
          "train_loss": 0.0037639739457517862,
          "validation_loss": 0.002939293859526515,
          "train_metric": 0.00037909368984401226,
          "validation_metric": 0.00029603490838781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.9929896849989746
        },
        {
          "epoch": 461,
          "updates": 461,
          "samples_seen": 472064,
          "online_loss": 0.00376397417858243,
          "train_loss": 0.003748988965526223,
          "validation_loss": 0.0029263824690133333,
          "train_metric": 0.0003775844816118479,
          "validation_metric": 0.00029473454924300313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.003434802998527
        },
        {
          "epoch": 462,
          "updates": 462,
          "samples_seen": 473088,
          "online_loss": 0.0037489894311875105,
          "train_loss": 0.0037341155111789703,
          "validation_loss": 0.0029135914519429207,
          "train_metric": 0.0003760864201467484,
          "validation_metric": 0.0002934462681878358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.014925430999938
        },
        {
          "epoch": 463,
          "updates": 463,
          "samples_seen": 474112,
          "online_loss": 0.003734115045517683,
          "train_loss": 0.0037193442694842815,
          "validation_loss": 0.002900931518524885,
          "train_metric": 0.00037459871964529157,
          "validation_metric": 0.0002921711711678654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.024915483998484
        },
        {
          "epoch": 464,
          "updates": 464,
          "samples_seen": 475136,
          "online_loss": 0.003719344502314925,
          "train_loss": 0.0037046847864985466,
          "validation_loss": 0.002888391725718975,
          "train_metric": 0.0003731223114300519,
          "validation_metric": 0.00029090821044519544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.036217565999323
        },
        {
          "epoch": 465,
          "updates": 465,
          "samples_seen": 476160,
          "online_loss": 0.003704684553667903,
          "train_loss": 0.003690125420689583,
          "validation_loss": 0.0028759813867509365,
          "train_metric": 0.0003716559149324894,
          "validation_metric": 0.00028965831734240055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.0466123029982555
        },
        {
          "epoch": 466,
          "updates": 466,
          "samples_seen": 477184,
          "online_loss": 0.0036901256535202265,
          "train_loss": 0.0036756666377186775,
          "validation_loss": 0.002863675821572542,
          "train_metric": 0.00037019967567175627,
          "validation_metric": 0.000288418959826231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.057049133998589
        },
        {
          "epoch": 467,
          "updates": 467,
          "samples_seen": 478208,
          "online_loss": 0.0036756666377186775,
          "train_loss": 0.0036613140255212784,
          "validation_loss": 0.0028514943551272154,
          "train_metric": 0.00036875411751680076,
          "validation_metric": 0.00028719205874949694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.067260440999235
        },
        {
          "epoch": 468,
          "updates": 468,
          "samples_seen": 479232,
          "online_loss": 0.0036613137926906347,
          "train_loss": 0.003647052915766835,
          "validation_loss": 0.002839436987414956,
          "train_metric": 0.0003673178143799305,
          "validation_metric": 0.00028597767231985927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.0773958879999554
        },
        {
          "epoch": 469,
          "updates": 469,
          "samples_seen": 480256,
          "online_loss": 0.0036470533814281225,
          "train_loss": 0.0036328937858343124,
          "validation_loss": 0.002827457617968321,
          "train_metric": 0.00036589172668755054,
          "validation_metric": 0.00028477120213210583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.087629649999144
        },
        {
          "epoch": 470,
          "updates": 470,
          "samples_seen": 481280,
          "online_loss": 0.0036328937858343124,
          "train_loss": 0.003618831280618906,
          "validation_loss": 0.002815625397488475,
          "train_metric": 0.0003644754469860345,
          "validation_metric": 0.0002835794584825635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.098187259998667
        },
        {
          "epoch": 471,
          "updates": 471,
          "samples_seen": 482304,
          "online_loss": 0.0036188315134495497,
          "train_loss": 0.003604861441999674,
          "validation_loss": 0.0028038874734193087,
          "train_metric": 0.0003630684223026037,
          "validation_metric": 0.0002823972317855805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.108879050998439
        },
        {
          "epoch": 472,
          "updates": 472,
          "samples_seen": 483328,
          "online_loss": 0.00360486120916903,
          "train_loss": 0.0035909865982830524,
          "validation_loss": 0.002792240818962455,
          "train_metric": 0.0003616710309870541,
          "validation_metric": 0.00028122428921051323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.119221021999692
        },
        {
          "epoch": 473,
          "updates": 473,
          "samples_seen": 484352,
          "online_loss": 0.003590986831113696,
          "train_loss": 0.003577203955501318,
          "validation_loss": 0.0027807282749563456,
          "train_metric": 0.0003602828655857593,
          "validation_metric": 0.00028006473439745605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.128594913998313
        },
        {
          "epoch": 474,
          "updates": 474,
          "samples_seen": 485376,
          "online_loss": 0.003577203955501318,
          "train_loss": 0.003563513280823827,
          "validation_loss": 0.002769287209957838,
          "train_metric": 0.00035890404251404107,
          "validation_metric": 0.0002789124264381826,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.137878432999059
        },
        {
          "epoch": 475,
          "updates": 475,
          "samples_seen": 486400,
          "online_loss": 0.0035635135136544704,
          "train_loss": 0.003549908986315131,
          "validation_loss": 0.002757968846708536,
          "train_metric": 0.0003575338050723076,
          "validation_metric": 0.0002777725167106837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.148595928998475
        },
        {
          "epoch": 476,
          "updates": 476,
          "samples_seen": 487424,
          "online_loss": 0.0035499087534844875,
          "train_loss": 0.0035363929346203804,
          "validation_loss": 0.0027467410545796156,
          "train_metric": 0.00035617256071418524,
          "validation_metric": 0.0002766416873782873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.1585486579988356
        },
        {
          "epoch": 477,
          "updates": 477,
          "samples_seen": 488448,
          "online_loss": 0.0035363929346203804,
          "train_loss": 0.0035229651257395744,
          "validation_loss": 0.0027356024365872145,
          "train_metric": 0.0003548201057128608,
          "validation_metric": 0.00027551985112950206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.167838574998314
        },
        {
          "epoch": 478,
          "updates": 478,
          "samples_seen": 489472,
          "online_loss": 0.0035229648929089308,
          "train_loss": 0.003509619738906622,
          "validation_loss": 0.002724571619182825,
          "train_metric": 0.00035347609082236886,
          "validation_metric": 0.0002744088415056467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.177377433999936
        },
        {
          "epoch": 479,
          "updates": 479,
          "samples_seen": 490496,
          "online_loss": 0.0035096202045679092,
          "train_loss": 0.0034963686484843493,
          "validation_loss": 0.0027136332355439663,
          "train_metric": 0.0003521414764691144,
          "validation_metric": 0.00027330726152285933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.1865542189989355
        },
        {
          "epoch": 480,
          "updates": 480,
          "samples_seen": 491520,
          "online_loss": 0.003496368881314993,
          "train_loss": 0.003483191365376115,
          "validation_loss": 0.0027027796022593975,
          "train_metric": 0.0003508143126964569,
          "validation_metric": 0.0002722140634432435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.196538252999744
        },
        {
          "epoch": 481,
          "updates": 481,
          "samples_seen": 492544,
          "online_loss": 0.003483191365376115,
          "train_loss": 0.003470100462436676,
          "validation_loss": 0.0026920209638774395,
          "train_metric": 0.0003494958218652755,
          "validation_metric": 0.00027113049873150885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.20646413599934
        },
        {
          "epoch": 482,
          "updates": 482,
          "samples_seen": 493568,
          "online_loss": 0.003470100462436676,
          "train_loss": 0.0034570919815450907,
          "validation_loss": 0.0026813519652932882,
          "train_metric": 0.00034818562562577426,
          "validation_metric": 0.0002700559562072158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.218185500998516
        },
        {
          "epoch": 483,
          "updates": 483,
          "samples_seen": 494592,
          "online_loss": 0.003457091748714447,
          "train_loss": 0.0034441628959029913,
          "validation_loss": 0.0026707614306360483,
          "train_metric": 0.00034688349114730954,
          "validation_metric": 0.0002689893008209765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.230163790000006
        },
        {
          "epoch": 484,
          "updates": 484,
          "samples_seen": 495616,
          "online_loss": 0.003444163128733635,
          "train_loss": 0.0034313183277845383,
          "validation_loss": 0.002660274039953947,
          "train_metric": 0.00034558982588350773,
          "validation_metric": 0.0002679330646060407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.239759845999288
        },
        {
          "epoch": 485,
          "updates": 485,
          "samples_seen": 496640,
          "online_loss": 0.0034313183277845383,
          "train_loss": 0.003418548032641411,
          "validation_loss": 0.002649854402989149,
          "train_metric": 0.00034430361120030284,
          "validation_metric": 0.0002668836386874318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.249913156998446
        },
        {
          "epoch": 486,
          "updates": 486,
          "samples_seen": 497664,
          "online_loss": 0.003418548032641411,
          "train_loss": 0.003405854105949402,
          "validation_loss": 0.0026395292952656746,
          "train_metric": 0.0003430251672398299,
          "validation_metric": 0.0002658437006175518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.260092631999214
        },
        {
          "epoch": 487,
          "updates": 487,
          "samples_seen": 498688,
          "online_loss": 0.003405854571610689,
          "train_loss": 0.0033932372462004423,
          "validation_loss": 0.0026292784605175257,
          "train_metric": 0.0003417543775867671,
          "validation_metric": 0.0002648113004397601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.27071883499957
        },
        {
          "epoch": 488,
          "updates": 488,
          "samples_seen": 499712,
          "online_loss": 0.0033932370133697987,
          "train_loss": 0.0033806967549026012,
          "validation_loss": 0.0026191086508333683,
          "train_metric": 0.00034049138776026666,
          "validation_metric": 0.00026378704933449626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.280872132998411
        },
        {
          "epoch": 489,
          "updates": 489,
          "samples_seen": 500736,
          "online_loss": 0.0033806967549026012,
          "train_loss": 0.003368232399225235,
          "validation_loss": 0.002609018702059984,
          "train_metric": 0.00033923605224117637,
          "validation_metric": 0.0002627707726787776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.290511118999348
        },
        {
          "epoch": 490,
          "updates": 490,
          "samples_seen": 501760,
          "online_loss": 0.003368233097717166,
          "train_loss": 0.0033558381255716085,
          "validation_loss": 0.0025990093126893044,
          "train_metric": 0.0003379877598490566,
          "validation_metric": 0.0002617627033032477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.29946541299978
        },
        {
          "epoch": 491,
          "updates": 491,
          "samples_seen": 502784,
          "online_loss": 0.003355838358402252,
          "train_loss": 0.0033435251098126173,
          "validation_loss": 0.002589072100818157,
          "train_metric": 0.00033674758742563426,
          "validation_metric": 0.0002607618225738406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.308942774998286
        },
        {
          "epoch": 492,
          "updates": 492,
          "samples_seen": 503808,
          "online_loss": 0.0033435251098126173,
          "train_loss": 0.00333127798512578,
          "validation_loss": 0.0025792191736400127,
          "train_metric": 0.000335514108883217,
          "validation_metric": 0.0002597694983705878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.318305110999063
        },
        {
          "epoch": 493,
          "updates": 493,
          "samples_seen": 504832,
          "online_loss": 0.00333127798512578,
          "train_loss": 0.0033191062975674868,
          "validation_loss": 0.0025694211944937706,
          "train_metric": 0.00033428819733671844,
          "validation_metric": 0.0002587827038951218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.3284372679991066
        },
        {
          "epoch": 494,
          "updates": 494,
          "samples_seen": 505856,
          "online_loss": 0.0033191062975674868,
          "train_loss": 0.0033069979399442673,
          "validation_loss": 0.0025597205385565758,
          "train_metric": 0.0003330687468405813,
          "validation_metric": 0.0002578056592028588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.337952788999246
        },
        {
          "epoch": 495,
          "updates": 495,
          "samples_seen": 506880,
          "online_loss": 0.0033069979399442673,
          "train_loss": 0.0032949713058769703,
          "validation_loss": 0.0025500748306512833,
          "train_metric": 0.0003318574163131416,
          "validation_metric": 0.00025683417334221303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.347041514998637
        },
        {
          "epoch": 496,
          "updates": 496,
          "samples_seen": 507904,
          "online_loss": 0.0032949713058769703,
          "train_loss": 0.0032830045092850924,
          "validation_loss": 0.0025405040942132473,
          "train_metric": 0.0003306521975900978,
          "validation_metric": 0.000255870254477486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.35636539099869
        },
        {
          "epoch": 497,
          "updates": 497,
          "samples_seen": 508928,
          "online_loss": 0.0032830045092850924,
          "train_loss": 0.0032711122184991837,
          "validation_loss": 0.0025310011114925146,
          "train_metric": 0.00032945445855148137,
          "validation_metric": 0.0002549131750129163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.366009860999839
        },
        {
          "epoch": 498,
          "updates": 498,
          "samples_seen": 509952,
          "online_loss": 0.0032711122184991837,
          "train_loss": 0.003259285120293498,
          "validation_loss": 0.002521575428545475,
          "train_metric": 0.0003282632678747177,
          "validation_metric": 0.00025396383716724813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.47822868899857
        },
        {
          "epoch": 499,
          "updates": 499,
          "samples_seen": 510976,
          "online_loss": 0.0032592853531241417,
          "train_loss": 0.003247522981837392,
          "validation_loss": 0.002512202365323901,
          "train_metric": 0.0003270786546636373,
          "validation_metric": 0.00025301979621872306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.487425677998544
        },
        {
          "epoch": 500,
          "updates": 500,
          "samples_seen": 512000,
          "online_loss": 0.003247522981837392,
          "train_loss": 0.003235829994082451,
          "validation_loss": 0.0025029112584888935,
          "train_metric": 0.0003259009390603751,
          "validation_metric": 0.00025208404986187816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.49685875599971
        },
        {
          "epoch": 501,
          "updates": 501,
          "samples_seen": 513024,
          "online_loss": 0.003235829994082451,
          "train_loss": 0.003224201500415802,
          "validation_loss": 0.00249367393553257,
          "train_metric": 0.00032472971361130476,
          "validation_metric": 0.0002511537168174982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.506401740000001
        },
        {
          "epoch": 502,
          "updates": 502,
          "samples_seen": 514048,
          "online_loss": 0.003224200801923871,
          "train_loss": 0.00321263843216002,
          "validation_loss": 0.002484516240656376,
          "train_metric": 0.0003235651820432395,
          "validation_metric": 0.0002502313582226634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.517561297998327
        },
        {
          "epoch": 503,
          "updates": 503,
          "samples_seen": 515072,
          "online_loss": 0.00321263843216002,
          "train_loss": 0.0032011379953473806,
          "validation_loss": 0.0024753951001912355,
          "train_metric": 0.0003224069077987224,
          "validation_metric": 0.0002493126958142966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.5300376399991364
        },
        {
          "epoch": 504,
          "updates": 504,
          "samples_seen": 516096,
          "online_loss": 0.003201137762516737,
          "train_loss": 0.0031897027511149645,
          "validation_loss": 0.002466369653120637,
          "train_metric": 0.00032125518191605806,
          "validation_metric": 0.0002484037249814719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.54511001299943
        },
        {
          "epoch": 505,
          "updates": 505,
          "samples_seen": 517120,
          "online_loss": 0.003189702983945608,
          "train_loss": 0.0031783373560756445,
          "validation_loss": 0.002457373309880495,
          "train_metric": 0.0003201105282641947,
          "validation_metric": 0.0002474976354278624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.558396163998623
        },
        {
          "epoch": 506,
          "updates": 506,
          "samples_seen": 518144,
          "online_loss": 0.003178337588906288,
          "train_loss": 0.0031670264434069395,
          "validation_loss": 0.002448471263051033,
          "train_metric": 0.00031897128792479634,
          "validation_metric": 0.00024660106282681227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.574997434998295
        },
        {
          "epoch": 507,
          "updates": 507,
          "samples_seen": 519168,
          "online_loss": 0.003167026210576296,
          "train_loss": 0.0031557781621813774,
          "validation_loss": 0.002439600182697177,
          "train_metric": 0.0003178383922204375,
          "validation_metric": 0.00024570757523179054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.584407092999754
        },
        {
          "epoch": 508,
          "updates": 508,
          "samples_seen": 520192,
          "online_loss": 0.0031557781621813774,
          "train_loss": 0.0031445955391973257,
          "validation_loss": 0.002430802211165428,
          "train_metric": 0.00031671213218942285,
          "validation_metric": 0.0002448215091135353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.597014234999733
        },
        {
          "epoch": 509,
          "updates": 509,
          "samples_seen": 521216,
          "online_loss": 0.003144595306366682,
          "train_loss": 0.003133467398583889,
          "validation_loss": 0.0024220645427703857,
          "train_metric": 0.0003155913727823645,
          "validation_metric": 0.00024394146748818457,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.608268449999741
        },
        {
          "epoch": 510,
          "updates": 510,
          "samples_seen": 522240,
          "online_loss": 0.003133467398583889,
          "train_loss": 0.0031223981641232967,
          "validation_loss": 0.00241336552426219,
          "train_metric": 0.00031447652145288885,
          "validation_metric": 0.00024306531122419983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.6299185269999725
        },
        {
          "epoch": 511,
          "updates": 511,
          "samples_seen": 523264,
          "online_loss": 0.0031223981641232967,
          "train_loss": 0.003111397149041295,
          "validation_loss": 0.0024047528859227896,
          "train_metric": 0.00031336850952357054,
          "validation_metric": 0.0002421978861093521,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.6481856429982145
        },
        {
          "epoch": 512,
          "updates": 512,
          "samples_seen": 524288,
          "online_loss": 0.003111397149041295,
          "train_loss": 0.0031004499178379774,
          "validation_loss": 0.0023961642291396856,
          "train_metric": 0.00031226599821820855,
          "validation_metric": 0.00024133289116434753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.6656617489989
        },
        {
          "epoch": 513,
          "updates": 513,
          "samples_seen": 525312,
          "online_loss": 0.0031004499178379774,
          "train_loss": 0.0030895627569407225,
          "validation_loss": 0.0023876747582107782,
          "train_metric": 0.00031116942409425974,
          "validation_metric": 0.00024047784972935915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.68447372799892
        },
        {
          "epoch": 514,
          "updates": 514,
          "samples_seen": 526336,
          "online_loss": 0.003089562989771366,
          "train_loss": 0.0030787279829382896,
          "validation_loss": 0.002379179000854492,
          "train_metric": 0.00031007814686745405,
          "validation_metric": 0.0002396222116658464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.6932709839984454
        },
        {
          "epoch": 515,
          "updates": 515,
          "samples_seen": 527360,
          "online_loss": 0.0030787275172770023,
          "train_loss": 0.0030679586343467236,
          "validation_loss": 0.002370800357311964,
          "train_metric": 0.00030899353441782296,
          "validation_metric": 0.00023877836065366864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.701818644998639
        },
        {
          "epoch": 516,
          "updates": 516,
          "samples_seen": 528384,
          "online_loss": 0.0030679581686854362,
          "train_loss": 0.003057240042835474,
          "validation_loss": 0.0023624140303581953,
          "train_metric": 0.0003079140151385218,
          "validation_metric": 0.00023793372383806854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.710589203999916
        },
        {
          "epoch": 517,
          "updates": 517,
          "samples_seen": 529408,
          "online_loss": 0.003057240042835474,
          "train_loss": 0.0030465773306787014,
          "validation_loss": 0.0023541415575891733,
          "train_metric": 0.00030684014200232923,
          "validation_metric": 0.0002371004957240075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.719679444999201
        },
        {
          "epoch": 518,
          "updates": 518,
          "samples_seen": 530432,
          "online_loss": 0.0030465777963399887,
          "train_loss": 0.0030359758529812098,
          "validation_loss": 0.0023458520881831646,
          "train_metric": 0.0003057723806705326,
          "validation_metric": 0.00023626559413969517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.7289016589984385
        },
        {
          "epoch": 519,
          "updates": 519,
          "samples_seen": 531456,
          "online_loss": 0.0030359758529812098,
          "train_loss": 0.0030254225712269545,
          "validation_loss": 0.0023376939352601767,
          "train_metric": 0.0003047094796784222,
          "validation_metric": 0.00023544397845398635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.738024753998616
        },
        {
          "epoch": 520,
          "updates": 520,
          "samples_seen": 532480,
          "online_loss": 0.0030254225712269545,
          "train_loss": 0.0030149284284561872,
          "validation_loss": 0.0023294882848858833,
          "train_metric": 0.00030365254497155547,
          "validation_metric": 0.00023461751698050648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.747109959998852
        },
        {
          "epoch": 521,
          "updates": 521,
          "samples_seen": 533504,
          "online_loss": 0.0030149281956255436,
          "train_loss": 0.003004488768056035,
          "validation_loss": 0.0023214404936879873,
          "train_metric": 0.00030260105268098414,
          "validation_metric": 0.00023380697530228645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.755966451999484
        },
        {
          "epoch": 522,
          "updates": 522,
          "samples_seen": 534528,
          "online_loss": 0.003004488768056035,
          "train_loss": 0.0029941026587039232,
          "validation_loss": 0.002313335658982396,
          "train_metric": 0.0003015550901181996,
          "validation_metric": 0.00023299071472138166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.764618162998886
        },
        {
          "epoch": 523,
          "updates": 523,
          "samples_seen": 535552,
          "online_loss": 0.002994102891534567,
          "train_loss": 0.0029837677720934153,
          "validation_loss": 0.0023053798358887434,
          "train_metric": 0.00030051416251808405,
          "validation_metric": 0.00023218941350933164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.773585153998283
        },
        {
          "epoch": 524,
          "updates": 524,
          "samples_seen": 536576,
          "online_loss": 0.0029837675392627716,
          "train_loss": 0.0029734878335148096,
          "validation_loss": 0.002297380706295371,
          "train_metric": 0.00029947885195724666,
          "validation_metric": 0.00023138376127462834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.782199175999267
        },
        {
          "epoch": 525,
          "updates": 525,
          "samples_seen": 537600,
          "online_loss": 0.0029734885320067406,
          "train_loss": 0.0029632544610649347,
          "validation_loss": 0.0022894940339028835,
          "train_metric": 0.00029844813980162144,
          "validation_metric": 0.00023058945953380316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.791216767998776
        },
        {
          "epoch": 526,
          "updates": 526,
          "samples_seen": 538624,
          "online_loss": 0.0029632544610649347,
          "train_loss": 0.00295308162458241,
          "validation_loss": 0.0022816213313490152,
          "train_metric": 0.0002974235685542226,
          "validation_metric": 0.0002297965547768399,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.799961059999987
        },
        {
          "epoch": 527,
          "updates": 527,
          "samples_seen": 539648,
          "online_loss": 0.002953081391751766,
          "train_loss": 0.002942955121397972,
          "validation_loss": 0.00227378960698843,
          "train_metric": 0.0002964035957120359,
          "validation_metric": 0.0002290077682118863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.808986247999201
        },
        {
          "epoch": 528,
          "updates": 528,
          "samples_seen": 540672,
          "online_loss": 0.002942954655736685,
          "train_loss": 0.0029328810051083565,
          "validation_loss": 0.0022660570684820414,
          "train_metric": 0.0002953890652861446,
          "validation_metric": 0.0002282289497088641,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.817546448999565
        },
        {
          "epoch": 529,
          "updates": 529,
          "samples_seen": 541696,
          "online_loss": 0.002932881470769644,
          "train_loss": 0.002922859275713563,
          "validation_loss": 0.0022582635283470154,
          "train_metric": 0.00029437971534207463,
          "validation_metric": 0.00022744404850527644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.826615048999884
        },
        {
          "epoch": 530,
          "updates": 530,
          "samples_seen": 542720,
          "online_loss": 0.002922859275713563,
          "train_loss": 0.002912885509431362,
          "validation_loss": 0.002250666730105877,
          "train_metric": 0.00029337513842619956,
          "validation_metric": 0.0002266789088025689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.835275453999202
        },
        {
          "epoch": 531,
          "updates": 531,
          "samples_seen": 543744,
          "online_loss": 0.002912885509431362,
          "train_loss": 0.002902960404753685,
          "validation_loss": 0.0022429144009947777,
          "train_metric": 0.0002923755382653326,
          "validation_metric": 0.0002258981257909909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.8438127169993095
        },
        {
          "epoch": 532,
          "updates": 532,
          "samples_seen": 544768,
          "online_loss": 0.002902960404753685,
          "train_loss": 0.002893091645091772,
          "validation_loss": 0.0022354493848979473,
          "train_metric": 0.0002913815842475742,
          "validation_metric": 0.00022514625743497163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.852813249999599
        },
        {
          "epoch": 533,
          "updates": 533,
          "samples_seen": 545792,
          "online_loss": 0.0028930911794304848,
          "train_loss": 0.002883265493437648,
          "validation_loss": 0.0022277403622865677,
          "train_metric": 0.00029039193759672344,
          "validation_metric": 0.00022436986910179257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.8615997139986575
        },
        {
          "epoch": 534,
          "updates": 534,
          "samples_seen": 546816,
          "online_loss": 0.0028832657262682915,
          "train_loss": 0.00287348753772676,
          "validation_loss": 0.002220406197011471,
          "train_metric": 0.0002894071221817285,
          "validation_metric": 0.000223631170229055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.8703715709998505
        },
        {
          "epoch": 535,
          "updates": 535,
          "samples_seen": 547840,
          "online_loss": 0.0028734873048961163,
          "train_loss": 0.0028637582436203957,
          "validation_loss": 0.002212714869529009,
          "train_metric": 0.00028842719621025026,
          "validation_metric": 0.00022285652812570333,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.976509166999676
        },
        {
          "epoch": 536,
          "updates": 536,
          "samples_seen": 548864,
          "online_loss": 0.0028637577779591084,
          "train_loss": 0.002854077611118555,
          "validation_loss": 0.002205550903454423,
          "train_metric": 0.0002874522178899497,
          "validation_metric": 0.00022213501506485045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.985080566999386
        },
        {
          "epoch": 537,
          "updates": 537,
          "samples_seen": 549888,
          "online_loss": 0.0028540773782879114,
          "train_loss": 0.0028444470372051,
          "validation_loss": 0.002197829307988286,
          "train_metric": 0.0002864822745323181,
          "validation_metric": 0.00022135733161121607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.993796460999874
        },
        {
          "epoch": 538,
          "updates": 538,
          "samples_seen": 550912,
          "online_loss": 0.0028444468043744564,
          "train_loss": 0.002834858139976859,
          "validation_loss": 0.0021909046918153763,
          "train_metric": 0.00028551658033393323,
          "validation_metric": 0.0002206598874181509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.002398508999249
        },
        {
          "epoch": 539,
          "updates": 539,
          "samples_seen": 551936,
          "online_loss": 0.00283485883846879,
          "train_loss": 0.0028253199998289347,
          "validation_loss": 0.0021830296609550714,
          "train_metric": 0.00028455586289055645,
          "validation_metric": 0.00021986676438245922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.011499279999043
        },
        {
          "epoch": 540,
          "updates": 540,
          "samples_seen": 552960,
          "online_loss": 0.0028253199998289347,
          "train_loss": 0.0028158226050436497,
          "validation_loss": 0.002176511101424694,
          "train_metric": 0.0002835993655025959,
          "validation_metric": 0.00021921024017501622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.020642461999159
        },
        {
          "epoch": 541,
          "updates": 541,
          "samples_seen": 553984,
          "online_loss": 0.0028158226050436497,
          "train_loss": 0.002806377364322543,
          "validation_loss": 0.0021682654041796923,
          "train_metric": 0.00028264804859645665,
          "validation_metric": 0.00021837977692484856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.029782859999614
        },
        {
          "epoch": 542,
          "updates": 542,
          "samples_seen": 555008,
          "online_loss": 0.002806377597153187,
          "train_loss": 0.002796980319544673,
          "validation_loss": 0.002162456512451172,
          "train_metric": 0.0002817016211338341,
          "validation_metric": 0.00021779474627692252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.0391281159991195
        },
        {
          "epoch": 543,
          "updates": 543,
          "samples_seen": 556032,
          "online_loss": 0.002796980319544673,
          "train_loss": 0.0027876223903149366,
          "validation_loss": 0.002153393579646945,
          "train_metric": 0.0002807590935844928,
          "validation_metric": 0.00021688193373847753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.048336942998503
        },
        {
          "epoch": 544,
          "updates": 544,
          "samples_seen": 557056,
          "online_loss": 0.0027876223903149366,
          "train_loss": 0.0027783168479800224,
          "validation_loss": 0.0021489430218935013,
          "train_metric": 0.0002798219211399555,
          "validation_metric": 0.0002164336619898677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.0575130979996175
        },
        {
          "epoch": 545,
          "updates": 545,
          "samples_seen": 558080,
          "online_loss": 0.0027783168479800224,
          "train_loss": 0.0027690583374351263,
          "validation_loss": 0.0021381124388426542,
          "train_metric": 0.0002788894053082913,
          "validation_metric": 0.0002153428504243493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.066377035998812
        },
        {
          "epoch": 546,
          "updates": 546,
          "samples_seen": 559104,
          "online_loss": 0.0027690583374351263,
          "train_loss": 0.0027598580345511436,
          "validation_loss": 0.002136436989530921,
          "train_metric": 0.0002779627393465489,
          "validation_metric": 0.0002151741209672764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.075434746999235
        },
        {
          "epoch": 547,
          "updates": 547,
          "samples_seen": 560128,
          "online_loss": 0.0027598578017205,
          "train_loss": 0.0027507178019732237,
          "validation_loss": 0.002121738623827696,
          "train_metric": 0.00027704224339686334,
          "validation_metric": 0.00021369376918300986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.084365884998988
        },
        {
          "epoch": 548,
          "updates": 548,
          "samples_seen": 561152,
          "online_loss": 0.0027507178019732237,
          "train_loss": 0.002741673029959202,
          "validation_loss": 0.002126123057678342,
          "train_metric": 0.00027613123529590666,
          "validation_metric": 0.0002141353179467842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.092408024998804
        },
        {
          "epoch": 549,
          "updates": 549,
          "samples_seen": 562176,
          "online_loss": 0.002741672797128558,
          "train_loss": 0.0027327851857990026,
          "validation_loss": 0.002102671191096306,
          "train_metric": 0.00027523611788637936,
          "validation_metric": 0.00021177338203415275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.101275027998781
        },
        {
          "epoch": 550,
          "updates": 550,
          "samples_seen": 563200,
          "online_loss": 0.0027327854186296463,
          "train_loss": 0.002724205609411001,
          "validation_loss": 0.0021210634149610996,
          "train_metric": 0.0002743719960562885,
          "validation_metric": 0.00021362573897931725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.109771115998228
        },
        {
          "epoch": 551,
          "updates": 551,
          "samples_seen": 564224,
          "online_loss": 0.002724205609411001,
          "train_loss": 0.002716305200010538,
          "validation_loss": 0.0020774556323885918,
          "train_metric": 0.00027357632643543184,
          "validation_metric": 0.00020923373813275248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.1188379369996255
        },
        {
          "epoch": 552,
          "updates": 552,
          "samples_seen": 565248,
          "online_loss": 0.0027163049671798944,
          "train_loss": 0.002709923079237342,
          "validation_loss": 0.0021299379877746105,
          "train_metric": 0.0002729335392359644,
          "validation_metric": 0.00021451957582030445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.127134466998541
        },
        {
          "epoch": 553,
          "updates": 553,
          "samples_seen": 566272,
          "online_loss": 0.0027099228464066982,
          "train_loss": 0.0027071996591985226,
          "validation_loss": 0.0020404295064508915,
          "train_metric": 0.00027265920653007925,
          "validation_metric": 0.00020550457702483982,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.136028723998606
        },
        {
          "epoch": 554,
          "updates": 554,
          "samples_seen": 567296,
          "online_loss": 0.0027071996591985226,
          "train_loss": 0.002712137997150421,
          "validation_loss": 0.0021782005205750465,
          "train_metric": 0.0002731565909925848,
          "validation_metric": 0.00021938039571978152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.1441092809982365
        },
        {
          "epoch": 555,
          "updates": 555,
          "samples_seen": 568320,
          "online_loss": 0.002712137997150421,
          "train_loss": 0.0027337153442204,
          "validation_loss": 0.0019953646697103977,
          "train_metric": 0.00027532983222045004,
          "validation_metric": 0.00020096582011319697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.153012252998451
        },
        {
          "epoch": 556,
          "updates": 556,
          "samples_seen": 569344,
          "online_loss": 0.0027337155770510435,
          "train_loss": 0.0027714737225323915,
          "validation_loss": 0.002308938652276993,
          "train_metric": 0.00027913274243474007,
          "validation_metric": 0.0002325478126294911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.161358264998853
        },
        {
          "epoch": 557,
          "updates": 557,
          "samples_seen": 570368,
          "online_loss": 0.002771473955363035,
          "train_loss": 0.002818348351866007,
          "validation_loss": 0.001979629509150982,
          "train_metric": 0.0002838537620846182,
          "validation_metric": 0.00019938104378525168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.1697244649985805
        },
        {
          "epoch": 558,
          "updates": 558,
          "samples_seen": 571392,
          "online_loss": 0.0028183485846966505,
          "train_loss": 0.0027872645296156406,
          "validation_loss": 0.0023447731509804726,
          "train_metric": 0.0002807230339385569,
          "validation_metric": 0.00023615699319634587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.177858173999994
        },
        {
          "epoch": 559,
          "updates": 559,
          "samples_seen": 572416,
          "online_loss": 0.002787264296784997,
          "train_loss": 0.0027034010272473097,
          "validation_loss": 0.001968086464330554,
          "train_metric": 0.0002722766366787255,
          "validation_metric": 0.00019821844762191176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.186564113999339
        },
        {
          "epoch": 560,
          "updates": 560,
          "samples_seen": 573440,
          "online_loss": 0.002703400794416666,
          "train_loss": 0.002638058038428426,
          "validation_loss": 0.002054137410596013,
          "train_metric": 0.00026569553301669657,
          "validation_metric": 0.00020688520453404635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.194870243998594
        },
        {
          "epoch": 561,
          "updates": 561,
          "samples_seen": 574464,
          "online_loss": 0.002638057805597782,
          "train_loss": 0.002662037266418338,
          "validation_loss": 0.002153353299945593,
          "train_metric": 0.0002681105979718268,
          "validation_metric": 0.00021687783009838313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.202837220998845
        },
        {
          "epoch": 562,
          "updates": 562,
          "samples_seen": 575488,
          "online_loss": 0.002662037033587694,
          "train_loss": 0.0027077305130660534,
          "validation_loss": 0.001947509590536356,
          "train_metric": 0.0002727127284742892,
          "validation_metric": 0.00019614602206274867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.211785980998684
        },
        {
          "epoch": 563,
          "updates": 563,
          "samples_seen": 576512,
          "online_loss": 0.0027077302802354097,
          "train_loss": 0.002675278577953577,
          "validation_loss": 0.002196977147832513,
          "train_metric": 0.0002694442227948457,
          "validation_metric": 0.00022127150441519916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.219790428998749
        },
        {
          "epoch": 564,
          "updates": 564,
          "samples_seen": 577536,
          "online_loss": 0.0026752783451229334,
          "train_loss": 0.0026147079188376665,
          "validation_loss": 0.001974873710423708,
          "train_metric": 0.00026334376889280975,
          "validation_metric": 0.00019890205294359475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.228144972999871
        },
        {
          "epoch": 565,
          "updates": 565,
          "samples_seen": 578560,
          "online_loss": 0.002614707686007023,
          "train_loss": 0.002606772817671299,
          "validation_loss": 0.001970214769244194,
          "train_metric": 0.00026254463591612875,
          "validation_metric": 0.0001984327973332256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.235741997999867
        },
        {
          "epoch": 566,
          "updates": 566,
          "samples_seen": 579584,
          "online_loss": 0.002606772817671299,
          "train_loss": 0.00263627409003675,
          "validation_loss": 0.0021508829668164253,
          "train_metric": 0.00026551587507128716,
          "validation_metric": 0.00021662905055563897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.243734548999782
        },
        {
          "epoch": 567,
          "updates": 567,
          "samples_seen": 580608,
          "online_loss": 0.002636273857206106,
          "train_loss": 0.0026305902283638716,
          "validation_loss": 0.0019262029090896249,
          "train_metric": 0.00026494337362237275,
          "validation_metric": 0.0001940000947797671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.252049900998827
        },
        {
          "epoch": 568,
          "updates": 568,
          "samples_seen": 581632,
          "online_loss": 0.0026305902283638716,
          "train_loss": 0.0025848757941275835,
          "validation_loss": 0.002050178125500679,
          "train_metric": 0.00026033923495560884,
          "validation_metric": 0.0002064864238491282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.259981654999137
        },
        {
          "epoch": 569,
          "updates": 569,
          "samples_seen": 582656,
          "online_loss": 0.0025848757941275835,
          "train_loss": 0.0025716947857290506,
          "validation_loss": 0.0020218570716679096,
          "train_metric": 0.0002590116928331554,
          "validation_metric": 0.00020363401563372463,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.26752964499974
        },
        {
          "epoch": 570,
          "updates": 570,
          "samples_seen": 583680,
          "online_loss": 0.002571694552898407,
          "train_loss": 0.0025905058719217777,
          "validation_loss": 0.0019172246102243662,
          "train_metric": 0.00026090629398822784,
          "validation_metric": 0.00019309586787130684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.276089632998264
        },
        {
          "epoch": 571,
          "updates": 571,
          "samples_seen": 584704,
          "online_loss": 0.002590506337583065,
          "train_loss": 0.0025830292142927647,
          "validation_loss": 0.002089118119329214,
          "train_metric": 0.0002601532614789903,
          "validation_metric": 0.0002104082959704101,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.283830818998467
        },
        {
          "epoch": 572,
          "updates": 572,
          "samples_seen": 585728,
          "online_loss": 0.0025830287486314774,
          "train_loss": 0.0025520005729049444,
          "validation_loss": 0.0019350918009877205,
          "train_metric": 0.0002570281212683767,
          "validation_metric": 0.00019489535770844668,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.291698103999806
        },
        {
          "epoch": 573,
          "updates": 573,
          "samples_seen": 586752,
          "online_loss": 0.0025520005729049444,
          "train_loss": 0.0025401676539331675,
          "validation_loss": 0.001942844595760107,
          "train_metric": 0.0002558364358264953,
          "validation_metric": 0.0001956761843757704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.29924697199931
        },
        {
          "epoch": 574,
          "updates": 574,
          "samples_seen": 587776,
          "online_loss": 0.002540167886763811,
          "train_loss": 0.002549593336880207,
          "validation_loss": 0.002044012537226081,
          "train_metric": 0.0002567857154645026,
          "validation_metric": 0.0002058654499705881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.4042407009983435
        },
        {
          "epoch": 575,
          "updates": 575,
          "samples_seen": 588800,
          "online_loss": 0.002549593336880207,
          "train_loss": 0.0025444829370826483,
          "validation_loss": 0.0018975564744323492,
          "train_metric": 0.00025627101422287524,
          "validation_metric": 0.00019111494475509971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.413631146999251
        },
        {
          "epoch": 576,
          "updates": 576,
          "samples_seen": 589824,
          "online_loss": 0.0025444829370826483,
          "train_loss": 0.002520662499591708,
          "validation_loss": 0.0019877064041793346,
          "train_metric": 0.00025387186906300485,
          "validation_metric": 0.0002001945103984326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.422033467999427
        },
        {
          "epoch": 577,
          "updates": 577,
          "samples_seen": 590848,
          "online_loss": 0.002520662499591708,
          "train_loss": 0.002510236343368888,
          "validation_loss": 0.0019648736342787743,
          "train_metric": 0.00025282177375629544,
          "validation_metric": 0.00019789488578680903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.429633245999867
        },
        {
          "epoch": 578,
          "updates": 578,
          "samples_seen": 591872,
          "online_loss": 0.0025102361105382442,
          "train_loss": 0.002514531370252371,
          "validation_loss": 0.0018909118371084332,
          "train_metric": 0.0002532544021960348,
          "validation_metric": 0.00019044570217374712,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.438359030998981
        },
        {
          "epoch": 579,
          "updates": 579,
          "samples_seen": 592896,
          "online_loss": 0.002514531370252371,
          "train_loss": 0.002508185338228941,
          "validation_loss": 0.0020024669356644154,
          "train_metric": 0.00025261519476771355,
          "validation_metric": 0.00020168111950624734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.446406330998798
        },
        {
          "epoch": 580,
          "updates": 580,
          "samples_seen": 593920,
          "online_loss": 0.0025081848725676537,
          "train_loss": 0.0024906331673264503,
          "validation_loss": 0.0018999510211870074,
          "train_metric": 0.000250847457209602,
          "validation_metric": 0.00019135611364617944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.454370140998435
        },
        {
          "epoch": 581,
          "updates": 581,
          "samples_seen": 594944,
          "online_loss": 0.0024906331673264503,
          "train_loss": 0.002480782102793455,
          "validation_loss": 0.0019089869456365705,
          "train_metric": 0.0002498552785255015,
          "validation_metric": 0.00019226616132073104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.462378878999516
        },
        {
          "epoch": 582,
          "updates": 582,
          "samples_seen": 595968,
          "online_loss": 0.002480782102793455,
          "train_loss": 0.0024810016620904207,
          "validation_loss": 0.001968235708773136,
          "train_metric": 0.00024987742654047906,
          "validation_metric": 0.00019823350885417312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.470322584998939
        },
        {
          "epoch": 583,
          "updates": 583,
          "samples_seen": 596992,
          "online_loss": 0.002481002127751708,
          "train_loss": 0.002475849585607648,
          "validation_loss": 0.0018689371645450592,
          "train_metric": 0.00024935847613960505,
          "validation_metric": 0.00018823247228283435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.479264298999624
        },
        {
          "epoch": 584,
          "updates": 584,
          "samples_seen": 598016,
          "online_loss": 0.0024758498184382915,
          "train_loss": 0.0024618699681013823,
          "validation_loss": 0.0019352416275069118,
          "train_metric": 0.00024795052013359964,
          "validation_metric": 0.00019491041894070804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.487118032999206
        },
        {
          "epoch": 585,
          "updates": 585,
          "samples_seen": 599040,
          "online_loss": 0.0024618697352707386,
          "train_loss": 0.0024521220475435257,
          "validation_loss": 0.0019077829783782363,
          "train_metric": 0.00024696876062080264,
          "validation_metric": 0.00019214490021113306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.494898156999625
        },
        {
          "epoch": 586,
          "updates": 586,
          "samples_seen": 600064,
          "online_loss": 0.0024521222803741693,
          "train_loss": 0.002449717838317156,
          "validation_loss": 0.0018644803203642368,
          "train_metric": 0.0002467266167514026,
          "validation_metric": 0.00018778361845761538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.5034710919990175
        },
        {
          "epoch": 587,
          "updates": 587,
          "samples_seen": 601088,
          "online_loss": 0.0024497180711477995,
          "train_loss": 0.002444739453494549,
          "validation_loss": 0.001936863991431892,
          "train_metric": 0.00024622524506412446,
          "validation_metric": 0.00019507385150063783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.511538075999852
        },
        {
          "epoch": 588,
          "updates": 588,
          "samples_seen": 602112,
          "online_loss": 0.002444739453494549,
          "train_loss": 0.0024338909424841404,
          "validation_loss": 0.001860598917119205,
          "train_metric": 0.0002451325417496264,
          "validation_metric": 0.00018739269580692053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.519881433998307
        },
        {
          "epoch": 589,
          "updates": 589,
          "samples_seen": 603136,
          "online_loss": 0.0024338907096534967,
          "train_loss": 0.0024241963401436806,
          "validation_loss": 0.0018790381727740169,
          "train_metric": 0.0002441561664454639,
          "validation_metric": 0.00018924982578027993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.528319401999397
        },
        {
          "epoch": 590,
          "updates": 590,
          "samples_seen": 604160,
          "online_loss": 0.0024241963401436806,
          "train_loss": 0.002419683150947094,
          "validation_loss": 0.0018998617306351662,
          "train_metric": 0.0002437016082694754,
          "validation_metric": 0.00019134710601065308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.536172802998408
        },
        {
          "epoch": 591,
          "updates": 591,
          "samples_seen": 605184,
          "online_loss": 0.0024196829181164503,
          "train_loss": 0.0024151625111699104,
          "validation_loss": 0.0018382904818281531,
          "train_metric": 0.00024324633704964072,
          "validation_metric": 0.00018514590919949114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.544959531998757
        },
        {
          "epoch": 592,
          "updates": 592,
          "samples_seen": 606208,
          "online_loss": 0.0024151625111699104,
          "train_loss": 0.0024064346216619015,
          "validation_loss": 0.0018918325658887625,
          "train_metric": 0.00024236728495452553,
          "validation_metric": 0.00019053847063332796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.552736952999112
        },
        {
          "epoch": 593,
          "updates": 593,
          "samples_seen": 607232,
          "online_loss": 0.0024064346216619015,
          "train_loss": 0.002397098345682025,
          "validation_loss": 0.0018517831340432167,
          "train_metric": 0.00024142696929629892,
          "validation_metric": 0.00018650481069926172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.560919178999029
        },
        {
          "epoch": 594,
          "updates": 594,
          "samples_seen": 608256,
          "online_loss": 0.002397098345682025,
          "train_loss": 0.00239098584279418,
          "validation_loss": 0.001839173724874854,
          "train_metric": 0.00024081135052256286,
          "validation_metric": 0.0001852348359534517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.568622158998551
        },
        {
          "epoch": 595,
          "updates": 595,
          "samples_seen": 609280,
          "online_loss": 0.00239098584279418,
          "train_loss": 0.002386205131188035,
          "validation_loss": 0.0018757872749119997,
          "train_metric": 0.0002403298276476562,
          "validation_metric": 0.0001889224222395569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.576622086999123
        },
        {
          "epoch": 596,
          "updates": 596,
          "samples_seen": 610304,
          "online_loss": 0.002386205131188035,
          "train_loss": 0.0023792923893779516,
          "validation_loss": 0.0018195078009739518,
          "train_metric": 0.00023963362036738545,
          "validation_metric": 0.00018325416021980345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.584972591999758
        },
        {
          "epoch": 597,
          "updates": 597,
          "samples_seen": 611328,
          "online_loss": 0.0023792923893779516,
          "train_loss": 0.00237071979790926,
          "validation_loss": 0.0018506436608731747,
          "train_metric": 0.00023877022613305598,
          "validation_metric": 0.0001863900397438556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.5930001019987685
        },
        {
          "epoch": 598,
          "updates": 598,
          "samples_seen": 612352,
          "online_loss": 0.0023707200307399035,
          "train_loss": 0.0023635211400687695,
          "validation_loss": 0.001836960669606924,
          "train_metric": 0.00023804523516446352,
          "validation_metric": 0.00018501197337172925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.600912695999796
        },
        {
          "epoch": 599,
          "updates": 599,
          "samples_seen": 613376,
          "online_loss": 0.002363521372899413,
          "train_loss": 0.0023580845445394516,
          "validation_loss": 0.0018106743227690458,
          "train_metric": 0.00023749763204250485,
          "validation_metric": 0.0001823644997784868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.609912428999451
        },
        {
          "epoch": 600,
          "updates": 600,
          "samples_seen": 614400,
          "online_loss": 0.0023580845445394516,
          "train_loss": 0.0023521678522229195,
          "validation_loss": 0.0018454823875799775,
          "train_metric": 0.000236901716561988,
          "validation_metric": 0.00018587024533189833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.6178154589997575
        },
        {
          "epoch": 601,
          "updates": 601,
          "samples_seen": 615424,
          "online_loss": 0.0023521678522229195,
          "train_loss": 0.0023447582498192787,
          "validation_loss": 0.0018017286201938987,
          "train_metric": 0.00023615549434907734,
          "validation_metric": 0.00018146350339520723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.632226678999359
        },
        {
          "epoch": 602,
          "updates": 602,
          "samples_seen": 616448,
          "online_loss": 0.0023447584826499224,
          "train_loss": 0.0023372068535536528,
          "validation_loss": 0.0018169153481721878,
          "train_metric": 0.00023539493849966675,
          "validation_metric": 0.00018299305520486087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.646731370999987
        },
        {
          "epoch": 603,
          "updates": 603,
          "samples_seen": 617472,
          "online_loss": 0.0023372070863842964,
          "train_loss": 0.0023308731615543365,
          "validation_loss": 0.0018155110301449895,
          "train_metric": 0.00023475699708797038,
          "validation_metric": 0.00018285161058884114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.654567392999525
        },
        {
          "epoch": 604,
          "updates": 604,
          "samples_seen": 618496,
          "online_loss": 0.0023308726958930492,
          "train_loss": 0.0023251755628734827,
          "validation_loss": 0.0017868875293061137,
          "train_metric": 0.00023418318596668541,
          "validation_metric": 0.00017996877431869507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.663352421999662
        },
        {
          "epoch": 605,
          "updates": 605,
          "samples_seen": 619520,
          "online_loss": 0.0023251755628734827,
          "train_loss": 0.0023188102059066296,
          "validation_loss": 0.0018148503731936216,
          "train_metric": 0.00023354207223746926,
          "validation_metric": 0.00018278506468050182,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.670990434999112
        },
        {
          "epoch": 606,
          "updates": 606,
          "samples_seen": 620544,
          "online_loss": 0.0023188102059066296,
          "train_loss": 0.0023117007222026587,
          "validation_loss": 0.0017816999461501837,
          "train_metric": 0.0002328260598005727,
          "validation_metric": 0.00017944630235433578,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.679806722999274
        },
        {
          "epoch": 607,
          "updates": 607,
          "samples_seen": 621568,
          "online_loss": 0.0023117007222026587,
          "train_loss": 0.0023047984577715397,
          "validation_loss": 0.0017889103619381785,
          "train_metric": 0.00023213088570628315,
          "validation_metric": 0.0001801725011318922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.6877281939996465
        },
        {
          "epoch": 608,
          "updates": 608,
          "samples_seen": 622592,
          "online_loss": 0.0023047984577715397,
          "train_loss": 0.0022986442781984806,
          "validation_loss": 0.0017905825516209006,
          "train_metric": 0.0002315110177733004,
          "validation_metric": 0.00018034092499874532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.695684879998225
        },
        {
          "epoch": 609,
          "updates": 609,
          "samples_seen": 623616,
          "online_loss": 0.0022986442781984806,
          "train_loss": 0.0022927408572286367,
          "validation_loss": 0.0017651834059506655,
          "train_metric": 0.00023091647017281502,
          "validation_metric": 0.00017778280016500503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.7041925659996195
        },
        {
          "epoch": 610,
          "updates": 610,
          "samples_seen": 624640,
          "online_loss": 0.0022927410900592804,
          "train_loss": 0.002286398783326149,
          "validation_loss": 0.001786543638445437,
          "train_metric": 0.00023027771385386586,
          "validation_metric": 0.00017993412620853633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.71240505699825
        },
        {
          "epoch": 611,
          "updates": 611,
          "samples_seen": 625664,
          "online_loss": 0.002286398783326149,
          "train_loss": 0.002279666718095541,
          "validation_loss": 0.0017598684644326568,
          "train_metric": 0.0002295996673638001,
          "validation_metric": 0.00017724750796332955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.72092341699863
        },
        {
          "epoch": 612,
          "updates": 612,
          "samples_seen": 626688,
          "online_loss": 0.0022796664852648973,
          "train_loss": 0.002273081336170435,
          "validation_loss": 0.00176477269269526,
          "train_metric": 0.00022893643472343683,
          "validation_metric": 0.000177741443621926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.728876154998943
        },
        {
          "epoch": 613,
          "updates": 613,
          "samples_seen": 627712,
          "online_loss": 0.0022730815690010786,
          "train_loss": 0.0022669299505650997,
          "validation_loss": 0.0017642735037952662,
          "train_metric": 0.00022831685782875866,
          "validation_metric": 0.00017769119585864246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.831124660999194
        },
        {
          "epoch": 614,
          "updates": 614,
          "samples_seen": 628736,
          "online_loss": 0.0022669294849038124,
          "train_loss": 0.0022609641309827566,
          "validation_loss": 0.001744152745231986,
          "train_metric": 0.0002277160092489794,
          "validation_metric": 0.00017566466704010963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.840226309999707
        },
        {
          "epoch": 615,
          "updates": 615,
          "samples_seen": 629760,
          "online_loss": 0.002260963898152113,
          "train_loss": 0.002254798775538802,
          "validation_loss": 0.0017598773119971156,
          "train_metric": 0.00022709507902618498,
          "validation_metric": 0.00017724839563015848,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.8486247919991
        },
        {
          "epoch": 616,
          "updates": 616,
          "samples_seen": 630784,
          "online_loss": 0.002254798775538802,
          "train_loss": 0.0022483887150883675,
          "validation_loss": 0.001736654550768435,
          "train_metric": 0.00022644948330707848,
          "validation_metric": 0.0001749094808474183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.857674696999311
        },
        {
          "epoch": 617,
          "updates": 617,
          "samples_seen": 631808,
          "online_loss": 0.002248388947919011,
          "train_loss": 0.002241993322968483,
          "validation_loss": 0.0017424678662791848,
          "train_metric": 0.00022580535733141005,
          "validation_metric": 0.00017549497715663165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.8657360059987695
        },
        {
          "epoch": 618,
          "updates": 618,
          "samples_seen": 632832,
          "online_loss": 0.002241993322968483,
          "train_loss": 0.0022358305286616087,
          "validation_loss": 0.0017374004237353802,
          "train_metric": 0.0002251846599392593,
          "validation_metric": 0.0001749846269376576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.873346782998851
        },
        {
          "epoch": 619,
          "updates": 619,
          "samples_seen": 633856,
          "online_loss": 0.002235830295830965,
          "train_loss": 0.002229858422651887,
          "validation_loss": 0.001723755500279367,
          "train_metric": 0.00022458315652329475,
          "validation_metric": 0.00017361034406349063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.882301308000024
        },
        {
          "epoch": 620,
          "updates": 620,
          "samples_seen": 634880,
          "online_loss": 0.0022298581898212433,
          "train_loss": 0.0022238660603761673,
          "validation_loss": 0.0017340016784146428,
          "train_metric": 0.00022397963039111346,
          "validation_metric": 0.0001746423076838255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.890239887999996
        },
        {
          "epoch": 621,
          "updates": 621,
          "samples_seen": 635904,
          "online_loss": 0.002223866293206811,
          "train_loss": 0.0022177258506417274,
          "validation_loss": 0.0017135600792244077,
          "train_metric": 0.000223361246753484,
          "validation_metric": 0.00017258351726923138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.89927277499919
        },
        {
          "epoch": 622,
          "updates": 622,
          "samples_seen": 636928,
          "online_loss": 0.002217726083472371,
          "train_loss": 0.002211526269093156,
          "validation_loss": 0.001721179811283946,
          "train_metric": 0.00022273683862295002,
          "validation_metric": 0.00017335094162262976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.907538684999963
        },
        {
          "epoch": 623,
          "updates": 623,
          "samples_seen": 637952,
          "online_loss": 0.002211526269093156,
          "train_loss": 0.002205400262027979,
          "validation_loss": 0.0017107403837144375,
          "train_metric": 0.00022211983741726726,
          "validation_metric": 0.00017229950753971934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.916742808998606
        },
        {
          "epoch": 624,
          "updates": 624,
          "samples_seen": 638976,
          "online_loss": 0.002205400262027979,
          "train_loss": 0.0021994118578732014,
          "validation_loss": 0.0017044596606865525,
          "train_metric": 0.00022151668963488191,
          "validation_metric": 0.000171666921232827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.925850825999078
        },
        {
          "epoch": 625,
          "updates": 625,
          "samples_seen": 640000,
          "online_loss": 0.0021994118578732014,
          "train_loss": 0.0021935091353952885,
          "validation_loss": 0.0017078588716685772,
          "train_metric": 0.00022092221479397267,
          "validation_metric": 0.00017200929869432002,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.933774026998435
        },
        {
          "epoch": 626,
          "updates": 626,
          "samples_seen": 641024,
          "online_loss": 0.0021935091353952885,
          "train_loss": 0.002187593374401331,
          "validation_loss": 0.00169177888892591,
          "train_metric": 0.0002203263866249472,
          "validation_metric": 0.00017038978694472462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.945884531998672
        },
        {
          "epoch": 627,
          "updates": 627,
          "samples_seen": 642048,
          "online_loss": 0.0021875936072319746,
          "train_loss": 0.0021816196385771036,
          "validation_loss": 0.0016993388999253511,
          "train_metric": 0.00021972472313791513,
          "validation_metric": 0.00017115117225330323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.953927668999313
        },
        {
          "epoch": 628,
          "updates": 628,
          "samples_seen": 643072,
          "online_loss": 0.00218161940574646,
          "train_loss": 0.002175624016672373,
          "validation_loss": 0.0016850258689373732,
          "train_metric": 0.0002191208623116836,
          "validation_metric": 0.00016970964497886598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.966619369999535
        },
        {
          "epoch": 629,
          "updates": 629,
          "samples_seen": 644096,
          "online_loss": 0.002175623783841729,
          "train_loss": 0.0021696656476706266,
          "validation_loss": 0.0016858032904565334,
          "train_metric": 0.00021852078498341143,
          "validation_metric": 0.0001697879342827946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.982780295998964
        },
        {
          "epoch": 630,
          "updates": 630,
          "samples_seen": 645120,
          "online_loss": 0.0021696656476706266,
          "train_loss": 0.002163785742595792,
          "validation_loss": 0.001681067980825901,
          "train_metric": 0.00021792856568936259,
          "validation_metric": 0.00016931099526118487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.994405434999862
        },
        {
          "epoch": 631,
          "updates": 631,
          "samples_seen": 646144,
          "online_loss": 0.002163785509765148,
          "train_loss": 0.0021579687017947435,
          "validation_loss": 0.001672255340963602,
          "train_metric": 0.00021734272013418376,
          "validation_metric": 0.0001684234302956611,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.00324647099842
        },
        {
          "epoch": 632,
          "updates": 632,
          "samples_seen": 647168,
          "online_loss": 0.0021579687017947435,
          "train_loss": 0.002152174012735486,
          "validation_loss": 0.0016755707329139113,
          "train_metric": 0.00021675908647011966,
          "validation_metric": 0.00016875735309440643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.015305063998312
        },
        {
          "epoch": 633,
          "updates": 633,
          "samples_seen": 648192,
          "online_loss": 0.002152174012735486,
          "train_loss": 0.002146371640264988,
          "validation_loss": 0.001661930582486093,
          "train_metric": 0.00021617466700263321,
          "validation_metric": 0.00016738357953727245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.027545660999749
        },
        {
          "epoch": 634,
          "updates": 634,
          "samples_seen": 649216,
          "online_loss": 0.0021463714074343443,
          "train_loss": 0.0021405599545687437,
          "validation_loss": 0.0016662625130265951,
          "train_metric": 0.00021558937442023307,
          "validation_metric": 0.0001678198459558189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.03603730299983
        },
        {
          "epoch": 635,
          "updates": 635,
          "samples_seen": 650240,
          "online_loss": 0.0021405601873993874,
          "train_loss": 0.0021347524598240852,
          "validation_loss": 0.0016550165601074696,
          "train_metric": 0.00021500446018762887,
          "validation_metric": 0.00016668719763401896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.052795658999457
        },
        {
          "epoch": 636,
          "updates": 636,
          "samples_seen": 651264,
          "online_loss": 0.0021347524598240852,
          "train_loss": 0.002128979656845331,
          "validation_loss": 0.0016545094549655914,
          "train_metric": 0.00021442303841467947,
          "validation_metric": 0.00016663613496348262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.069976102999135
        },
        {
          "epoch": 637,
          "updates": 637,
          "samples_seen": 652288,
          "online_loss": 0.002128979889675975,
          "train_loss": 0.002123254118487239,
          "validation_loss": 0.0016495687887072563,
          "train_metric": 0.00021384637511800975,
          "validation_metric": 0.00016613854677416384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.0838580499985255
        },
        {
          "epoch": 638,
          "updates": 638,
          "samples_seen": 653312,
          "online_loss": 0.002123254118487239,
          "train_loss": 0.002117567230015993,
          "validation_loss": 0.0016427022637799382,
          "train_metric": 0.000213273597182706,
          "validation_metric": 0.00016544695245102048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.096890713999528
        },
        {
          "epoch": 639,
          "updates": 639,
          "samples_seen": 654336,
          "online_loss": 0.002117567230015993,
          "train_loss": 0.0021119045559316874,
          "validation_loss": 0.0016432582633569837,
          "train_metric": 0.00021270332217682153,
          "validation_metric": 0.0001655029773246497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.116764041998977
        },
        {
          "epoch": 640,
          "updates": 640,
          "samples_seen": 655360,
          "online_loss": 0.002111904788762331,
          "train_loss": 0.002106258412823081,
          "validation_loss": 0.0016326000913977623,
          "train_metric": 0.00021213463332969695,
          "validation_metric": 0.00016442952619399875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.12964103899867
        },
        {
          "epoch": 641,
          "updates": 641,
          "samples_seen": 656384,
          "online_loss": 0.002106258412823081,
          "train_loss": 0.0021006218157708645,
          "validation_loss": 0.0016351414378732443,
          "train_metric": 0.00021156691946089268,
          "validation_metric": 0.00016468545072712004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.13764637299937
        },
        {
          "epoch": 642,
          "updates": 642,
          "samples_seen": 657408,
          "online_loss": 0.0021006218157708645,
          "train_loss": 0.0020949936006218195,
          "validation_loss": 0.00162444484885782,
          "train_metric": 0.00021100010781083256,
          "validation_metric": 0.00016360817244276404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.146140225999261
        },
        {
          "epoch": 643,
          "updates": 643,
          "samples_seen": 658432,
          "online_loss": 0.002094993833452463,
          "train_loss": 0.002089381916448474,
          "validation_loss": 0.0016253708163276315,
          "train_metric": 0.0002104348677676171,
          "validation_metric": 0.0001637014065636322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.15406868899845
        },
        {
          "epoch": 644,
          "updates": 644,
          "samples_seen": 659456,
          "online_loss": 0.002089381916448474,
          "train_loss": 0.0020837902557104826,
          "validation_loss": 0.001617494854144752,
          "train_metric": 0.00020987172320019454,
          "validation_metric": 0.00016290815256070346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.162762853999084
        },
        {
          "epoch": 645,
          "updates": 645,
          "samples_seen": 660480,
          "online_loss": 0.002083790022879839,
          "train_loss": 0.002078232355415821,
          "validation_loss": 0.0016150153242051601,
          "train_metric": 0.00020931196922902018,
          "validation_metric": 0.0001626584562472999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.171586053998908
        },
        {
          "epoch": 646,
          "updates": 646,
          "samples_seen": 661504,
          "online_loss": 0.002078232355415821,
          "train_loss": 0.002072698902338743,
          "validation_loss": 0.0016108723357319832,
          "train_metric": 0.00020875465997960418,
          "validation_metric": 0.00016224116552621126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.180522498998471
        },
        {
          "epoch": 647,
          "updates": 647,
          "samples_seen": 662528,
          "online_loss": 0.002072699135169387,
          "train_loss": 0.002067194553092122,
          "validation_loss": 0.0016049188561737537,
          "train_metric": 0.0002082002320094034,
          "validation_metric": 0.00016164153930731118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.188484550999419
        },
        {
          "epoch": 648,
          "updates": 648,
          "samples_seen": 663552,
          "online_loss": 0.002067194553092122,
          "train_loss": 0.0020617065019905567,
          "validation_loss": 0.0016038015019148588,
          "train_metric": 0.00020764750661328435,
          "validation_metric": 0.00016152900934685022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.196973093999986
        },
        {
          "epoch": 649,
          "updates": 649,
          "samples_seen": 664576,
          "online_loss": 0.002061706269159913,
          "train_loss": 0.0020562447607517242,
          "validation_loss": 0.00159552157856524,
          "train_metric": 0.00020709741511382163,
          "validation_metric": 0.0001606950827408582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.205335485999967
        },
        {
          "epoch": 650,
          "updates": 650,
          "samples_seen": 665600,
          "online_loss": 0.0020562447607517242,
          "train_loss": 0.0020508025772869587,
          "validation_loss": 0.0015961243771016598,
          "train_metric": 0.00020654933177866042,
          "validation_metric": 0.00016075580788310617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.213426874999641
        },
        {
          "epoch": 651,
          "updates": 651,
          "samples_seen": 666624,
          "online_loss": 0.002050802344456315,
          "train_loss": 0.0020453750621527433,
          "validation_loss": 0.0015868793707340956,
          "train_metric": 0.00020600268908310682,
          "validation_metric": 0.00015982465993147343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.319464454998524
        },
        {
          "epoch": 652,
          "updates": 652,
          "samples_seen": 667648,
          "online_loss": 0.002045375294983387,
          "train_loss": 0.002039965707808733,
          "validation_loss": 0.0015878576086834073,
          "train_metric": 0.00020545786537695676,
          "validation_metric": 0.00015992320550139993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.327716845999021
        },
        {
          "epoch": 653,
          "updates": 653,
          "samples_seen": 668672,
          "online_loss": 0.002039965707808733,
          "train_loss": 0.0020345754455775023,
          "validation_loss": 0.0015787315787747502,
          "train_metric": 0.00020491496252361685,
          "validation_metric": 0.00015900404832791537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.336094256999786
        },
        {
          "epoch": 654,
          "updates": 654,
          "samples_seen": 669696,
          "online_loss": 0.0020345752127468586,
          "train_loss": 0.0020291979890316725,
          "validation_loss": 0.0015793072525411844,
          "train_metric": 0.00020437338389456272,
          "validation_metric": 0.00015906205226201564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.344178730998465
        },
        {
          "epoch": 655,
          "updates": 655,
          "samples_seen": 670720,
          "online_loss": 0.0020291979890316725,
          "train_loss": 0.002023849170655012,
          "validation_loss": 0.0015708948485553265,
          "train_metric": 0.00020383470109663904,
          "validation_metric": 0.0001582147815497592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.352459594998436
        },
        {
          "epoch": 656,
          "updates": 656,
          "samples_seen": 671744,
          "online_loss": 0.002023849403485656,
          "train_loss": 0.002018516417592764,
          "validation_loss": 0.0015706917038187385,
          "train_metric": 0.00020329756080172956,
          "validation_metric": 0.00015819432155694813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.361258408998765
        },
        {
          "epoch": 657,
          "updates": 657,
          "samples_seen": 672768,
          "online_loss": 0.0020185161847621202,
          "train_loss": 0.0020132081117480993,
          "validation_loss": 0.0015631052665412426,
          "train_metric": 0.00020276296709198505,
          "validation_metric": 0.00015743022959213704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.369929176998994
        },
        {
          "epoch": 658,
          "updates": 658,
          "samples_seen": 673792,
          "online_loss": 0.002013208344578743,
          "train_loss": 0.002007915638387203,
          "validation_loss": 0.0015621738275513053,
          "train_metric": 0.00020222993043716997,
          "validation_metric": 0.00015733642794657499,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.3787571389984805
        },
        {
          "epoch": 659,
          "updates": 659,
          "samples_seen": 674816,
          "online_loss": 0.002007915871217847,
          "train_loss": 0.0020026478450745344,
          "validation_loss": 0.0015553137054666877,
          "train_metric": 0.00020169935305602849,
          "validation_metric": 0.00015664550301153213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.386941847998969
        },
        {
          "epoch": 660,
          "updates": 660,
          "samples_seen": 675840,
          "online_loss": 0.0020026476122438908,
          "train_loss": 0.001997394021600485,
          "validation_loss": 0.0015538681764155626,
          "train_metric": 0.00020117020176257938,
          "validation_metric": 0.0001564998965477571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.395823287999519
        },
        {
          "epoch": 661,
          "updates": 661,
          "samples_seen": 676864,
          "online_loss": 0.0019973942544311285,
          "train_loss": 0.0019921662751585245,
          "validation_loss": 0.001547383377328515,
          "train_metric": 0.0002006436843657866,
          "validation_metric": 0.00015584679204039276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.404191168998295
        },
        {
          "epoch": 662,
          "updates": 662,
          "samples_seen": 677888,
          "online_loss": 0.001992166042327881,
          "train_loss": 0.001986957620829344,
          "validation_loss": 0.0015457968693226576,
          "train_metric": 0.000200119087821804,
          "validation_metric": 0.00015568699745927006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.413022156999432
        },
        {
          "epoch": 663,
          "updates": 663,
          "samples_seen": 678912,
          "online_loss": 0.0019869571551680565,
          "train_loss": 0.001981767825782299,
          "validation_loss": 0.00153932289686054,
          "train_metric": 0.0001995964121306315,
          "validation_metric": 0.00015503498434554785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.421558133999497
        },
        {
          "epoch": 664,
          "updates": 664,
          "samples_seen": 679936,
          "online_loss": 0.0019817680586129427,
          "train_loss": 0.001976599683985114,
          "validation_loss": 0.0015380139229819179,
          "train_metric": 0.00019907589012291282,
          "validation_metric": 0.00015490312944166362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.430848736999906
        },
        {
          "epoch": 665,
          "updates": 665,
          "samples_seen": 680960,
          "online_loss": 0.0019765999168157578,
          "train_loss": 0.0019714527297765017,
          "validation_loss": 0.0015310666058212519,
          "train_metric": 0.00019855749269481748,
          "validation_metric": 0.00015420341514982283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.442749357998764
        },
        {
          "epoch": 666,
          "updates": 666,
          "samples_seen": 681984,
          "online_loss": 0.0019714527297765017,
          "train_loss": 0.0019663257990032434,
          "validation_loss": 0.001530581503175199,
          "train_metric": 0.00019804116163868457,
          "validation_metric": 0.0001541545643704012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.465362976998222
        },
        {
          "epoch": 667,
          "updates": 667,
          "samples_seen": 683008,
          "online_loss": 0.001966326031833887,
          "train_loss": 0.001961221219971776,
          "validation_loss": 0.0015224909875541925,
          "train_metric": 0.00019752704247366637,
          "validation_metric": 0.00015333971532527357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.4802831539982435
        },
        {
          "epoch": 668,
          "updates": 668,
          "samples_seen": 684032,
          "online_loss": 0.0019612209871411324,
          "train_loss": 0.0019561396911740303,
          "validation_loss": 0.001523706247098744,
          "train_metric": 0.00019701523706316948,
          "validation_metric": 0.00015346211148425937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.4942264199999045
        },
        {
          "epoch": 669,
          "updates": 669,
          "samples_seen": 685056,
          "online_loss": 0.0019561396911740303,
          "train_loss": 0.0019510877318680286,
          "validation_loss": 0.0015133607666939497,
          "train_metric": 0.00019650640024337918,
          "validation_metric": 0.0001524201361462474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.50858009699914
        },
        {
          "epoch": 670,
          "updates": 670,
          "samples_seen": 686080,
          "online_loss": 0.0019510876154527068,
          "train_loss": 0.0019460762850940228,
          "validation_loss": 0.0015177734894677997,
          "train_metric": 0.00019600166706368327,
          "validation_metric": 0.00015286458074115217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.518312590998903
        },
        {
          "epoch": 671,
          "updates": 671,
          "samples_seen": 687104,
          "online_loss": 0.0019460765179246664,
          "train_loss": 0.0019411108223721385,
          "validation_loss": 0.0015031392686069012,
          "train_metric": 0.00019550157594494522,
          "validation_metric": 0.00015139069000724703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.53024408499914
        },
        {
          "epoch": 672,
          "updates": 672,
          "samples_seen": 688128,
          "online_loss": 0.0019411105895414948,
          "train_loss": 0.0019362256862223148,
          "validation_loss": 0.0015136920846998692,
          "train_metric": 0.00019500954658724368,
          "validation_metric": 0.0001524535327916965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.542050066998854
        },
        {
          "epoch": 673,
          "updates": 673,
          "samples_seen": 689152,
          "online_loss": 0.001936225569806993,
          "train_loss": 0.0019314814126119018,
          "validation_loss": 0.0014908029697835445,
          "train_metric": 0.00019453173445072025,
          "validation_metric": 0.0001501482183812186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.552000424999278
        },
        {
          "epoch": 674,
          "updates": 674,
          "samples_seen": 690176,
          "online_loss": 0.0019314816454425454,
          "train_loss": 0.0019269885960966349,
          "validation_loss": 0.0015134833520278335,
          "train_metric": 0.0001940792571986094,
          "validation_metric": 0.00015243253437802196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.564752664999105
        },
        {
          "epoch": 675,
          "updates": 675,
          "samples_seen": 691200,
          "online_loss": 0.0019269885960966349,
          "train_loss": 0.001922992873005569,
          "validation_loss": 0.0014743809588253498,
          "train_metric": 0.00019367678032722324,
          "validation_metric": 0.00014849424769636244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.57772768399991
        },
        {
          "epoch": 676,
          "updates": 676,
          "samples_seen": 692224,
          "online_loss": 0.001922992873005569,
          "train_loss": 0.001919939648360014,
          "validation_loss": 0.00152209447696805,
          "train_metric": 0.00019336931291036308,
          "validation_metric": 0.0001532997703179717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.586144256998523
        },
        {
          "epoch": 677,
          "updates": 677,
          "samples_seen": 693248,
          "online_loss": 0.001919940230436623,
          "train_loss": 0.001918958150781691,
          "validation_loss": 0.0014509434113278985,
          "train_metric": 0.00019327043264638633,
          "validation_metric": 0.0001461337087675929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.5980778739995
        },
        {
          "epoch": 678,
          "updates": 678,
          "samples_seen": 694272,
          "online_loss": 0.0019189580343663692,
          "train_loss": 0.0019217869266867638,
          "validation_loss": 0.0015525863273069263,
          "train_metric": 0.0001935553300427273,
          "validation_metric": 0.00015637080650776625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.6118589069992595
        },
        {
          "epoch": 679,
          "updates": 679,
          "samples_seen": 695296,
          "online_loss": 0.001921786810271442,
          "train_loss": 0.0019334717653691769,
          "validation_loss": 0.001420109416358173,
          "train_metric": 0.000194732187082991,
          "validation_metric": 0.00014302819909062237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.62389219799843
        },
        {
          "epoch": 680,
          "updates": 680,
          "samples_seen": 696320,
          "online_loss": 0.0019334719981998205,
          "train_loss": 0.001958497567102313,
          "validation_loss": 0.0016388422809541225,
          "train_metric": 0.00019725272431969643,
          "validation_metric": 0.00016505818348377943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.633485049998853
        },
        {
          "epoch": 681,
          "updates": 681,
          "samples_seen": 697344,
          "online_loss": 0.001958497567102313,
          "train_loss": 0.0020141613204032183,
          "validation_loss": 0.0014036456122994423,
          "train_metric": 0.00020285895152483135,
          "validation_metric": 0.0001413700229022652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.648253122999449
        },
        {
          "epoch": 682,
          "updates": 682,
          "samples_seen": 698368,
          "online_loss": 0.0020141613204032183,
          "train_loss": 0.0020805280655622482,
          "validation_loss": 0.0018213123548775911,
          "train_metric": 0.00020954315550625324,
          "validation_metric": 0.00018343589908909053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.657016029999795
        },
        {
          "epoch": 683,
          "updates": 683,
          "samples_seen": 699392,
          "online_loss": 0.0020805280655622482,
          "train_loss": 0.0021671282593160868,
          "validation_loss": 0.001436756458133459,
          "train_metric": 0.00021826522424817085,
          "validation_metric": 0.0001447048271074891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.665852209998775
        },
        {
          "epoch": 684,
          "updates": 684,
          "samples_seen": 700416,
          "online_loss": 0.002167128026485443,
          "train_loss": 0.002114185830578208,
          "validation_loss": 0.001865258440375328,
          "train_metric": 0.00021293305326253176,
          "validation_metric": 0.0001878620096249506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.674080558999776
        },
        {
          "epoch": 685,
          "updates": 685,
          "samples_seen": 701440,
          "online_loss": 0.0021141860634088516,
          "train_loss": 0.0019962741062045097,
          "validation_loss": 0.0013890336267650127,
          "train_metric": 0.00020105739531572908,
          "validation_metric": 0.00013989838771522045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.684001109999372
        },
        {
          "epoch": 686,
          "updates": 686,
          "samples_seen": 702464,
          "online_loss": 0.001996273873373866,
          "train_loss": 0.0018782539991661906,
          "validation_loss": 0.0015031842049211264,
          "train_metric": 0.0001891708525363356,
          "validation_metric": 0.00015139521565288305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.693011549999937
        },
        {
          "epoch": 687,
          "updates": 687,
          "samples_seen": 703488,
          "online_loss": 0.0018782541155815125,
          "train_loss": 0.0018946656491607428,
          "validation_loss": 0.0015527310315519571,
          "train_metric": 0.00019082377548329532,
          "validation_metric": 0.00015638538752682507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.7018657389999134
        },
        {
          "epoch": 688,
          "updates": 688,
          "samples_seen": 704512,
          "online_loss": 0.0018946656491607428,
          "train_loss": 0.001985089387744665,
          "validation_loss": 0.0013820987660437822,
          "train_metric": 0.00019993091700598598,
          "validation_metric": 0.00013919993944000453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.711578212998575
        },
        {
          "epoch": 689,
          "updates": 689,
          "samples_seen": 705536,
          "online_loss": 0.0019850891549140215,
          "train_loss": 0.0019884819630533457,
          "validation_loss": 0.0017097856616601348,
          "train_metric": 0.00020027261052746326,
          "validation_metric": 0.00017220336303580552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.720258849998572
        },
        {
          "epoch": 690,
          "updates": 690,
          "samples_seen": 706560,
          "online_loss": 0.0019884819630533457,
          "train_loss": 0.0019115949980914593,
          "validation_loss": 0.0013734352542087436,
          "train_metric": 0.00019252885249443352,
          "validation_metric": 0.0001383273774990812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.730186483999205
        },
        {
          "epoch": 691,
          "updates": 691,
          "samples_seen": 707584,
          "online_loss": 0.0019115949980914593,
          "train_loss": 0.0018485735636204481,
          "validation_loss": 0.0014478611992672086,
          "train_metric": 0.000186181569006294,
          "validation_metric": 0.00014582327276002616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.853251546999672
        },
        {
          "epoch": 692,
          "updates": 692,
          "samples_seen": 708608,
          "online_loss": 0.0018485735636204481,
          "train_loss": 0.0018800219986587763,
          "validation_loss": 0.0015501697780564427,
          "train_metric": 0.0001893489243229851,
          "validation_metric": 0.00015612739662174135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.861971757998617
        },
        {
          "epoch": 693,
          "updates": 693,
          "samples_seen": 709632,
          "online_loss": 0.0018800218822434545,
          "train_loss": 0.0019304216839373112,
          "validation_loss": 0.0013621925609186292,
          "train_metric": 0.00019442499615252018,
          "validation_metric": 0.00013719503476750106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.870436244998928
        },
        {
          "epoch": 694,
          "updates": 694,
          "samples_seen": 710656,
          "online_loss": 0.0019304215675219893,
          "train_loss": 0.0018983891932293773,
          "validation_loss": 0.001587212085723877,
          "train_metric": 0.00019119882199447602,
          "validation_metric": 0.00015985820209607482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.878450095999142
        },
        {
          "epoch": 695,
          "updates": 695,
          "samples_seen": 711680,
          "online_loss": 0.0018983889603987336,
          "train_loss": 0.0018406093586236238,
          "validation_loss": 0.0013864602660760283,
          "train_metric": 0.00018537942378316075,
          "validation_metric": 0.000139639203553088,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.886172214999533
        },
        {
          "epoch": 696,
          "updates": 696,
          "samples_seen": 712704,
          "online_loss": 0.0018406093586236238,
          "train_loss": 0.0018357096705585718,
          "validation_loss": 0.0013878941535949707,
          "train_metric": 0.00018488598288968205,
          "validation_metric": 0.00013978361675981432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.89443668199965
        },
        {
          "epoch": 697,
          "updates": 697,
          "samples_seen": 713728,
          "online_loss": 0.0018357099033892155,
          "train_loss": 0.0018694999162107706,
          "validation_loss": 0.0015535675920546055,
          "train_metric": 0.00018828916654456407,
          "validation_metric": 0.00015646967221982777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.902284941999824
        },
        {
          "epoch": 698,
          "updates": 698,
          "samples_seen": 714752,
          "online_loss": 0.0018694997997954488,
          "train_loss": 0.0018730014562606812,
          "validation_loss": 0.0013509334530681372,
          "train_metric": 0.00018864187586586922,
          "validation_metric": 0.00013606106222141534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.911158174998491
        },
        {
          "epoch": 699,
          "updates": 699,
          "samples_seen": 715776,
          "online_loss": 0.0018730016890913248,
          "train_loss": 0.0018314882181584835,
          "validation_loss": 0.0014810776337981224,
          "train_metric": 0.0001844607904786244,
          "validation_metric": 0.000149168714415282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.918909745998462
        },
        {
          "epoch": 700,
          "updates": 700,
          "samples_seen": 716800,
          "online_loss": 0.0018314882181584835,
          "train_loss": 0.0018119470914825797,
          "validation_loss": 0.0014239440206438303,
          "train_metric": 0.00018249267304781824,
          "validation_metric": 0.0001434144505765289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.926960666998639
        },
        {
          "epoch": 701,
          "updates": 701,
          "samples_seen": 717824,
          "online_loss": 0.0018119469750672579,
          "train_loss": 0.0018307564314454794,
          "validation_loss": 0.0013581969542428851,
          "train_metric": 0.00018438708502799273,
          "validation_metric": 0.00013679264520760626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.935591166999075
        },
        {
          "epoch": 702,
          "updates": 702,
          "samples_seen": 718848,
          "online_loss": 0.0018307563150301576,
          "train_loss": 0.0018392687197774649,
          "validation_loss": 0.001518802484497428,
          "train_metric": 0.00018524439656175673,
          "validation_metric": 0.0001529682194814086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.9472854629984795
        },
        {
          "epoch": 703,
          "updates": 703,
          "samples_seen": 719872,
          "online_loss": 0.0018392687197774649,
          "train_loss": 0.0018183066276833415,
          "validation_loss": 0.0013504362432286143,
          "train_metric": 0.00018313316104467958,
          "validation_metric": 0.00013601100363302976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.956193178999456
        },
        {
          "epoch": 704,
          "updates": 704,
          "samples_seen": 720896,
          "online_loss": 0.0018183065112680197,
          "train_loss": 0.0017957284580916166,
          "validation_loss": 0.0014063906855881214,
          "train_metric": 0.00018085920601151884,
          "validation_metric": 0.0001416465238435194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.964445022998916
        },
        {
          "epoch": 705,
          "updates": 705,
          "samples_seen": 721920,
          "online_loss": 0.0017957284580916166,
          "train_loss": 0.001800251193344593,
          "validation_loss": 0.0014460165984928608,
          "train_metric": 0.00018131473916582763,
          "validation_metric": 0.00014563751756213605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.972055862999696
        },
        {
          "epoch": 706,
          "updates": 706,
          "samples_seen": 722944,
          "online_loss": 0.0018002514261752367,
          "train_loss": 0.0018114617560058832,
          "validation_loss": 0.0013410523533821106,
          "train_metric": 0.00018244379316456616,
          "validation_metric": 0.0001350658858427778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.981175164999513
        },
        {
          "epoch": 707,
          "updates": 707,
          "samples_seen": 723968,
          "online_loss": 0.0018114617560058832,
          "train_loss": 0.001801176811568439,
          "validation_loss": 0.0014617597917094827,
          "train_metric": 0.0001814079296309501,
          "validation_metric": 0.0001472230942454189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.989748239999244
        },
        {
          "epoch": 708,
          "updates": 708,
          "samples_seen": 724992,
          "online_loss": 0.0018011769279837608,
          "train_loss": 0.0017824231181293726,
          "validation_loss": 0.0013620277168229222,
          "train_metric": 0.00017951913469005376,
          "validation_metric": 0.00013717844558414072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.9972129089983355
        },
        {
          "epoch": 709,
          "updates": 709,
          "samples_seen": 726016,
          "online_loss": 0.0017824231181293726,
          "train_loss": 0.0017775208689272404,
          "validation_loss": 0.001364647294394672,
          "train_metric": 0.0001790254027582705,
          "validation_metric": 0.00013744228635914624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.004936303999784
        },
        {
          "epoch": 710,
          "updates": 710,
          "samples_seen": 727040,
          "online_loss": 0.0017775208689272404,
          "train_loss": 0.0017842266242951155,
          "validation_loss": 0.001441621920093894,
          "train_metric": 0.00017970077169593424,
          "validation_metric": 0.0001451948774047196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.012866438999481
        },
        {
          "epoch": 711,
          "updates": 711,
          "samples_seen": 728064,
          "online_loss": 0.0017842263914644718,
          "train_loss": 0.0017830010037869215,
          "validation_loss": 0.0013326364569365978,
          "train_metric": 0.00017957731324713677,
          "validation_metric": 0.00013421826588455588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.021333467999284
        },
        {
          "epoch": 712,
          "updates": 712,
          "samples_seen": 729088,
          "online_loss": 0.0017830011202022433,
          "train_loss": 0.0017697521252557635,
          "validation_loss": 0.0014098354149609804,
          "train_metric": 0.00017824297538027167,
          "validation_metric": 0.00014199345605447888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.029195318998973
        },
        {
          "epoch": 713,
          "updates": 713,
          "samples_seen": 730112,
          "online_loss": 0.0017697522416710854,
          "train_loss": 0.0017601054860278964,
          "validation_loss": 0.001369165489450097,
          "train_metric": 0.0001772713876562193,
          "validation_metric": 0.0001378973393002525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.037649225998393
        },
        {
          "epoch": 714,
          "updates": 714,
          "samples_seen": 731136,
          "online_loss": 0.0017601054860278964,
          "train_loss": 0.0017611179500818253,
          "validation_loss": 0.0013397164875641465,
          "train_metric": 0.00017737333837430924,
          "validation_metric": 0.0001349313388345763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.04588829199929
        },
        {
          "epoch": 715,
          "updates": 715,
          "samples_seen": 732160,
          "online_loss": 0.0017611178336665034,
          "train_loss": 0.0017624442698433995,
          "validation_loss": 0.0014166701585054398,
          "train_metric": 0.0001775068958522752,
          "validation_metric": 0.00014268183440435678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.054221804999543
        },
        {
          "epoch": 716,
          "updates": 716,
          "samples_seen": 733184,
          "online_loss": 0.0017624440370127559,
          "train_loss": 0.0017559423577040434,
          "validation_loss": 0.0013256837846711278,
          "train_metric": 0.0001768520742189139,
          "validation_metric": 0.00013351801317185163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.063077065999096
        },
        {
          "epoch": 717,
          "updates": 717,
          "samples_seen": 734208,
          "online_loss": 0.0017559423577040434,
          "train_loss": 0.0017460505478084087,
          "validation_loss": 0.0013716394314542413,
          "train_metric": 0.000175855791894719,
          "validation_metric": 0.00013814651174470782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.071554133999598
        },
        {
          "epoch": 718,
          "updates": 718,
          "samples_seen": 735232,
          "online_loss": 0.001746050431393087,
          "train_loss": 0.0017417330527678132,
          "validation_loss": 0.001366583164781332,
          "train_metric": 0.00017542099521961063,
          "validation_metric": 0.00013763726747129112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.0791634219986
        },
        {
          "epoch": 719,
          "updates": 719,
          "samples_seen": 736256,
          "online_loss": 0.001741733169183135,
          "train_loss": 0.0017422345699742436,
          "validation_loss": 0.001324447919614613,
          "train_metric": 0.00017547149036545306,
          "validation_metric": 0.00013339353608898818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.087716312998964
        },
        {
          "epoch": 720,
          "updates": 720,
          "samples_seen": 737280,
          "online_loss": 0.0017422343371436,
          "train_loss": 0.0017398599302396178,
          "validation_loss": 0.0013884995132684708,
          "train_metric": 0.0001752323005348444,
          "validation_metric": 0.00013984457473270595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.096595446999345
        },
        {
          "epoch": 721,
          "updates": 721,
          "samples_seen": 738304,
          "online_loss": 0.0017398599302396178,
          "train_loss": 0.0017330123810097575,
          "validation_loss": 0.0013207430019974709,
          "train_metric": 0.0001745426416164264,
          "validation_metric": 0.0001330203958787024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.108235682999293
        },
        {
          "epoch": 722,
          "updates": 722,
          "samples_seen": 739328,
          "online_loss": 0.0017330122645944357,
          "train_loss": 0.001726160990074277,
          "validation_loss": 0.0013471769634634256,
          "train_metric": 0.00017385263345204294,
          "validation_metric": 0.00013568272697739303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.117064328998822
        },
        {
          "epoch": 723,
          "updates": 723,
          "samples_seen": 740352,
          "online_loss": 0.001726160990074277,
          "train_loss": 0.0017233710968866944,
          "validation_loss": 0.0013552344171330333,
          "train_metric": 0.00017357162141706795,
          "validation_metric": 0.00013649424363393337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.126589810999576
        },
        {
          "epoch": 724,
          "updates": 724,
          "samples_seen": 741376,
          "online_loss": 0.0017233710968866944,
          "train_loss": 0.0017222287133336067,
          "validation_loss": 0.0013120736693963408,
          "train_metric": 0.00017345655942335725,
          "validation_metric": 0.00013214725186116993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.13594161799847
        },
        {
          "epoch": 725,
          "updates": 725,
          "samples_seen": 742400,
          "online_loss": 0.0017222288297489285,
          "train_loss": 0.001718695741146803,
          "validation_loss": 0.0013626120053231716,
          "train_metric": 0.000173100721440278,
          "validation_metric": 0.000137237278977409,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.145203682999636
        },
        {
          "epoch": 726,
          "updates": 726,
          "samples_seen": 743424,
          "online_loss": 0.001718695741146803,
          "train_loss": 0.0017127852188423276,
          "validation_loss": 0.0013125556288287044,
          "train_metric": 0.0001725054462440312,
          "validation_metric": 0.00013219579705037177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.15315383699999
        },
        {
          "epoch": 727,
          "updates": 727,
          "samples_seen": 744448,
          "online_loss": 0.0017127852188423276,
          "train_loss": 0.0017075289506465197,
          "validation_loss": 0.0013298570411279798,
          "train_metric": 0.00017197606211993843,
          "validation_metric": 0.00013393833069130778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.161179546999847
        },
        {
          "epoch": 728,
          "updates": 728,
          "samples_seen": 745472,
          "online_loss": 0.0017075289506465197,
          "train_loss": 0.0017046664142981172,
          "validation_loss": 0.001338590169325471,
          "train_metric": 0.00017168775957543403,
          "validation_metric": 0.00013481789210345596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.169020991999787
        },
        {
          "epoch": 729,
          "updates": 729,
          "samples_seen": 746496,
          "online_loss": 0.0017046664142981172,
          "train_loss": 0.001702522044070065,
          "validation_loss": 0.001301600132137537,
          "train_metric": 0.00017147179460152984,
          "validation_metric": 0.00013109238352626562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.177649404999102
        },
        {
          "epoch": 730,
          "updates": 730,
          "samples_seen": 747520,
          "online_loss": 0.0017025221604853868,
          "train_loss": 0.0016989430878311396,
          "validation_loss": 0.001341617084108293,
          "train_metric": 0.0001711113436613232,
          "validation_metric": 0.00013512276927940547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.282852324999112
        },
        {
          "epoch": 731,
          "updates": 731,
          "samples_seen": 748544,
          "online_loss": 0.0016989430878311396,
          "train_loss": 0.0016939605120569468,
          "validation_loss": 0.0013024421641603112,
          "train_metric": 0.00017060950631275773,
          "validation_metric": 0.00013117719208821654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.290977182999995
        },
        {
          "epoch": 732,
          "updates": 732,
          "samples_seen": 749568,
          "online_loss": 0.0016939605120569468,
          "train_loss": 0.0016893591964617372,
          "validation_loss": 0.0013166975695639849,
          "train_metric": 0.00017014608602039516,
          "validation_metric": 0.00013261294225230813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.298482894999324
        },
        {
          "epoch": 733,
          "updates": 733,
          "samples_seen": 750592,
          "online_loss": 0.0016893591964617372,
          "train_loss": 0.0016860768664628267,
          "validation_loss": 0.001319337636232376,
          "train_metric": 0.00016981549561023712,
          "validation_metric": 0.00013287884939927608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.306758397999147
        },
        {
          "epoch": 734,
          "updates": 734,
          "samples_seen": 751616,
          "online_loss": 0.001686076750047505,
          "train_loss": 0.0016834293492138386,
          "validation_loss": 0.0012916193809360266,
          "train_metric": 0.00016954884631559253,
          "validation_metric": 0.0001300871663261205,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.315357562998543
        },
        {
          "epoch": 735,
          "updates": 735,
          "samples_seen": 752640,
          "online_loss": 0.0016834294656291604,
          "train_loss": 0.0016800768207758665,
          "validation_loss": 0.0013227497693151236,
          "train_metric": 0.0001692111836746335,
          "validation_metric": 0.0001332225074293092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.324461352998696
        },
        {
          "epoch": 736,
          "updates": 736,
          "samples_seen": 753664,
          "online_loss": 0.001680075889453292,
          "train_loss": 0.0016758567653596401,
          "validation_loss": 0.0012895343825221062,
          "train_metric": 0.0001687861658865586,
          "validation_metric": 0.00012987718218937516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.33331537399863
        },
        {
          "epoch": 737,
          "updates": 737,
          "samples_seen": 754688,
          "online_loss": 0.001675856881774962,
          "train_loss": 0.0016715567326173186,
          "validation_loss": 0.0013045419473201036,
          "train_metric": 0.00016835307178553194,
          "validation_metric": 0.00013138868962414563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.341303309998693
        },
        {
          "epoch": 738,
          "updates": 738,
          "samples_seen": 755712,
          "online_loss": 0.0016715566162019968,
          "train_loss": 0.001667940290644765,
          "validation_loss": 0.0012992852134630084,
          "train_metric": 0.00016798883734736592,
          "validation_metric": 0.00013085926184430718,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.349044233998939
        },
        {
          "epoch": 739,
          "updates": 739,
          "samples_seen": 756736,
          "online_loss": 0.0016679404070600867,
          "train_loss": 0.001664896379224956,
          "validation_loss": 0.0012828425969928503,
          "train_metric": 0.0001676822721492499,
          "validation_metric": 0.00012920322478748858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.357827488998737
        },
        {
          "epoch": 740,
          "updates": 740,
          "samples_seen": 757760,
          "online_loss": 0.0016648966120555997,
          "train_loss": 0.0016617524670436978,
          "validation_loss": 0.0013048691907897592,
          "train_metric": 0.00016736563702579588,
          "validation_metric": 0.00013142163516022265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.365541322998979
        },
        {
          "epoch": 741,
          "updates": 741,
          "samples_seen": 758784,
          "online_loss": 0.0016617524670436978,
          "train_loss": 0.0016581143718212843,
          "validation_loss": 0.0012762666447088122,
          "train_metric": 0.00016699920524843037,
          "validation_metric": 0.0001285409089177847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.373773563998839
        },
        {
          "epoch": 742,
          "updates": 742,
          "samples_seen": 759808,
          "online_loss": 0.0016581144882366061,
          "train_loss": 0.0016541534569114447,
          "validation_loss": 0.0012930119410157204,
          "train_metric": 0.00016660029359627515,
          "validation_metric": 0.00013022743223700672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.391405777998443
        },
        {
          "epoch": 743,
          "updates": 743,
          "samples_seen": 760832,
          "online_loss": 0.0016541536897420883,
          "train_loss": 0.0016503665829077363,
          "validation_loss": 0.0012799034593626857,
          "train_metric": 0.00016621888789813966,
          "validation_metric": 0.0001289071951759979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.406138102998739
        },
        {
          "epoch": 744,
          "updates": 744,
          "samples_seen": 761856,
          "online_loss": 0.0016503664664924145,
          "train_loss": 0.0016469677211716771,
          "validation_loss": 0.001275191898457706,
          "train_metric": 0.00016587655409239233,
          "validation_metric": 0.00012843265722040087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.42800872999942
        },
        {
          "epoch": 745,
          "updates": 745,
          "samples_seen": 762880,
          "online_loss": 0.0016469690017402172,
          "train_loss": 0.001643802272155881,
          "validation_loss": 0.001285863807424903,
          "train_metric": 0.00016555775073356926,
          "validation_metric": 0.0001295074907829985,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.446577455999432
        },
        {
          "epoch": 746,
          "updates": 746,
          "samples_seen": 763904,
          "online_loss": 0.001643802272155881,
          "train_loss": 0.0016405193600803614,
          "validation_loss": 0.001264053164049983,
          "train_metric": 0.00016522708756383508,
          "validation_metric": 0.00012731079186778516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.465366598999026
        },
        {
          "epoch": 747,
          "updates": 747,
          "samples_seen": 764928,
          "online_loss": 0.0016405193600803614,
          "train_loss": 0.001636990113183856,
          "validation_loss": 0.0012805832084268332,
          "train_metric": 0.00016487167158629745,
          "validation_metric": 0.00012897566193714738,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.483348495999962
        },
        {
          "epoch": 748,
          "updates": 748,
          "samples_seen": 765952,
          "online_loss": 0.0016369902295991778,
          "train_loss": 0.001633313950151205,
          "validation_loss": 0.0012624443043023348,
          "train_metric": 0.00016450138355139643,
          "validation_metric": 0.00012714877084363252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.501974333999897
        },
        {
          "epoch": 749,
          "updates": 749,
          "samples_seen": 766976,
          "online_loss": 0.0016333134844899178,
          "train_loss": 0.0016297274269163609,
          "validation_loss": 0.0012681400403380394,
          "train_metric": 0.0001641401759115979,
          "validation_metric": 0.00012772240734193474,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.519705631999386
        },
        {
          "epoch": 750,
          "updates": 750,
          "samples_seen": 768000,
          "online_loss": 0.0016297274269163609,
          "train_loss": 0.0016263379948213696,
          "validation_loss": 0.0012658312916755676,
          "train_metric": 0.00016379878798034042,
          "validation_metric": 0.00012748988228850067,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.527552793999348
        },
        {
          "epoch": 751,
          "updates": 751,
          "samples_seen": 769024,
          "online_loss": 0.0016263379948213696,
          "train_loss": 0.0016230909386649728,
          "validation_loss": 0.0012547969818115234,
          "train_metric": 0.0001634717482374981,
          "validation_metric": 0.00012637856707442552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.536199583999405
        },
        {
          "epoch": 752,
          "updates": 752,
          "samples_seen": 770048,
          "online_loss": 0.001623090822249651,
          "train_loss": 0.001619837828911841,
          "validation_loss": 0.001265494036488235,
          "train_metric": 0.00016314411186613142,
          "validation_metric": 0.00012745591811835766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.544159540999317
        },
        {
          "epoch": 753,
          "updates": 753,
          "samples_seen": 771072,
          "online_loss": 0.001619837712496519,
          "train_loss": 0.0016164765693247318,
          "validation_loss": 0.001248164800927043,
          "train_metric": 0.00016280560521408916,
          "validation_metric": 0.0001257105905096978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.552671275998364
        },
        {
          "epoch": 754,
          "updates": 754,
          "samples_seen": 772096,
          "online_loss": 0.0016164765693247318,
          "train_loss": 0.0016130245057865977,
          "validation_loss": 0.0012590683763846755,
          "train_metric": 0.00016245791630353779,
          "validation_metric": 0.0001268087507924065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.560727489999408
        },
        {
          "epoch": 755,
          "updates": 755,
          "samples_seen": 773120,
          "online_loss": 0.001613024272955954,
          "train_loss": 0.001609556726180017,
          "validation_loss": 0.0012464505853131413,
          "train_metric": 0.0001621086848899722,
          "validation_metric": 0.00012553793203551322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.569365596998978
        },
        {
          "epoch": 756,
          "updates": 756,
          "samples_seen": 774144,
          "online_loss": 0.0016095568425953388,
          "train_loss": 0.0016061646165326238,
          "validation_loss": 0.0012479325523599982,
          "train_metric": 0.00016176700592041016,
          "validation_metric": 0.00012568717647809535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.577464897000027
        },
        {
          "epoch": 757,
          "updates": 757,
          "samples_seen": 775168,
          "online_loss": 0.001606164500117302,
          "train_loss": 0.001602867734618485,
          "validation_loss": 0.0012471671216189861,
          "train_metric": 0.0001614349748706445,
          "validation_metric": 0.00012561010953504592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.585530426998957
        },
        {
          "epoch": 758,
          "updates": 758,
          "samples_seen": 776192,
          "online_loss": 0.0016028678510338068,
          "train_loss": 0.0015996366273611784,
          "validation_loss": 0.0012383281718939543,
          "train_metric": 0.00016110955039039254,
          "validation_metric": 0.00012471988156903535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.59417659799874
        },
        {
          "epoch": 759,
          "updates": 759,
          "samples_seen": 777216,
          "online_loss": 0.001599636161699891,
          "train_loss": 0.0015964019112288952,
          "validation_loss": 0.0012452347436919808,
          "train_metric": 0.00016078374756034464,
          "validation_metric": 0.00012541547766886652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.601867603998471
        },
        {
          "epoch": 760,
          "updates": 760,
          "samples_seen": 778240,
          "online_loss": 0.0015964019112288952,
          "train_loss": 0.0015931327361613512,
          "validation_loss": 0.0012314766645431519,
          "train_metric": 0.00016045451047830284,
          "validation_metric": 0.00012402981519699097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.610854470998675
        },
        {
          "epoch": 761,
          "updates": 761,
          "samples_seen": 779264,
          "online_loss": 0.001593132852576673,
          "train_loss": 0.001589822699315846,
          "validation_loss": 0.001239642035216093,
          "train_metric": 0.00016012112610042095,
          "validation_metric": 0.00012485218758229166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.618621325998902
        },
        {
          "epoch": 762,
          "updates": 762,
          "samples_seen": 780288,
          "online_loss": 0.0015898228157311678,
          "train_loss": 0.0015865013701841235,
          "validation_loss": 0.0012286114506423473,
          "train_metric": 0.00015978662122506648,
          "validation_metric": 0.00012374123616609722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.630438199999844
        },
        {
          "epoch": 763,
          "updates": 763,
          "samples_seen": 781312,
          "online_loss": 0.0015865016030147672,
          "train_loss": 0.0015832027420401573,
          "validation_loss": 0.001231284230016172,
          "train_metric": 0.00015945438644848764,
          "validation_metric": 0.000124010446597822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.645582563998687
        },
        {
          "epoch": 764,
          "updates": 764,
          "samples_seen": 782336,
          "online_loss": 0.0015832027420401573,
          "train_loss": 0.0015799511456862092,
          "validation_loss": 0.0012266729027032852,
          "train_metric": 0.00015912689559627324,
          "validation_metric": 0.00012354599311947823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.664026521999403
        },
        {
          "epoch": 765,
          "updates": 765,
          "samples_seen": 783360,
          "online_loss": 0.0015799510292708874,
          "train_loss": 0.001576735987327993,
          "validation_loss": 0.0012227788101881742,
          "train_metric": 0.00015880304272286594,
          "validation_metric": 0.00012315378990024328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.682709142998647
        },
        {
          "epoch": 766,
          "updates": 766,
          "samples_seen": 784384,
          "online_loss": 0.001576735987327993,
          "train_loss": 0.0015735552879050374,
          "validation_loss": 0.0012248523999005556,
          "train_metric": 0.00015848272596485913,
          "validation_metric": 0.0001233626389876008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.70117132499945
        },
        {
          "epoch": 767,
          "updates": 767,
          "samples_seen": 785408,
          "online_loss": 0.0015735552879050374,
          "train_loss": 0.0015703761018812656,
          "validation_loss": 0.0012157168239355087,
          "train_metric": 0.00015816254017408937,
          "validation_metric": 0.0001224425359396264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.72002178799994
        },
        {
          "epoch": 768,
          "updates": 768,
          "samples_seen": 786432,
          "online_loss": 0.0015703761018812656,
          "train_loss": 0.0015672006411477923,
          "validation_loss": 0.0012207994004711509,
          "train_metric": 0.00015784268907736987,
          "validation_metric": 0.00012295442866161466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.85626032899927
        },
        {
          "epoch": 769,
          "updates": 769,
          "samples_seen": 787456,
          "online_loss": 0.0015672005247324705,
          "train_loss": 0.0015640180790796876,
          "validation_loss": 0.0012104035122320056,
          "train_metric": 0.00015752216859254986,
          "validation_metric": 0.00012190741836093366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.864623049999864
        },
        {
          "epoch": 770,
          "updates": 770,
          "samples_seen": 788480,
          "online_loss": 0.0015640180790796876,
          "train_loss": 0.001560828648507595,
          "validation_loss": 0.001215440919622779,
          "train_metric": 0.00015720093506388366,
          "validation_metric": 0.00012241475633345544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.87240645699967
        },
        {
          "epoch": 771,
          "updates": 771,
          "samples_seen": 789504,
          "online_loss": 0.0015608287649229169,
          "train_loss": 0.0015576491132378578,
          "validation_loss": 0.0012065785704180598,
          "train_metric": 0.00015688072016928345,
          "validation_metric": 0.0001215221855090931,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.880620218998956
        },
        {
          "epoch": 772,
          "updates": 772,
          "samples_seen": 790528,
          "online_loss": 0.0015576491132378578,
          "train_loss": 0.0015544816851615906,
          "validation_loss": 0.0012086748611181974,
          "train_metric": 0.0001565617130836472,
          "validation_metric": 0.00012173330469522625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.888515265000024
        },
        {
          "epoch": 773,
          "updates": 773,
          "samples_seen": 791552,
          "online_loss": 0.0015544816851615906,
          "train_loss": 0.0015513281105086207,
          "validation_loss": 0.0012032999657094479,
          "train_metric": 0.0001562440738780424,
          "validation_metric": 0.000121191973448731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.897066341000027
        },
        {
          "epoch": 774,
          "updates": 774,
          "samples_seen": 792576,
          "online_loss": 0.001551327994093299,
          "train_loss": 0.0015481999143958092,
          "validation_loss": 0.0012021218426525593,
          "train_metric": 0.00015592903946526349,
          "validation_metric": 0.0001210733171319589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.905768406999414
        },
        {
          "epoch": 775,
          "updates": 775,
          "samples_seen": 793600,
          "online_loss": 0.0015481999143958092,
          "train_loss": 0.0015450895298272371,
          "validation_loss": 0.001200182014144957,
          "train_metric": 0.0001556157658342272,
          "validation_metric": 0.00012087794311810285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.914725665999867
        },
        {
          "epoch": 776,
          "updates": 776,
          "samples_seen": 794624,
          "online_loss": 0.0015450895298272371,
          "train_loss": 0.001541991252452135,
          "validation_loss": 0.0011956248199567199,
          "train_metric": 0.0001553037145640701,
          "validation_metric": 0.00012041895388392732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.925093515999833
        },
        {
          "epoch": 777,
          "updates": 777,
          "samples_seen": 795648,
          "online_loss": 0.0015419911360368133,
          "train_loss": 0.0015389093896374106,
          "validation_loss": 0.001196538214571774,
          "train_metric": 0.0001549933076603338,
          "validation_metric": 0.00012051093654008582,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.934700760999476
        },
        {
          "epoch": 778,
          "updates": 778,
          "samples_seen": 796672,
          "online_loss": 0.0015389093896374106,
          "train_loss": 0.0015358389355242252,
          "validation_loss": 0.0011899422388523817,
          "train_metric": 0.00015468407946173102,
          "validation_metric": 0.00011984661978203803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.948806101998343
        },
        {
          "epoch": 779,
          "updates": 779,
          "samples_seen": 797696,
          "online_loss": 0.0015358389355242252,
          "train_loss": 0.0015327746514230967,
          "validation_loss": 0.0011925437720492482,
          "train_metric": 0.0001543754624435678,
          "validation_metric": 0.00012010864156764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.976974492999943
        },
        {
          "epoch": 780,
          "updates": 780,
          "samples_seen": 798720,
          "online_loss": 0.0015327747678384185,
          "train_loss": 0.0015297232894226909,
          "validation_loss": 0.001184496097266674,
          "train_metric": 0.0001540681259939447,
          "validation_metric": 0.00011929812171729282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.041715625999132
        },
        {
          "epoch": 781,
          "updates": 781,
          "samples_seen": 799744,
          "online_loss": 0.001529723173007369,
          "train_loss": 0.0015266790287569165,
          "validation_loss": 0.0011880409438163042,
          "train_metric": 0.00015376153169199824,
          "validation_metric": 0.00011965513112954795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.100505255999451
        },
        {
          "epoch": 782,
          "updates": 782,
          "samples_seen": 800768,
          "online_loss": 0.0015266791451722383,
          "train_loss": 0.0015236434992402792,
          "validation_loss": 0.0011795946629717946,
          "train_metric": 0.00015345579595305026,
          "validation_metric": 0.00011880445526912808,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.142868019998787
        },
        {
          "epoch": 783,
          "updates": 783,
          "samples_seen": 801792,
          "online_loss": 0.0015236434992402792,
          "train_loss": 0.0015206180978566408,
          "validation_loss": 0.001183528220281005,
          "train_metric": 0.00015315109340008348,
          "validation_metric": 0.00011920061660930514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.169677959998808
        },
        {
          "epoch": 784,
          "updates": 784,
          "samples_seen": 802816,
          "online_loss": 0.0015206180978566408,
          "train_loss": 0.001517601776868105,
          "validation_loss": 0.0011746439849957824,
          "train_metric": 0.00015284727851394564,
          "validation_metric": 0.00011830584116978571,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.225887110998883
        },
        {
          "epoch": 785,
          "updates": 785,
          "samples_seen": 803840,
          "online_loss": 0.001517601776868105,
          "train_loss": 0.0015146021032705903,
          "validation_loss": 0.0011789336567744613,
          "train_metric": 0.00015254516620188951,
          "validation_metric": 0.00011873789480887353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.28215953399922
        },
        {
          "epoch": 786,
          "updates": 786,
          "samples_seen": 804864,
          "online_loss": 0.0015146019868552685,
          "train_loss": 0.0015116101130843163,
          "validation_loss": 0.0011698086746037006,
          "train_metric": 0.0001522438251413405,
          "validation_metric": 0.00011781883949879557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.343313666999165
        },
        {
          "epoch": 787,
          "updates": 787,
          "samples_seen": 805888,
          "online_loss": 0.0015116099966689944,
          "train_loss": 0.001508634421043098,
          "validation_loss": 0.0011746673844754696,
          "train_metric": 0.00015194412844721228,
          "validation_metric": 0.0001183081985800527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.379412268999658
        },
        {
          "epoch": 788,
          "updates": 788,
          "samples_seen": 806912,
          "online_loss": 0.001508634421043098,
          "train_loss": 0.0015056702541187406,
          "validation_loss": 0.0011645962949842215,
          "train_metric": 0.0001516455813543871,
          "validation_metric": 0.00011729387915693223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.406778866999957
        },
        {
          "epoch": 789,
          "updates": 789,
          "samples_seen": 807936,
          "online_loss": 0.0015056701377034187,
          "train_loss": 0.0015027258777990937,
          "validation_loss": 0.001170670846477151,
          "train_metric": 0.00015134902787394822,
          "validation_metric": 0.00011790567805292085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.452655419998337
        },
        {
          "epoch": 790,
          "updates": 790,
          "samples_seen": 808960,
          "online_loss": 0.0015027258777990937,
          "train_loss": 0.0014998003607615829,
          "validation_loss": 0.0011591004440560937,
          "train_metric": 0.00015105438069440424,
          "validation_metric": 0.00011674033885356039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.463295905999985
        },
        {
          "epoch": 791,
          "updates": 791,
          "samples_seen": 809984,
          "online_loss": 0.0014998003607615829,
          "train_loss": 0.001496903016231954,
          "validation_loss": 0.0011674410197883844,
          "train_metric": 0.00015076258569024503,
          "validation_metric": 0.00011758038453990594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.476997531999586
        },
        {
          "epoch": 792,
          "updates": 792,
          "samples_seen": 811008,
          "online_loss": 0.0014969028998166323,
          "train_loss": 0.001494038151577115,
          "validation_loss": 0.001152791897766292,
          "train_metric": 0.00015047407941892743,
          "validation_metric": 0.00011610498040681705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.492648289999124
        },
        {
          "epoch": 793,
          "updates": 793,
          "samples_seen": 812032,
          "online_loss": 0.0014940383844077587,
          "train_loss": 0.0014912211336195469,
          "validation_loss": 0.0011652546236291528,
          "train_metric": 0.00015019031707197428,
          "validation_metric": 0.00011736017040675506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.502027221999015
        },
        {
          "epoch": 794,
          "updates": 794,
          "samples_seen": 813056,
          "online_loss": 0.0014912211336195469,
          "train_loss": 0.001488482696004212,
          "validation_loss": 0.001145426300354302,
          "train_metric": 0.00014991451462265104,
          "validation_metric": 0.00011536314559634775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.51877186299862
        },
        {
          "epoch": 795,
          "updates": 795,
          "samples_seen": 814080,
          "online_loss": 0.00148848257958889,
          "train_loss": 0.0014858638169243932,
          "validation_loss": 0.0011651689419522882,
          "train_metric": 0.00014965079026296735,
          "validation_metric": 0.00011735154112102464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.532721166999181
        },
        {
          "epoch": 796,
          "updates": 796,
          "samples_seen": 815104,
          "online_loss": 0.001485863933339715,
          "train_loss": 0.001483445637859404,
          "validation_loss": 0.0011359886266291142,
          "train_metric": 0.00014940720575395972,
          "validation_metric": 0.00011441262176958844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.547823317998336
        },
        {
          "epoch": 797,
          "updates": 797,
          "samples_seen": 816128,
          "online_loss": 0.0014834454050287604,
          "train_loss": 0.0014813641319051385,
          "validation_loss": 0.0011688442900776863,
          "train_metric": 0.0001491975854150951,
          "validation_metric": 0.00011772171274060383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.55948270799854
        },
        {
          "epoch": 798,
          "updates": 798,
          "samples_seen": 817152,
          "online_loss": 0.0014813642483204603,
          "train_loss": 0.0014799063792452216,
          "validation_loss": 0.0011234190315008163,
          "train_metric": 0.00014905074203852564,
          "validation_metric": 0.00011314664880046621,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.569773159999386
        },
        {
          "epoch": 799,
          "updates": 799,
          "samples_seen": 818176,
          "online_loss": 0.0014799062628298998,
          "train_loss": 0.0014794613234698772,
          "validation_loss": 0.0011804585810750723,
          "train_metric": 0.00014900592213962227,
          "validation_metric": 0.00011889146117027849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.584409378998316
        },
        {
          "epoch": 800,
          "updates": 800,
          "samples_seen": 819200,
          "online_loss": 0.0014794613234698772,
          "train_loss": 0.00148109532892704,
          "validation_loss": 0.0011061812983825803,
          "train_metric": 0.0001491705043008551,
          "validation_metric": 0.00011141053983010352,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.598800007999671
        },
        {
          "epoch": 801,
          "updates": 801,
          "samples_seen": 820224,
          "online_loss": 0.0014810955617576838,
          "train_loss": 0.001485896296799183,
          "validation_loss": 0.0012093705590814352,
          "train_metric": 0.00014965402078814805,
          "validation_metric": 0.00012180336489109322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.620509504999063
        },
        {
          "epoch": 802,
          "updates": 802,
          "samples_seen": 821248,
          "online_loss": 0.001485896180383861,
          "train_loss": 0.0014978121034801006,
          "validation_loss": 0.0010857253801077604,
          "train_metric": 0.00015085414634086192,
          "validation_metric": 0.00010935028694802895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.630874705999304
        },
        {
          "epoch": 803,
          "updates": 803,
          "samples_seen": 822272,
          "online_loss": 0.0014978119870647788,
          "train_loss": 0.0015186696546152234,
          "validation_loss": 0.001277600647881627,
          "train_metric": 0.0001529548317193985,
          "validation_metric": 0.00012867525219917297,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.639794548998907
        },
        {
          "epoch": 804,
          "updates": 804,
          "samples_seen": 823296,
          "online_loss": 0.0015186695381999016,
          "train_loss": 0.0015614491421729326,
          "validation_loss": 0.0010763505706563592,
          "train_metric": 0.0001572634355397895,
          "validation_metric": 0.00010840607137652114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.655866012999468
        },
        {
          "epoch": 805,
          "updates": 805,
          "samples_seen": 824320,
          "online_loss": 0.0015614491421729326,
          "train_loss": 0.0016162056708708405,
          "validation_loss": 0.0014191673835739493,
          "train_metric": 0.00016277832037303597,
          "validation_metric": 0.00014293333515524864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.665562175998275
        },
        {
          "epoch": 806,
          "updates": 806,
          "samples_seen": 825344,
          "online_loss": 0.0016162055544555187,
          "train_loss": 0.0017084054416045547,
          "validation_loss": 0.0011112913489341736,
          "train_metric": 0.00017206434858962893,
          "validation_metric": 0.00011192519741598517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.675682981998762
        },
        {
          "epoch": 807,
          "updates": 807,
          "samples_seen": 826368,
          "online_loss": 0.0017084055580198765,
          "train_loss": 0.0017405794933438301,
          "validation_loss": 0.001569105195812881,
          "train_metric": 0.00017530479817651212,
          "validation_metric": 0.0001580345124239102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.68514654199862
        },
        {
          "epoch": 808,
          "updates": 808,
          "samples_seen": 827392,
          "online_loss": 0.001740579609759152,
          "train_loss": 0.001744047156535089,
          "validation_loss": 0.0011210466036573052,
          "train_metric": 0.00017565405869390815,
          "validation_metric": 0.00011290771362837404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.816774099999748
        },
        {
          "epoch": 809,
          "updates": 809,
          "samples_seen": 828416,
          "online_loss": 0.0017440472729504108,
          "train_loss": 0.0016020801849663258,
          "validation_loss": 0.0014003480318933725,
          "train_metric": 0.00016135563782881945,
          "validation_metric": 0.00014103793364483863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.82918346599945
        },
        {
          "epoch": 810,
          "updates": 810,
          "samples_seen": 829440,
          "online_loss": 0.001602079952135682,
          "train_loss": 0.0014773337170481682,
          "validation_loss": 0.001067213248461485,
          "train_metric": 0.00014879163063596934,
          "validation_metric": 0.00010748580825747922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.842007853998439
        },
        {
          "epoch": 811,
          "updates": 811,
          "samples_seen": 830464,
          "online_loss": 0.0014773337170481682,
          "train_loss": 0.0014451323077082634,
          "validation_loss": 0.0010952511802315712,
          "train_metric": 0.00014554844528902322,
          "validation_metric": 0.00011030968744307756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.850696668998353
        },
        {
          "epoch": 812,
          "updates": 812,
          "samples_seen": 831488,
          "online_loss": 0.0014451323077082634,
          "train_loss": 0.0015091150999069214,
          "validation_loss": 0.0012791913468390703,
          "train_metric": 0.0001519925135653466,
          "validation_metric": 0.0001288354687858373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.861073889998806
        },
        {
          "epoch": 813,
          "updates": 813,
          "samples_seen": 832512,
          "online_loss": 0.0015091149834915996,
          "train_loss": 0.0015855152159929276,
          "validation_loss": 0.0010695067467167974,
          "train_metric": 0.00015968730440363288,
          "validation_metric": 0.00010771679080789909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.869302713999787
        },
        {
          "epoch": 814,
          "updates": 814,
          "samples_seen": 833536,
          "online_loss": 0.0015855153324082494,
          "train_loss": 0.0015630966518074274,
          "validation_loss": 0.0013586238492280245,
          "train_metric": 0.00015742937102913857,
          "validation_metric": 0.00013683561701327562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.88062912299938
        },
        {
          "epoch": 815,
          "updates": 815,
          "samples_seen": 834560,
          "online_loss": 0.0015630967682227492,
          "train_loss": 0.001488621928729117,
          "validation_loss": 0.001054943772032857,
          "train_metric": 0.00014992854266893119,
          "validation_metric": 0.00010625008144415915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.892714149998938
        },
        {
          "epoch": 816,
          "updates": 816,
          "samples_seen": 835584,
          "online_loss": 0.001488621812313795,
          "train_loss": 0.001431002514436841,
          "validation_loss": 0.0011272981064394116,
          "train_metric": 0.00014412534073926508,
          "validation_metric": 0.00011353733134455979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.901316863999455
        },
        {
          "epoch": 817,
          "updates": 817,
          "samples_seen": 836608,
          "online_loss": 0.0014310023980215192,
          "train_loss": 0.0014487472362816334,
          "validation_loss": 0.0011827102862298489,
          "train_metric": 0.0001459125051042065,
          "validation_metric": 0.00011911825276911259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.913990273998934
        },
        {
          "epoch": 818,
          "updates": 818,
          "samples_seen": 837632,
          "online_loss": 0.0014487471198663116,
          "train_loss": 0.0014990696217864752,
          "validation_loss": 0.001047329162247479,
          "train_metric": 0.00015098082076292485,
          "validation_metric": 0.00010548315185587853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.926165907998438
        },
        {
          "epoch": 819,
          "updates": 819,
          "samples_seen": 838656,
          "online_loss": 0.0014990696217864752,
          "train_loss": 0.0015007060719653964,
          "validation_loss": 0.0012767288135364652,
          "train_metric": 0.00015114560665097088,
          "validation_metric": 0.00012858746049460024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.935286219999398
        },
        {
          "epoch": 820,
          "updates": 820,
          "samples_seen": 839680,
          "online_loss": 0.001500705722719431,
          "train_loss": 0.0014591822400689125,
          "validation_loss": 0.0010501670185476542,
          "train_metric": 0.0001469635171815753,
          "validation_metric": 0.00010576895147096366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.948713438998311
        },
        {
          "epoch": 821,
          "updates": 821,
          "samples_seen": 840704,
          "online_loss": 0.0014591824728995562,
          "train_loss": 0.0014194226823747158,
          "validation_loss": 0.0011232407996430993,
          "train_metric": 0.0001429590629413724,
          "validation_metric": 0.00011312869901303202,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.960597983999833
        },
        {
          "epoch": 822,
          "updates": 822,
          "samples_seen": 841728,
          "online_loss": 0.0014194226823747158,
          "train_loss": 0.0014254073612391949,
          "validation_loss": 0.0011504247086122632,
          "train_metric": 0.0001435618323739618,
          "validation_metric": 0.00011586656182771549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.969568477999928
        },
        {
          "epoch": 823,
          "updates": 823,
          "samples_seen": 842752,
          "online_loss": 0.0014254073612391949,
          "train_loss": 0.0014549378538504243,
          "validation_loss": 0.0010462969075888395,
          "train_metric": 0.00014653602556791157,
          "validation_metric": 0.00010537918569752946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.985042779999276
        },
        {
          "epoch": 824,
          "updates": 824,
          "samples_seen": 843776,
          "online_loss": 0.0014549382030963898,
          "train_loss": 0.0014582976000383496,
          "validation_loss": 0.001222836785018444,
          "train_metric": 0.00014687440125271678,
          "validation_metric": 0.00012315962521824986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.99867273899872
        },
        {
          "epoch": 825,
          "updates": 825,
          "samples_seen": 844800,
          "online_loss": 0.0014582978328689933,
          "train_loss": 0.0014333132421597838,
          "validation_loss": 0.001047078869305551,
          "train_metric": 0.00014435805496759713,
          "validation_metric": 0.00010545796249061823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.013586728999144
        },
        {
          "epoch": 826,
          "updates": 826,
          "samples_seen": 845824,
          "online_loss": 0.0014333132421597838,
          "train_loss": 0.0014072676422074437,
          "validation_loss": 0.001111502991989255,
          "train_metric": 0.00014173483941704035,
          "validation_metric": 0.00011194650869583711,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.025739497999894
        },
        {
          "epoch": 827,
          "updates": 827,
          "samples_seen": 846848,
          "online_loss": 0.0014072676422074437,
          "train_loss": 0.0014082829002290964,
          "validation_loss": 0.0011234843404963613,
          "train_metric": 0.00014183708117343485,
          "validation_metric": 0.00011315324081806466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.034139997998864
        },
        {
          "epoch": 828,
          "updates": 828,
          "samples_seen": 847872,
          "online_loss": 0.0014082830166444182,
          "train_loss": 0.001424905494786799,
          "validation_loss": 0.0010419454192742705,
          "train_metric": 0.000143511249916628,
          "validation_metric": 0.00010494093294255435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.048890777999986
        },
        {
          "epoch": 829,
          "updates": 829,
          "samples_seen": 848896,
          "online_loss": 0.0014249053783714771,
          "train_loss": 0.0014277255395427346,
          "validation_loss": 0.0011788597330451012,
          "train_metric": 0.00014379527419805527,
          "validation_metric": 0.00011873044422827661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.060046368998883
        },
        {
          "epoch": 830,
          "updates": 830,
          "samples_seen": 849920,
          "online_loss": 0.001427725306712091,
          "train_loss": 0.0014126955065876245,
          "validation_loss": 0.0010447119129821658,
          "train_metric": 0.0001422815112164244,
          "validation_metric": 0.00010521955118747428,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.068261792999692
        },
        {
          "epoch": 831,
          "updates": 831,
          "samples_seen": 850944,
          "online_loss": 0.0014126955065876245,
          "train_loss": 0.001395409693941474,
          "validation_loss": 0.001102758920751512,
          "train_metric": 0.00014054054918233305,
          "validation_metric": 0.00011106583406217396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.08073537799828
        },
        {
          "epoch": 832,
          "updates": 832,
          "samples_seen": 851968,
          "online_loss": 0.001395409693941474,
          "train_loss": 0.0013934188755229115,
          "validation_loss": 0.0011024518171325326,
          "train_metric": 0.00014034003834240139,
          "validation_metric": 0.0001110349185182713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.091980071998478
        },
        {
          "epoch": 833,
          "updates": 833,
          "samples_seen": 852992,
          "online_loss": 0.0013934188755229115,
          "train_loss": 0.0014020642265677452,
          "validation_loss": 0.001042062765918672,
          "train_metric": 0.00014121078129392117,
          "validation_metric": 0.00010495274182176217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.100687371999811
        },
        {
          "epoch": 834,
          "updates": 834,
          "samples_seen": 854016,
          "online_loss": 0.001402064342983067,
          "train_loss": 0.0014046953292563558,
          "validation_loss": 0.0011445862473919988,
          "train_metric": 0.00014147577167022973,
          "validation_metric": 0.00011527854803716764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.113028348998341
        },
        {
          "epoch": 835,
          "updates": 835,
          "samples_seen": 855040,
          "online_loss": 0.0014046953292563558,
          "train_loss": 0.0013962844386696815,
          "validation_loss": 0.001038319431245327,
          "train_metric": 0.00014062867558095604,
          "validation_metric": 0.00010457572352606803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.124950083998556
        },
        {
          "epoch": 836,
          "updates": 836,
          "samples_seen": 856064,
          "online_loss": 0.0013962843222543597,
          "train_loss": 0.001384301926009357,
          "validation_loss": 0.0010932838777080178,
          "train_metric": 0.0001394217979395762,
          "validation_metric": 0.00011011154856532812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.134195560998705
        },
        {
          "epoch": 837,
          "updates": 837,
          "samples_seen": 857088,
          "online_loss": 0.0013843018095940351,
          "train_loss": 0.0013800878077745438,
          "validation_loss": 0.0010789736406877637,
          "train_metric": 0.00013899740588385612,
          "validation_metric": 0.00010867027594940737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.147224242999073
        },
        {
          "epoch": 838,
          "updates": 838,
          "samples_seen": 858112,
          "online_loss": 0.0013800878077745438,
          "train_loss": 0.00138350750785321,
          "validation_loss": 0.0010409834794700146,
          "train_metric": 0.00013934180606156588,
          "validation_metric": 0.00010484403173904866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.160864045999915
        },
        {
          "epoch": 839,
          "updates": 839,
          "samples_seen": 859136,
          "online_loss": 0.0013835076242685318,
          "train_loss": 0.001385822775773704,
          "validation_loss": 0.001116677769459784,
          "train_metric": 0.00013957498595118523,
          "validation_metric": 0.00011246769281569868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.170036779998554
        },
        {
          "epoch": 840,
          "updates": 840,
          "samples_seen": 860160,
          "online_loss": 0.0013858228921890259,
          "train_loss": 0.001381853362545371,
          "validation_loss": 0.001034015673212707,
          "train_metric": 0.00013917523028794676,
          "validation_metric": 0.00010414225835120305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.185022790999938
        },
        {
          "epoch": 841,
          "updates": 841,
          "samples_seen": 861184,
          "online_loss": 0.001381853362545371,
          "train_loss": 0.0013737144181504846,
          "validation_loss": 0.0010888216784223914,
          "train_metric": 0.00013835547724738717,
          "validation_metric": 0.00010966213449137285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.198017131999222
        },
        {
          "epoch": 842,
          "updates": 842,
          "samples_seen": 862208,
          "online_loss": 0.0013737145345658064,
          "train_loss": 0.00136827421374619,
          "validation_loss": 0.0010597860673442483,
          "train_metric": 0.00013780758308712393,
          "validation_metric": 0.00010673775977920741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.211550621999777
        },
        {
          "epoch": 843,
          "updates": 843,
          "samples_seen": 863232,
          "online_loss": 0.0013682739809155464,
          "train_loss": 0.0013678808463737369,
          "validation_loss": 0.001043570227921009,
          "train_metric": 0.0001377679582219571,
          "validation_metric": 0.00010510457650525495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.225436381999316
        },
        {
          "epoch": 844,
          "updates": 844,
          "samples_seen": 864256,
          "online_loss": 0.001367880729958415,
          "train_loss": 0.0013691758504137397,
          "validation_loss": 0.001091290032491088,
          "train_metric": 0.00013789838703814894,
          "validation_metric": 0.00010991073941113427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.235444735999408
        },
        {
          "epoch": 845,
          "updates": 845,
          "samples_seen": 865280,
          "online_loss": 0.0013691757339984179,
          "train_loss": 0.0013679348630830646,
          "validation_loss": 0.0010287367040291429,
          "train_metric": 0.00013777338608633727,
          "validation_metric": 0.00010361059685237706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.250085896999735
        },
        {
          "epoch": 846,
          "updates": 846,
          "samples_seen": 866304,
          "online_loss": 0.0013679349794983864,
          "train_loss": 0.0013631260953843594,
          "validation_loss": 0.0010821121977642179,
          "train_metric": 0.00013728905469179153,
          "validation_metric": 0.00010898637265199795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.26114373099881
        },
        {
          "epoch": 847,
          "updates": 847,
          "samples_seen": 867328,
          "online_loss": 0.0013631258625537157,
          "train_loss": 0.0013578807702288032,
          "validation_loss": 0.0010421142214909196,
          "train_metric": 0.00013676079106517136,
          "validation_metric": 0.00010495794413145632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.412795546999405
        },
        {
          "epoch": 848,
          "updates": 848,
          "samples_seen": 868352,
          "online_loss": 0.001357880886644125,
          "train_loss": 0.0013549389550462365,
          "validation_loss": 0.0010477855103090405,
          "train_metric": 0.00013646449951920658,
          "validation_metric": 0.00010552912863204256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.42084264299956
        },
        {
          "epoch": 849,
          "updates": 849,
          "samples_seen": 869376,
          "online_loss": 0.0013549396535381675,
          "train_loss": 0.0013543505920097232,
          "validation_loss": 0.0010679515544325113,
          "train_metric": 0.00013640524412039667,
          "validation_metric": 0.00010756016126833856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.428613662999851
        },
        {
          "epoch": 850,
          "updates": 850,
          "samples_seen": 870400,
          "online_loss": 0.001354350708425045,
          "train_loss": 0.0013539367355406284,
          "validation_loss": 0.001027473364956677,
          "train_metric": 0.00013636353833135217,
          "validation_metric": 0.00010348335490562022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.43720299099914
        },
        {
          "epoch": 851,
          "updates": 851,
          "samples_seen": 871424,
          "online_loss": 0.0013539366191253066,
          "train_loss": 0.0013517525512725115,
          "validation_loss": 0.0010729195782914758,
          "train_metric": 0.00013614355702884495,
          "validation_metric": 0.00010806052887346596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.445124611998835
        },
        {
          "epoch": 852,
          "updates": 852,
          "samples_seen": 872448,
          "online_loss": 0.0013517525512725115,
          "train_loss": 0.0013480433262884617,
          "validation_loss": 0.0010284252930432558,
          "train_metric": 0.00013576999481301755,
          "validation_metric": 0.0001035792229231447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.452654367998548
        },
        {
          "epoch": 853,
          "updates": 853,
          "samples_seen": 873472,
          "online_loss": 0.0013480434427037835,
          "train_loss": 0.0013442581985145807,
          "validation_loss": 0.0010511206928640604,
          "train_metric": 0.00013538877828978002,
          "validation_metric": 0.00010586502321530133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.46026361999975
        },
        {
          "epoch": 854,
          "updates": 854,
          "samples_seen": 874496,
          "online_loss": 0.0013442581985145807,
          "train_loss": 0.0013417263980954885,
          "validation_loss": 0.0010438533499836922,
          "train_metric": 0.0001351337559754029,
          "validation_metric": 0.00010513308370718732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.467973840999548
        },
        {
          "epoch": 855,
          "updates": 855,
          "samples_seen": 875520,
          "online_loss": 0.0013417262816801667,
          "train_loss": 0.001340391463600099,
          "validation_loss": 0.0010280914139002562,
          "train_metric": 0.00013499932538252324,
          "validation_metric": 0.00010354560799896717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.475665576999745
        },
        {
          "epoch": 856,
          "updates": 856,
          "samples_seen": 876544,
          "online_loss": 0.001340391463600099,
          "train_loss": 0.0013392014661803842,
          "validation_loss": 0.0010563019895926118,
          "train_metric": 0.00013487946125678718,
          "validation_metric": 0.00010638686217134818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.48312870799964
        },
        {
          "epoch": 857,
          "updates": 857,
          "samples_seen": 877568,
          "online_loss": 0.0013392013497650623,
          "train_loss": 0.0013372459216043353,
          "validation_loss": 0.0010198046220466495,
          "train_metric": 0.00013468250108417124,
          "validation_metric": 0.00010271097562508658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.491560387999925
        },
        {
          "epoch": 858,
          "updates": 858,
          "samples_seen": 878592,
          "online_loss": 0.0013372459216043353,
          "train_loss": 0.0013343989849090576,
          "validation_loss": 0.0010504358215257525,
          "train_metric": 0.00013439579925034195,
          "validation_metric": 0.00010579604713711888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.49910816099873
        },
        {
          "epoch": 859,
          "updates": 859,
          "samples_seen": 879616,
          "online_loss": 0.0013343991013243794,
          "train_loss": 0.0013313323725014925,
          "validation_loss": 0.0010247674072161317,
          "train_metric": 0.00013408692029770464,
          "validation_metric": 0.00010321081208530813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.50656889199854
        },
        {
          "epoch": 860,
          "updates": 860,
          "samples_seen": 880640,
          "online_loss": 0.0013313323725014925,
          "train_loss": 0.0013287311885505915,
          "validation_loss": 0.0010323271853849292,
          "train_metric": 0.00013382492761593312,
          "validation_metric": 0.00010397219739388674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.514550022999174
        },
        {
          "epoch": 861,
          "updates": 861,
          "samples_seen": 881664,
          "online_loss": 0.0013287310721352696,
          "train_loss": 0.0013267986942082644,
          "validation_loss": 0.0010353225516155362,
          "train_metric": 0.00013363029574975371,
          "validation_metric": 0.00010427388042444363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.52256284899886
        },
        {
          "epoch": 862,
          "updates": 862,
          "samples_seen": 882688,
          "online_loss": 0.0013267985777929425,
          "train_loss": 0.0013252061326056719,
          "validation_loss": 0.001017454662360251,
          "train_metric": 0.00013346990454010665,
          "validation_metric": 0.00010247431055177003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.531545766998534
        },
        {
          "epoch": 863,
          "updates": 863,
          "samples_seen": 883712,
          "online_loss": 0.0013252061326056719,
          "train_loss": 0.0013234477955847979,
          "validation_loss": 0.0010399807943031192,
          "train_metric": 0.00013329283683560789,
          "validation_metric": 0.00010474305599927902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.540729303998887
        },
        {
          "epoch": 864,
          "updates": 864,
          "samples_seen": 884736,
          "online_loss": 0.0013234479120001197,
          "train_loss": 0.0013212725752964616,
          "validation_loss": 0.0010124484542757273,
          "train_metric": 0.00013307372864801437,
          "validation_metric": 0.00010197010124102235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.549521298999025
        },
        {
          "epoch": 865,
          "updates": 865,
          "samples_seen": 885760,
          "online_loss": 0.0013212725752964616,
          "train_loss": 0.001318751135841012,
          "validation_loss": 0.0010334941325709224,
          "train_metric": 0.00013281978317536414,
          "validation_metric": 0.00010408974048914388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.557389688998228
        },
        {
          "epoch": 866,
          "updates": 866,
          "samples_seen": 886784,
          "online_loss": 0.0013187512522563338,
          "train_loss": 0.001316209090873599,
          "validation_loss": 0.0010158841032534838,
          "train_metric": 0.00013256375677883625,
          "validation_metric": 0.00010231611668132246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.564880626998274
        },
        {
          "epoch": 867,
          "updates": 867,
          "samples_seen": 887808,
          "online_loss": 0.001316209090873599,
          "train_loss": 0.0013138956855982542,
          "validation_loss": 0.001020844210870564,
          "train_metric": 0.0001323307806160301,
          "validation_metric": 0.00010281568393111229,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.574074856998777
        },
        {
          "epoch": 868,
          "updates": 868,
          "samples_seen": 888832,
          "online_loss": 0.0013138956855982542,
          "train_loss": 0.001311879837885499,
          "validation_loss": 0.0010220904368907213,
          "train_metric": 0.00013212772319093347,
          "validation_metric": 0.00010294120875187218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.582128078998721
        },
        {
          "epoch": 869,
          "updates": 869,
          "samples_seen": 889856,
          "online_loss": 0.0013118797214701772,
          "train_loss": 0.0013100365176796913,
          "validation_loss": 0.0010099181672558188,
          "train_metric": 0.00013194206985644996,
          "validation_metric": 0.00010171526082558557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.5909778239984
        },
        {
          "epoch": 870,
          "updates": 870,
          "samples_seen": 890880,
          "online_loss": 0.0013100364012643695,
          "train_loss": 0.0013081661891192198,
          "validation_loss": 0.0010246530873700976,
          "train_metric": 0.0001317537244176492,
          "validation_metric": 0.00010319929424440488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.599128139998356
        },
        {
          "epoch": 871,
          "updates": 871,
          "samples_seen": 891904,
          "online_loss": 0.0013081664219498634,
          "train_loss": 0.0013061384670436382,
          "validation_loss": 0.0010043809888884425,
          "train_metric": 0.00013154947373550385,
          "validation_metric": 0.00010115756595041603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.608129459998963
        },
        {
          "epoch": 872,
          "updates": 872,
          "samples_seen": 892928,
          "online_loss": 0.0013061384670436382,
          "train_loss": 0.0013039454352110624,
          "validation_loss": 0.0010206979932263494,
          "train_metric": 0.0001313286047661677,
          "validation_metric": 0.00010280095739290118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.615966374998607
        },
        {
          "epoch": 873,
          "updates": 873,
          "samples_seen": 893952,
          "online_loss": 0.0013039452023804188,
          "train_loss": 0.001301672775298357,
          "validation_loss": 0.0010042276699095964,
          "train_metric": 0.0001310997031396255,
          "validation_metric": 0.00010114212636835873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.624754605998532
        },
        {
          "epoch": 874,
          "updates": 874,
          "samples_seen": 894976,
          "online_loss": 0.001301672775298357,
          "train_loss": 0.0012994291028007865,
          "validation_loss": 0.0010124664986506104,
          "train_metric": 0.0001308737409999594,
          "validation_metric": 0.00010197191295446828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.632741642999463
        },
        {
          "epoch": 875,
          "updates": 875,
          "samples_seen": 896000,
          "online_loss": 0.0012994291028007865,
          "train_loss": 0.0012972964905202389,
          "validation_loss": 0.0010068059200420976,
          "train_metric": 0.00013065895473118871,
          "validation_metric": 0.00010140179074369371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.640376619998278
        },
        {
          "epoch": 876,
          "updates": 876,
          "samples_seen": 897024,
          "online_loss": 0.0012972967233508825,
          "train_loss": 0.0012952795950695872,
          "validation_loss": 0.0010036715539172292,
          "train_metric": 0.00013045579544268548,
          "validation_metric": 0.00010108610877068713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.64868940999986
        },
        {
          "epoch": 877,
          "updates": 877,
          "samples_seen": 898048,
          "online_loss": 0.0012952793622389436,
          "train_loss": 0.0012933282414451241,
          "validation_loss": 0.001008764491416514,
          "train_metric": 0.00013025925727561116,
          "validation_metric": 0.00010159904923057184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.656895378999252
        },
        {
          "epoch": 878,
          "updates": 878,
          "samples_seen": 899072,
          "online_loss": 0.0012933280086144805,
          "train_loss": 0.0012913960963487625,
          "validation_loss": 0.0009969488019123673,
          "train_metric": 0.00013006468361709267,
          "validation_metric": 0.00010040903725894168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.665079673999571
        },
        {
          "epoch": 879,
          "updates": 879,
          "samples_seen": 900096,
          "online_loss": 0.0012913958635181189,
          "train_loss": 0.0012894216924905777,
          "validation_loss": 0.0010078465566039085,
          "train_metric": 0.00012986580259166658,
          "validation_metric": 0.00010150663001695648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.673395762998553
        },
        {
          "epoch": 880,
          "updates": 880,
          "samples_seen": 901120,
          "online_loss": 0.001289421459659934,
          "train_loss": 0.001287394785322249,
          "validation_loss": 0.0009931100066751242,
          "train_metric": 0.0001296616974286735,
          "validation_metric": 0.00010002239287132397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.682011221999346
        },
        {
          "epoch": 881,
          "updates": 881,
          "samples_seen": 902144,
          "online_loss": 0.0012873949017375708,
          "train_loss": 0.0012853150255978107,
          "validation_loss": 0.0010039452463388443,
          "train_metric": 0.00012945222260896116,
          "validation_metric": 0.00010111368465004489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.690184859999135
        },
        {
          "epoch": 882,
          "updates": 882,
          "samples_seen": 903168,
          "online_loss": 0.0012853150255978107,
          "train_loss": 0.001283218851312995,
          "validation_loss": 0.0009917718125507236,
          "train_metric": 0.00012924110342282802,
          "validation_metric": 9.988761303247884e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.698307910999574
        },
        {
          "epoch": 883,
          "updates": 883,
          "samples_seen": 904192,
          "online_loss": 0.001283218851312995,
          "train_loss": 0.0012811340857297182,
          "validation_loss": 0.000998220988549292,
          "train_metric": 0.0001290311338379979,
          "validation_metric": 0.00010053714504465461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.706583093999143
        },
        {
          "epoch": 884,
          "updates": 884,
          "samples_seen": 905216,
          "online_loss": 0.0012811340857297182,
          "train_loss": 0.0012790770269930363,
          "validation_loss": 0.00099168939050287,
          "train_metric": 0.00012882397277280688,
          "validation_metric": 9.987931844079867e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.714781447999485
        },
        {
          "epoch": 885,
          "updates": 885,
          "samples_seen": 906240,
          "online_loss": 0.0012790770269930363,
          "train_loss": 0.0012770609464496374,
          "validation_loss": 0.0009921081364154816,
          "train_metric": 0.0001286208862438798,
          "validation_metric": 9.992148989113048e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.822432796998328
        },
        {
          "epoch": 886,
          "updates": 886,
          "samples_seen": 907264,
          "online_loss": 0.0012770608300343156,
          "train_loss": 0.0012750753667205572,
          "validation_loss": 0.0009915875270962715,
          "train_metric": 0.00012842091382481158,
          "validation_metric": 9.986905934056267e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.83098479399996
        },
        {
          "epoch": 887,
          "updates": 887,
          "samples_seen": 908288,
          "online_loss": 0.001275075483135879,
          "train_loss": 0.0012731155147776008,
          "validation_loss": 0.000986491795629263,
          "train_metric": 0.00012822354619856924,
          "validation_metric": 9.935584967024624e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.840209819998563
        },
        {
          "epoch": 888,
          "updates": 888,
          "samples_seen": 909312,
          "online_loss": 0.0012731155147776008,
          "train_loss": 0.0012711705639958382,
          "validation_loss": 0.0009906413033604622,
          "train_metric": 0.00012802763376384974,
          "validation_metric": 9.97737588477321e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.848508390999996
        },
        {
          "epoch": 889,
          "updates": 889,
          "samples_seen": 910336,
          "online_loss": 0.0012711705639958382,
          "train_loss": 0.0012692250311374664,
          "validation_loss": 0.0009819130646064878,
          "train_metric": 0.00012783169222529978,
          "validation_metric": 9.889467764878646e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.85734212099851
        },
        {
          "epoch": 890,
          "updates": 890,
          "samples_seen": 911360,
          "online_loss": 0.0012692250311374664,
          "train_loss": 0.0012672776356339455,
          "validation_loss": 0.000988751300610602,
          "train_metric": 0.00012763556151185185,
          "validation_metric": 9.958341252058744e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.865049105999788
        },
        {
          "epoch": 891,
          "updates": 891,
          "samples_seen": 912384,
          "online_loss": 0.0012672777520492673,
          "train_loss": 0.0012653219746425748,
          "validation_loss": 0.00097838812507689,
          "train_metric": 0.00012743858678732067,
          "validation_metric": 9.85396618489176e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.873869085999104
        },
        {
          "epoch": 892,
          "updates": 892,
          "samples_seen": 913408,
          "online_loss": 0.001265321858227253,
          "train_loss": 0.0012633617734536529,
          "validation_loss": 0.00098606722895056,
          "train_metric": 0.00012724116095341742,
          "validation_metric": 9.931308886734769e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.881787781998355
        },
        {
          "epoch": 893,
          "updates": 893,
          "samples_seen": 914432,
          "online_loss": 0.001263361657038331,
          "train_loss": 0.0012613990111276507,
          "validation_loss": 0.0009756077197380364,
          "train_metric": 0.00012704348773695529,
          "validation_metric": 9.825963934417814e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.890539551999609
        },
        {
          "epoch": 894,
          "updates": 894,
          "samples_seen": 915456,
          "online_loss": 0.001261398894712329,
          "train_loss": 0.0012594371801242232,
          "validation_loss": 0.0009828282054513693,
          "train_metric": 0.00012684590183198452,
          "validation_metric": 9.898684947984293e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.898410193998643
        },
        {
          "epoch": 895,
          "updates": 895,
          "samples_seen": 916480,
          "online_loss": 0.001259437296539545,
          "train_loss": 0.0012574788415804505,
          "validation_loss": 0.0009731058962643147,
          "train_metric": 0.00012664867972489446,
          "validation_metric": 9.800767293199897e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.906924324999636
        },
        {
          "epoch": 896,
          "updates": 896,
          "samples_seen": 917504,
          "online_loss": 0.0012574789579957724,
          "train_loss": 0.001255521667189896,
          "validation_loss": 0.0009793107165023685,
          "train_metric": 0.00012645153037738055,
          "validation_metric": 9.863259037956595e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.91484625799967
        },
        {
          "epoch": 897,
          "updates": 897,
          "samples_seen": 918528,
          "online_loss": 0.0012555219000205398,
          "train_loss": 0.0012535701971501112,
          "validation_loss": 0.0009706805576570332,
          "train_metric": 0.00012625499221030623,
          "validation_metric": 9.776339720701799e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.924666085998979
        },
        {
          "epoch": 898,
          "updates": 898,
          "samples_seen": 919552,
          "online_loss": 0.0012535701971501112,
          "train_loss": 0.001251629670150578,
          "validation_loss": 0.0009758500382304192,
          "train_metric": 0.00012605955998878926,
          "validation_metric": 9.828404290601611e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.932897243999832
        },
        {
          "epoch": 899,
          "updates": 899,
          "samples_seen": 920576,
          "online_loss": 0.001251629670150578,
          "train_loss": 0.0012496974086388946,
          "validation_loss": 0.000968201260548085,
          "train_metric": 0.0001258649572264403,
          "validation_metric": 9.75136790657416e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.94182702499893
        },
        {
          "epoch": 900,
          "updates": 900,
          "samples_seen": 921600,
          "online_loss": 0.0012496974086388946,
          "train_loss": 0.0012477701529860497,
          "validation_loss": 0.0009726286516524851,
          "train_metric": 0.0001256708346772939,
          "validation_metric": 9.795958612812683e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.95112181499826
        },
        {
          "epoch": 901,
          "updates": 901,
          "samples_seen": 922624,
          "online_loss": 0.0012477701529860497,
          "train_loss": 0.00124585279263556,
          "validation_loss": 0.0009655911126174033,
          "train_metric": 0.00012547773076221347,
          "validation_metric": 9.725079871714115e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.961155158998736
        },
        {
          "epoch": 902,
          "updates": 902,
          "samples_seen": 923648,
          "online_loss": 0.0012458529090508819,
          "train_loss": 0.001243940438143909,
          "validation_loss": 0.0009696908527985215,
          "train_metric": 0.00012528513616416603,
          "validation_metric": 9.766371658770368e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.969037368999125
        },
        {
          "epoch": 903,
          "updates": 903,
          "samples_seen": 924672,
          "online_loss": 0.0012439405545592308,
          "train_loss": 0.0012420356506481767,
          "validation_loss": 0.0009626745013520122,
          "train_metric": 0.00012509328371379524,
          "validation_metric": 9.695703920442611e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.978640764999
        },
        {
          "epoch": 904,
          "updates": 904,
          "samples_seen": 925696,
          "online_loss": 0.0012420356506481767,
          "train_loss": 0.0012401427375152707,
          "validation_loss": 0.0009670417057350278,
          "train_metric": 0.00012490262452047318,
          "validation_metric": 9.739689267007634e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.988266065998687
        },
        {
          "epoch": 905,
          "updates": 905,
          "samples_seen": 926720,
          "online_loss": 0.0012401427375152707,
          "train_loss": 0.0012382562272250652,
          "validation_loss": 0.0009593793074600399,
          "train_metric": 0.00012471263471525162,
          "validation_metric": 9.662517550168559e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.997381317998588
        },
        {
          "epoch": 906,
          "updates": 906,
          "samples_seen": 927744,
          "online_loss": 0.0012382562272250652,
          "train_loss": 0.0012363861314952374,
          "validation_loss": 0.0009648416889831424,
          "train_metric": 0.00012452428927645087,
          "validation_metric": 9.717531793285161e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.00359814499825
        },
        {
          "epoch": 907,
          "updates": 907,
          "samples_seen": 928768,
          "online_loss": 0.0012363861314952374,
          "train_loss": 0.0012345288414508104,
          "validation_loss": 0.0009556396398693323,
          "train_metric": 0.000124337209854275,
          "validation_metric": 9.624852827982977e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.009531675999824
        },
        {
          "epoch": 908,
          "updates": 908,
          "samples_seen": 929792,
          "online_loss": 0.0012345287250354886,
          "train_loss": 0.0012326923897489905,
          "validation_loss": 0.0009633249719627202,
          "train_metric": 0.00012415225501172245,
          "validation_metric": 9.702256647869945e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.01460266599861
        },
        {
          "epoch": 909,
          "updates": 909,
          "samples_seen": 930816,
          "online_loss": 0.0012326922733336687,
          "train_loss": 0.0012308910954743624,
          "validation_loss": 0.0009512589313089848,
          "train_metric": 0.0001239708362845704,
          "validation_metric": 9.580731421010569e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.023155004999353
        },
        {
          "epoch": 910,
          "updates": 910,
          "samples_seen": 931840,
          "online_loss": 0.0012308910954743624,
          "train_loss": 0.0012291355524212122,
          "validation_loss": 0.0009628881234675646,
          "train_metric": 0.00012379404506646097,
          "validation_metric": 9.697856876300648e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.028607916998226
        },
        {
          "epoch": 911,
          "updates": 911,
          "samples_seen": 932864,
          "online_loss": 0.0012291355524212122,
          "train_loss": 0.0012274582404643297,
          "validation_loss": 0.0009457716951146722,
          "train_metric": 0.00012362509733065963,
          "validation_metric": 9.525466884952039e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.038763002999985
        },
        {
          "epoch": 912,
          "updates": 912,
          "samples_seen": 933888,
          "online_loss": 0.001227458007633686,
          "train_loss": 0.0012259100330993533,
          "validation_loss": 0.0009643145021982491,
          "train_metric": 0.00012346915900707245,
          "validation_metric": 9.712222527014092e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.044721716998538
        },
        {
          "epoch": 913,
          "updates": 913,
          "samples_seen": 934912,
          "online_loss": 0.0012259101495146751,
          "train_loss": 0.001224579755216837,
          "validation_loss": 0.0009385041194036603,
          "train_metric": 0.00012333517952356488,
          "validation_metric": 9.45226929616183e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.053859903999182
        },
        {
          "epoch": 914,
          "updates": 914,
          "samples_seen": 935936,
          "online_loss": 0.001224579755216837,
          "train_loss": 0.00122360500972718,
          "validation_loss": 0.0009693398606032133,
          "train_metric": 0.00012323702685534954,
          "validation_metric": 9.762836270965636e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.05985995699848
        },
        {
          "epoch": 915,
          "updates": 915,
          "samples_seen": 936960,
          "online_loss": 0.0012236051261425018,
          "train_loss": 0.0012233237503096461,
          "validation_loss": 0.0009284677798859775,
          "train_metric": 0.00012320867972448468,
          "validation_metric": 9.351185872219503e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.066001845998471
        },
        {
          "epoch": 916,
          "updates": 916,
          "samples_seen": 937984,
          "online_loss": 0.0012233237503096461,
          "train_loss": 0.001224128296598792,
          "validation_loss": 0.0009818548569455743,
          "train_metric": 0.00012328971934039146,
          "validation_metric": 9.888882050290704e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.074172813998302
        },
        {
          "epoch": 917,
          "updates": 917,
          "samples_seen": 939008,
          "online_loss": 0.001224128296598792,
          "train_loss": 0.0012271867599338293,
          "validation_loss": 0.0009149583056569099,
          "train_metric": 0.0001235977397300303,
          "validation_metric": 9.215125464834273e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.079955509998399
        },
        {
          "epoch": 918,
          "updates": 918,
          "samples_seen": 940032,
          "online_loss": 0.0012271866435185075,
          "train_loss": 0.0012336347717791796,
          "validation_loss": 0.0010110160801559687,
          "train_metric": 0.00012424717715475708,
          "validation_metric": 0.00010182583355344832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.085136095998678
        },
        {
          "epoch": 919,
          "updates": 919,
          "samples_seen": 941056,
          "online_loss": 0.0012336346553638577,
          "train_loss": 0.0012477028649300337,
          "validation_loss": 0.0009004448656924069,
          "train_metric": 0.0001256640680367127,
          "validation_metric": 9.068950021173805e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.090920437998648
        },
        {
          "epoch": 920,
          "updates": 920,
          "samples_seen": 942080,
          "online_loss": 0.0012477028649300337,
          "train_loss": 0.001271414221264422,
          "validation_loss": 0.0010787427891045809,
          "train_metric": 0.00012805218284484,
          "validation_metric": 0.00010864702926483005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.096198847999403
        },
        {
          "epoch": 921,
          "updates": 921,
          "samples_seen": 943104,
          "online_loss": 0.0012714143376797438,
          "train_loss": 0.0013192427577450871,
          "validation_loss": 0.0009017993579618633,
          "train_metric": 0.00013286928879097104,
          "validation_metric": 9.082594624487683e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.101379709998582
        },
        {
          "epoch": 922,
          "updates": 922,
          "samples_seen": 944128,
          "online_loss": 0.0013192429905757308,
          "train_loss": 0.0013829497620463371,
          "validation_loss": 0.0012256242334842682,
          "train_metric": 0.00013928560656495392,
          "validation_metric": 0.00012344037531875074,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.106303850998302
        },
        {
          "epoch": 923,
          "updates": 923,
          "samples_seen": 945152,
          "online_loss": 0.0013829494128003716,
          "train_loss": 0.0014974591322243214,
          "validation_loss": 0.000960623612627387,
          "train_metric": 0.00015081861056387424,
          "validation_metric": 9.675048931967467e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.111394938998274
        },
        {
          "epoch": 924,
          "updates": 924,
          "samples_seen": 946176,
          "online_loss": 0.001497459365054965,
          "train_loss": 0.0015629478730261326,
          "validation_loss": 0.0014243029290810227,
          "train_metric": 0.00015741438255645335,
          "validation_metric": 0.0001434505684301257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.116478612999344
        },
        {
          "epoch": 925,
          "updates": 925,
          "samples_seen": 947200,
          "online_loss": 0.0015629479894414544,
          "train_loss": 0.001625338802114129,
          "validation_loss": 0.0010155944619327784,
          "train_metric": 0.00016369814693462104,
          "validation_metric": 0.00010228695464320481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.198224977999416
        },
        {
          "epoch": 926,
          "updates": 926,
          "samples_seen": 948224,
          "online_loss": 0.001625338802114129,
          "train_loss": 0.0014786601532250643,
          "validation_loss": 0.001328574842773378,
          "train_metric": 0.00014892523176968098,
          "validation_metric": 0.00013380919699557126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.204090485998677
        },
        {
          "epoch": 927,
          "updates": 927,
          "samples_seen": 949248,
          "online_loss": 0.0014786601532250643,
          "train_loss": 0.0013108461862429976,
          "validation_loss": 0.0008900425746105611,
          "train_metric": 0.00013202364789322019,
          "validation_metric": 8.964182779891416e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.209939269998358
        },
        {
          "epoch": 928,
          "updates": 928,
          "samples_seen": 950272,
          "online_loss": 0.0013108461862429976,
          "train_loss": 0.001200845930725336,
          "validation_loss": 0.0009529967792332172,
          "train_metric": 0.0001209448091685772,
          "validation_metric": 9.598234464647248e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.21519052199983
        },
        {
          "epoch": 929,
          "updates": 929,
          "samples_seen": 951296,
          "online_loss": 0.001200845930725336,
          "train_loss": 0.0012368918396532536,
          "validation_loss": 0.0010351696982979774,
          "train_metric": 0.00012457520642783493,
          "validation_metric": 0.00010425849905004725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.220887864999895
        },
        {
          "epoch": 930,
          "updates": 930,
          "samples_seen": 952320,
          "online_loss": 0.0012368918396532536,
          "train_loss": 0.0013487533433362842,
          "validation_loss": 0.0009078563889488578,
          "train_metric": 0.0001358414883725345,
          "validation_metric": 9.143598435912281e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.226047234998987
        },
        {
          "epoch": 931,
          "updates": 931,
          "samples_seen": 953344,
          "online_loss": 0.0013487533433362842,
          "train_loss": 0.0013817554572597146,
          "validation_loss": 0.0012297973735257983,
          "train_metric": 0.00013916536408942193,
          "validation_metric": 0.00012386067828629166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.231360392999704
        },
        {
          "epoch": 932,
          "updates": 932,
          "samples_seen": 954368,
          "online_loss": 0.0013817556900903583,
          "train_loss": 0.0013270596973598003,
          "validation_loss": 0.000897020217962563,
          "train_metric": 0.00013365659106057137,
          "validation_metric": 9.03445907169953e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.236687793998499
        },
        {
          "epoch": 933,
          "updates": 933,
          "samples_seen": 955392,
          "online_loss": 0.0013270596973598003,
          "train_loss": 0.0012190721463412046,
          "validation_loss": 0.0010085785761475563,
          "train_metric": 0.00012278046051505953,
          "validation_metric": 0.00010158035001950338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.241809373999786
        },
        {
          "epoch": 934,
          "updates": 934,
          "samples_seen": 956416,
          "online_loss": 0.0012190717970952392,
          "train_loss": 0.0011910214088857174,
          "validation_loss": 0.000943630060646683,
          "train_metric": 0.00011995532986475155,
          "validation_metric": 9.503895853413269e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.246970515998328
        },
        {
          "epoch": 935,
          "updates": 935,
          "samples_seen": 957440,
          "online_loss": 0.0011910215253010392,
          "train_loss": 0.001246530213393271,
          "validation_loss": 0.0008770579006522894,
          "train_metric": 0.0001255459792446345,
          "validation_metric": 8.833406900521368e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.253172841999913
        },
        {
          "epoch": 936,
          "updates": 936,
          "samples_seen": 958464,
          "online_loss": 0.001246530213393271,
          "train_loss": 0.001289769890718162,
          "validation_loss": 0.0011152034858241677,
          "train_metric": 0.0001299009018111974,
          "validation_metric": 0.00011231921962462366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.258419738998782
        },
        {
          "epoch": 937,
          "updates": 937,
          "samples_seen": 959488,
          "online_loss": 0.001289769890718162,
          "train_loss": 0.0012721286620944738,
          "validation_loss": 0.0008824219112284482,
          "train_metric": 0.00012812412751372904,
          "validation_metric": 8.887430158210918e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.263717149999138
        },
        {
          "epoch": 938,
          "updates": 938,
          "samples_seen": 960512,
          "online_loss": 0.001272128545679152,
          "train_loss": 0.0012063595931977034,
          "validation_loss": 0.0009982319315895438,
          "train_metric": 0.00012150011752964929,
          "validation_metric": 0.00010053826554212719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.269142164999721
        },
        {
          "epoch": 939,
          "updates": 939,
          "samples_seen": 961536,
          "online_loss": 0.0012063594767823815,
          "train_loss": 0.0011799458879977465,
          "validation_loss": 0.0009301662212237716,
          "train_metric": 0.00011883984552696347,
          "validation_metric": 9.368294558953494e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.27426602199921
        },
        {
          "epoch": 940,
          "updates": 940,
          "samples_seen": 962560,
          "online_loss": 0.0011799458879977465,
          "train_loss": 0.0012083055917173624,
          "validation_loss": 0.0008822389645501971,
          "train_metric": 0.00012169610999990255,
          "validation_metric": 8.885587158147246e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.279367504999755
        },
        {
          "epoch": 941,
          "updates": 941,
          "samples_seen": 963584,
          "online_loss": 0.0012083054753020406,
          "train_loss": 0.0012380988337099552,
          "validation_loss": 0.001053960295394063,
          "train_metric": 0.00012469678767956793,
          "validation_metric": 0.00010615102655719966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.284800683999492
        },
        {
          "epoch": 942,
          "updates": 942,
          "samples_seen": 964608,
          "online_loss": 0.0012380987172946334,
          "train_loss": 0.00123079854529351,
          "validation_loss": 0.0008727540844120085,
          "train_metric": 0.00012396153761073947,
          "validation_metric": 8.790059655439109e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.290581642999314
        },
        {
          "epoch": 943,
          "updates": 943,
          "samples_seen": 965632,
          "online_loss": 0.0012307986617088318,
          "train_loss": 0.0011911747278645635,
          "validation_loss": 0.0009750477620400488,
          "train_metric": 0.00011997076217085123,
          "validation_metric": 9.820323612075299e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.29608142699908
        },
        {
          "epoch": 944,
          "updates": 944,
          "samples_seen": 966656,
          "online_loss": 0.0011911748442798853,
          "train_loss": 0.0011718004243448377,
          "validation_loss": 0.0009144674986600876,
          "train_metric": 0.00011801944492617622,
          "validation_metric": 9.210180724039674e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.301589932998468
        },
        {
          "epoch": 945,
          "updates": 945,
          "samples_seen": 967680,
          "online_loss": 0.0011718003079295158,
          "train_loss": 0.001186491223052144,
          "validation_loss": 0.0008792012231424451,
          "train_metric": 0.00011949904001085088,
          "validation_metric": 8.854992483975366e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.306729026999164
        },
        {
          "epoch": 946,
          "updates": 946,
          "samples_seen": 968704,
          "online_loss": 0.001186491223052144,
          "train_loss": 0.0012052246602252126,
          "validation_loss": 0.0010077969636768103,
          "train_metric": 0.00012138581951148808,
          "validation_metric": 0.0001015016168821603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.311867991998952
        },
        {
          "epoch": 947,
          "updates": 947,
          "samples_seen": 969728,
          "online_loss": 0.0012052246602252126,
          "train_loss": 0.0012018803972750902,
          "validation_loss": 0.0008724419167265296,
          "train_metric": 0.0001210490008816123,
          "validation_metric": 8.786916441749781e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.317992602998856
        },
        {
          "epoch": 948,
          "updates": 948,
          "samples_seen": 970752,
          "online_loss": 0.001201880513690412,
          "train_loss": 0.0011780490167438984,
          "validation_loss": 0.0009599740151315928,
          "train_metric": 0.00011864879343193024,
          "validation_metric": 9.668506390880793e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.323428021998552
        },
        {
          "epoch": 949,
          "updates": 949,
          "samples_seen": 971776,
          "online_loss": 0.0011780491331592202,
          "train_loss": 0.001163978362455964,
          "validation_loss": 0.0009061984601430595,
          "train_metric": 0.00011723162606358528,
          "validation_metric": 9.126898657996207e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.328546844999437
        },
        {
          "epoch": 950,
          "updates": 950,
          "samples_seen": 972800,
          "online_loss": 0.001163978362455964,
          "train_loss": 0.0011703298659995198,
          "validation_loss": 0.0008804580429568887,
          "train_metric": 0.00011787134280893952,
          "validation_metric": 8.867651195032522e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.333843787999285
        },
        {
          "epoch": 951,
          "updates": 951,
          "samples_seen": 973824,
          "online_loss": 0.001170329749584198,
          "train_loss": 0.001181771862320602,
          "validation_loss": 0.0009703395771794021,
          "train_metric": 0.00011902373080374673,
          "validation_metric": 9.772904741112143e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.339074114999676
        },
        {
          "epoch": 952,
          "updates": 952,
          "samples_seen": 974848,
          "online_loss": 0.001181771862320602,
          "train_loss": 0.001181175233796239,
          "validation_loss": 0.0008665920468047261,
          "train_metric": 0.00011896363866981119,
          "validation_metric": 8.727995736990124e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.344714093998846
        },
        {
          "epoch": 953,
          "updates": 953,
          "samples_seen": 975872,
          "online_loss": 0.001181175117380917,
          "train_loss": 0.0011672027176246047,
          "validation_loss": 0.0009425262687727809,
          "train_metric": 0.00011755638843169436,
          "validation_metric": 9.49277964537032e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.352948008998283
        },
        {
          "epoch": 954,
          "updates": 954,
          "samples_seen": 976896,
          "online_loss": 0.0011672027176246047,
          "train_loss": 0.0011566011235117912,
          "validation_loss": 0.0008938173996284604,
          "train_metric": 0.00011648863437585533,
          "validation_metric": 9.002201113617048e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.358359040999858
        },
        {
          "epoch": 955,
          "updates": 955,
          "samples_seen": 977920,
          "online_loss": 0.0011566011235117912,
          "train_loss": 0.0011578453704714775,
          "validation_loss": 0.0008827471756376326,
          "train_metric": 0.00011661395546980202,
          "validation_metric": 8.890705794328824e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.363673288999053
        },
        {
          "epoch": 956,
          "updates": 956,
          "samples_seen": 978944,
          "online_loss": 0.0011578454868867993,
          "train_loss": 0.0011644322657957673,
          "validation_loss": 0.0009452062076888978,
          "train_metric": 0.00011727735545719042,
          "validation_metric": 9.519770537735894e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.369233742998404
        },
        {
          "epoch": 957,
          "updates": 957,
          "samples_seen": 979968,
          "online_loss": 0.0011644323822110891,
          "train_loss": 0.0011655993293970823,
          "validation_loss": 0.000868197821546346,
          "train_metric": 0.00011739489855244756,
          "validation_metric": 8.744170190766454e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.37445080499856
        },
        {
          "epoch": 958,
          "updates": 958,
          "samples_seen": 980992,
          "online_loss": 0.0011655993293970823,
          "train_loss": 0.0011580055579543114,
          "validation_loss": 0.0009338286472484469,
          "train_metric": 0.00011663007899187505,
          "validation_metric": 9.405180753674358e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.379586119999658
        },
        {
          "epoch": 959,
          "updates": 959,
          "samples_seen": 982016,
          "online_loss": 0.0011580055579543114,
          "train_loss": 0.0011497994419187307,
          "validation_loss": 0.0008833558531478047,
          "train_metric": 0.00011580361024243757,
          "validation_metric": 8.896837243810296e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.386152609999044
        },
        {
          "epoch": 960,
          "updates": 960,
          "samples_seen": 983040,
          "online_loss": 0.0011497994419187307,
          "train_loss": 0.0011476074578240514,
          "validation_loss": 0.000885270070284605,
          "train_metric": 0.00011558281403267756,
          "validation_metric": 8.916116348700598e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.39141911599836
        },
        {
          "epoch": 961,
          "updates": 961,
          "samples_seen": 984064,
          "online_loss": 0.0011476073414087296,
          "train_loss": 0.001150436932221055,
          "validation_loss": 0.0009206347749568522,
          "train_metric": 0.00011586779146455228,
          "validation_metric": 9.272296301787719e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.396776951998618
        },
        {
          "epoch": 962,
          "updates": 962,
          "samples_seen": 985088,
          "online_loss": 0.0011504368158057332,
          "train_loss": 0.0011522539425641298,
          "validation_loss": 0.000866347283590585,
          "train_metric": 0.00011605078907450661,
          "validation_metric": 8.725532825337723e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.402872562999619
        },
        {
          "epoch": 963,
          "updates": 963,
          "samples_seen": 986112,
          "online_loss": 0.0011522539425641298,
          "train_loss": 0.001149010960943997,
          "validation_loss": 0.0009244393440894783,
          "train_metric": 0.00011572418588912115,
          "validation_metric": 9.310615132562816e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.408115516998805
        },
        {
          "epoch": 964,
          "updates": 964,
          "samples_seen": 987136,
          "online_loss": 0.001149010844528675,
          "train_loss": 0.0011433117324486375,
          "validation_loss": 0.0008755133021622896,
          "train_metric": 0.00011515017104102299,
          "validation_metric": 8.81784872035496e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.547994249998737
        },
        {
          "epoch": 965,
          "updates": 965,
          "samples_seen": 988160,
          "online_loss": 0.0011433118488639593,
          "train_loss": 0.0011394537286832929,
          "validation_loss": 0.000891581759788096,
          "train_metric": 0.00011476162035251036,
          "validation_metric": 8.979684207588434e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.561379493999993
        },
        {
          "epoch": 966,
          "updates": 966,
          "samples_seen": 989184,
          "online_loss": 0.0011394538450986147,
          "train_loss": 0.0011391950538381934,
          "validation_loss": 0.0009018811979331076,
          "train_metric": 0.00011473555059637874,
          "validation_metric": 9.083417535293847e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.568010277998837
        },
        {
          "epoch": 967,
          "updates": 967,
          "samples_seen": 990208,
          "online_loss": 0.0011391950538381934,
          "train_loss": 0.0011403220705688,
          "validation_loss": 0.0008680981118232012,
          "train_metric": 0.00011484907008707523,
          "validation_metric": 8.743165381019935e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.58117835899975
        },
        {
          "epoch": 968,
          "updates": 968,
          "samples_seen": 991232,
          "online_loss": 0.0011403221869841218,
          "train_loss": 0.0011396093759685755,
          "validation_loss": 0.0009125121869146824,
          "train_metric": 0.00011477727821329609,
          "validation_metric": 9.190487617161125e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.595056397000008
        },
        {
          "epoch": 969,
          "updates": 969,
          "samples_seen": 992256,
          "online_loss": 0.0011396094923838973,
          "train_loss": 0.0011364973615854979,
          "validation_loss": 0.0008669067174196243,
          "train_metric": 0.00011446386633906513,
          "validation_metric": 8.731166599318385e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.600341292998564
        },
        {
          "epoch": 970,
          "updates": 970,
          "samples_seen": 993280,
          "online_loss": 0.0011364974780008197,
          "train_loss": 0.0011327004758641124,
          "validation_loss": 0.0008930914918892086,
          "train_metric": 0.00011408144928282127,
          "validation_metric": 8.994890231406316e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.61370727999929
        },
        {
          "epoch": 971,
          "updates": 971,
          "samples_seen": 994304,
          "online_loss": 0.0011327004758641124,
          "train_loss": 0.0011304001091048121,
          "validation_loss": 0.0008830142323859036,
          "train_metric": 0.00011384976096451283,
          "validation_metric": 8.89339717105031e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.619350835999285
        },
        {
          "epoch": 972,
          "updates": 972,
          "samples_seen": 995328,
          "online_loss": 0.001130400225520134,
          "train_loss": 0.0011298884637653828,
          "validation_loss": 0.000870901218149811,
          "train_metric": 0.00011379822535673156,
          "validation_metric": 8.771398279350251e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.628478081998765
        },
        {
          "epoch": 973,
          "updates": 973,
          "samples_seen": 996352,
          "online_loss": 0.0011298884637653828,
          "train_loss": 0.0011297522578388453,
          "validation_loss": 0.0008987498003989458,
          "train_metric": 0.0001137844956247136,
          "validation_metric": 9.051879897015169e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.641953913998805
        },
        {
          "epoch": 974,
          "updates": 974,
          "samples_seen": 997376,
          "online_loss": 0.0011297520250082016,
          "train_loss": 0.0011285861255601048,
          "validation_loss": 0.000863432593178004,
          "train_metric": 0.00011366706166882068,
          "validation_metric": 8.696176519151777e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.651650630999939
        },
        {
          "epoch": 975,
          "updates": 975,
          "samples_seen": 998400,
          "online_loss": 0.0011285861255601048,
          "train_loss": 0.001126125454902649,
          "validation_loss": 0.0008931998163461685,
          "train_metric": 0.00011341924255248159,
          "validation_metric": 8.995982352644205e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.664679588999206
        },
        {
          "epoch": 976,
          "updates": 976,
          "samples_seen": 999424,
          "online_loss": 0.0011261256877332926,
          "train_loss": 0.0011234126286581159,
          "validation_loss": 0.0008689611568115652,
          "train_metric": 0.00011314600851619616,
          "validation_metric": 8.7518579675816e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.677119772999504
        },
        {
          "epoch": 977,
          "updates": 977,
          "samples_seen": 1000448,
          "online_loss": 0.0011234127450734377,
          "train_loss": 0.001121434848755598,
          "validation_loss": 0.0008746622479520738,
          "train_metric": 0.00011294680734863505,
          "validation_metric": 8.809277642285451e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.68256952199954
        },
        {
          "epoch": 978,
          "updates": 978,
          "samples_seen": 1001472,
          "online_loss": 0.0011214347323402762,
          "train_loss": 0.0011203918838873506,
          "validation_loss": 0.0008812457090243697,
          "train_metric": 0.00011284176434855908,
          "validation_metric": 8.875584171619266e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.689542536998488
        },
        {
          "epoch": 979,
          "updates": 979,
          "samples_seen": 1002496,
          "online_loss": 0.0011203918838873506,
          "train_loss": 0.0011196635896340013,
          "validation_loss": 0.0008619240252301097,
          "train_metric": 0.00011276842269580811,
          "validation_metric": 8.680982864461839e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.695325018999938
        },
        {
          "epoch": 980,
          "updates": 980,
          "samples_seen": 1003520,
          "online_loss": 0.001119663706049323,
          "train_loss": 0.0011184938484802842,
          "validation_loss": 0.0008863006951287389,
          "train_metric": 0.00011265060311416164,
          "validation_metric": 8.92649550223723e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.700509159998546
        },
        {
          "epoch": 981,
          "updates": 981,
          "samples_seen": 1004544,
          "online_loss": 0.0011184938484802842,
          "train_loss": 0.0011166803305968642,
          "validation_loss": 0.000860413012560457,
          "train_metric": 0.0001124679547501728,
          "validation_metric": 8.665764471516013e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.706223486999079
        },
        {
          "epoch": 982,
          "updates": 982,
          "samples_seen": 1005568,
          "online_loss": 0.0011166803305968642,
          "train_loss": 0.0011145443422719836,
          "validation_loss": 0.0008782324148342013,
          "train_metric": 0.00011225282651139423,
          "validation_metric": 8.845234697218984e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.71150307699827
        },
        {
          "epoch": 983,
          "updates": 983,
          "samples_seen": 1006592,
          "online_loss": 0.0011145443422719836,
          "train_loss": 0.0011126184836030006,
          "validation_loss": 0.0008666800567880273,
          "train_metric": 0.0001120588494814001,
          "validation_metric": 8.728883403819054e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.716604502998962
        },
        {
          "epoch": 984,
          "updates": 984,
          "samples_seen": 1007616,
          "online_loss": 0.0011126184836030006,
          "train_loss": 0.0011111624771729112,
          "validation_loss": 0.0008651454118080437,
          "train_metric": 0.00011191221710760146,
          "validation_metric": 8.713427087059245e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.721870885998214
        },
        {
          "epoch": 985,
          "updates": 985,
          "samples_seen": 1008640,
          "online_loss": 0.0011111624771729112,
          "train_loss": 0.0011100381379947066,
          "validation_loss": 0.0008737234747968614,
          "train_metric": 0.00011179898865520954,
          "validation_metric": 8.799823262961581e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.727063230999192
        },
        {
          "epoch": 986,
          "updates": 986,
          "samples_seen": 1009664,
          "online_loss": 0.0011100382544100285,
          "train_loss": 0.001108901109546423,
          "validation_loss": 0.0008567421464249492,
          "train_metric": 0.00011168445780640468,
          "validation_metric": 8.628793875686824e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.73282629699861
        },
        {
          "epoch": 987,
          "updates": 987,
          "samples_seen": 1010688,
          "online_loss": 0.0011089009931311011,
          "train_loss": 0.001107491203583777,
          "validation_loss": 0.0008743372745811939,
          "train_metric": 0.0001115424747695215,
          "validation_metric": 8.80600418895483e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.738213124999675
        },
        {
          "epoch": 988,
          "updates": 988,
          "samples_seen": 1011712,
          "online_loss": 0.0011074915528297424,
          "train_loss": 0.0011058050440624356,
          "validation_loss": 0.0008557377150282264,
          "train_metric": 0.00011137264664284885,
          "validation_metric": 8.618677384220064e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.74417349199939
        },
        {
          "epoch": 989,
          "updates": 989,
          "samples_seen": 1012736,
          "online_loss": 0.0011058051604777575,
          "train_loss": 0.0011040230747312307,
          "validation_loss": 0.0008677478763274848,
          "train_metric": 0.0001111931778723374,
          "validation_metric": 8.73963872436434e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.749382262998552
        },
        {
          "epoch": 990,
          "updates": 990,
          "samples_seen": 1013760,
          "online_loss": 0.0011040230747312307,
          "train_loss": 0.0011023572878912091,
          "validation_loss": 0.000859776686411351,
          "train_metric": 0.00011102539428975433,
          "validation_metric": 8.65935580804944e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.754676147998907
        },
        {
          "epoch": 991,
          "updates": 991,
          "samples_seen": 1014784,
          "online_loss": 0.0011023572878912091,
          "train_loss": 0.0011009065201506019,
          "validation_loss": 0.0008586258045397699,
          "train_metric": 0.0001108792785089463,
          "validation_metric": 8.647764479974285e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.760003339999457
        },
        {
          "epoch": 992,
          "updates": 992,
          "samples_seen": 1015808,
          "online_loss": 0.0011009067529812455,
          "train_loss": 0.0010996084893122315,
          "validation_loss": 0.0008637391729280353,
          "train_metric": 0.00011074854410253465,
          "validation_metric": 8.699264435563236e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.765453135998541
        },
        {
          "epoch": 993,
          "updates": 993,
          "samples_seen": 1016832,
          "online_loss": 0.0010996084893122315,
          "train_loss": 0.0010983278043568134,
          "validation_loss": 0.0008519160910509527,
          "train_metric": 0.00011061956320190802,
          "validation_metric": 8.580186840845272e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.771219560998361
        },
        {
          "epoch": 994,
          "updates": 994,
          "samples_seen": 1017856,
          "online_loss": 0.0010983278043568134,
          "train_loss": 0.001096949796192348,
          "validation_loss": 0.0008638591971248388,
          "train_metric": 0.00011048075975850224,
          "validation_metric": 8.700472972122952e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.776428335999299
        },
        {
          "epoch": 995,
          "updates": 995,
          "samples_seen": 1018880,
          "online_loss": 0.0010969496797770262,
          "train_loss": 0.0010954441968351603,
          "validation_loss": 0.0008495688671246171,
          "train_metric": 0.00011032914335373789,
          "validation_metric": 8.556546526961029e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.78320552399964
        },
        {
          "epoch": 996,
          "updates": 996,
          "samples_seen": 1019904,
          "online_loss": 0.001095444313250482,
          "train_loss": 0.0010938623454421759,
          "validation_loss": 0.0008595650433562696,
          "train_metric": 0.00011016980715794489,
          "validation_metric": 8.657224680064246e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.788793552999778
        },
        {
          "epoch": 997,
          "updates": 997,
          "samples_seen": 1020928,
          "online_loss": 0.0010938623454421759,
          "train_loss": 0.0010922836372628808,
          "validation_loss": 0.0008507499005645514,
          "train_metric": 0.00011001082020811737,
          "validation_metric": 8.568441262468696e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.794041618999472
        },
        {
          "epoch": 998,
          "updates": 998,
          "samples_seen": 1021952,
          "online_loss": 0.0010922836372628808,
          "train_loss": 0.0010907781543210149,
          "validation_loss": 0.0008532565552741289,
          "train_metric": 0.00010985917469952255,
          "validation_metric": 8.593688107794151e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.799184456998773
        },
        {
          "epoch": 999,
          "updates": 999,
          "samples_seen": 1022976,
          "online_loss": 0.0010907782707363367,
          "train_loss": 0.00108935812022537,
          "validation_loss": 0.0008529435726813972,
          "train_metric": 0.0001097161730285734,
          "validation_metric": 8.590535435359925e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.804337009998562
        },
        {
          "epoch": 1000,
          "updates": 1000,
          "samples_seen": 1024000,
          "online_loss": 0.00108935812022537,
          "train_loss": 0.001087991870008409,
          "validation_loss": 0.0008474442875012755,
          "train_metric": 0.00010957855556625873,
          "validation_metric": 8.535148663213477e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.809967384999254
        },
        {
          "epoch": 1001,
          "updates": 1001,
          "samples_seen": 1025024,
          "online_loss": 0.001087991870008409,
          "train_loss": 0.0010866356315091252,
          "validation_loss": 0.0008536216337233782,
          "train_metric": 0.00010944197128992528,
          "validation_metric": 8.597364649176598e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.815189100998396
        },
        {
          "epoch": 1002,
          "updates": 1002,
          "samples_seen": 1026048,
          "online_loss": 0.001086635747924447,
          "train_loss": 0.00108524935785681,
          "validation_loss": 0.0008436364587396383,
          "train_metric": 0.0001093023456633091,
          "validation_metric": 8.496797818224877e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.901158025999393
        },
        {
          "epoch": 1003,
          "updates": 1003,
          "samples_seen": 1027072,
          "online_loss": 0.00108524935785681,
          "train_loss": 0.0010838144225999713,
          "validation_loss": 0.0008518046815879643,
          "train_metric": 0.00010915781604126096,
          "validation_metric": 8.579064888181165e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.906459780999285
        },
        {
          "epoch": 1004,
          "updates": 1004,
          "samples_seen": 1028096,
          "online_loss": 0.0010838144225999713,
          "train_loss": 0.0010823437478393316,
          "validation_loss": 0.000842145353090018,
          "train_metric": 0.00010900970664806664,
          "validation_metric": 8.481779514113441e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.912083748999066
        },
        {
          "epoch": 1005,
          "updates": 1005,
          "samples_seen": 1029120,
          "online_loss": 0.0010823437478393316,
          "train_loss": 0.001080861664377153,
          "validation_loss": 0.0008481693803332746,
          "train_metric": 0.00010886043310165405,
          "validation_metric": 8.542451541870832e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.917105892998734
        },
        {
          "epoch": 1006,
          "updates": 1006,
          "samples_seen": 1030144,
          "online_loss": 0.0010808617807924747,
          "train_loss": 0.0010793919209390879,
          "validation_loss": 0.0008422804530709982,
          "train_metric": 0.00010871240374399349,
          "validation_metric": 8.483140118187293e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.922254738999982
        },
        {
          "epoch": 1007,
          "updates": 1007,
          "samples_seen": 1031168,
          "online_loss": 0.001079391804523766,
          "train_loss": 0.00107794557698071,
          "validation_loss": 0.0008437252254225314,
          "train_metric": 0.00010856673907255754,
          "validation_metric": 8.497692761011422e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.927345921998494
        },
        {
          "epoch": 1008,
          "updates": 1008,
          "samples_seen": 1032192,
          "online_loss": 0.00107794557698071,
          "train_loss": 0.0010765284532681108,
          "validation_loss": 0.0008425986743532121,
          "train_metric": 0.00010842402116395533,
          "validation_metric": 8.486345177516341e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.932464739998977
        },
        {
          "epoch": 1009,
          "updates": 1009,
          "samples_seen": 1033216,
          "online_loss": 0.0010765285696834326,
          "train_loss": 0.001075137173756957,
          "validation_loss": 0.0008394489996135235,
          "train_metric": 0.00010828388622030616,
          "validation_metric": 8.454622002318501e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.938035248998858
        },
        {
          "epoch": 1010,
          "updates": 1010,
          "samples_seen": 1034240,
          "online_loss": 0.0010751372901722789,
          "train_loss": 0.001073754159733653,
          "validation_loss": 0.0008422346436418593,
          "train_metric": 0.000108144580735825,
          "validation_metric": 8.482678094878793e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.94326549699872
        },
        {
          "epoch": 1011,
          "updates": 1011,
          "samples_seen": 1035264,
          "online_loss": 0.001073754159733653,
          "train_loss": 0.0010723693994805217,
          "validation_loss": 0.0008360397187061608,
          "train_metric": 0.00010800512973219156,
          "validation_metric": 8.420286030741408e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.94896089499889
        },
        {
          "epoch": 1012,
          "updates": 1012,
          "samples_seen": 1036288,
          "online_loss": 0.0010723695158958435,
          "train_loss": 0.0010709785856306553,
          "validation_loss": 0.0008409128640778363,
          "train_metric": 0.00010786502389237285,
          "validation_metric": 8.469367458019406e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.95445381799982
        },
        {
          "epoch": 1013,
          "updates": 1013,
          "samples_seen": 1037312,
          "online_loss": 0.0010709787020459771,
          "train_loss": 0.0010695750825107098,
          "validation_loss": 0.0008337442413903773,
          "train_metric": 0.00010772368841571733,
          "validation_metric": 8.397167403018102e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.960314116999143
        },
        {
          "epoch": 1014,
          "updates": 1014,
          "samples_seen": 1038336,
          "online_loss": 0.0010695750825107098,
          "train_loss": 0.001068162964656949,
          "validation_loss": 0.0008387715788558125,
          "train_metric": 0.00010758145072031766,
          "validation_metric": 8.447800064459443e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.965685736999149
        },
        {
          "epoch": 1015,
          "updates": 1015,
          "samples_seen": 1039360,
          "online_loss": 0.0010681628482416272,
          "train_loss": 0.0010667475871741772,
          "validation_loss": 0.0008322414360009134,
          "train_metric": 0.0001074389074346982,
          "validation_metric": 8.382031228393316e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.971467373999985
        },
        {
          "epoch": 1016,
          "updates": 1016,
          "samples_seen": 1040384,
          "online_loss": 0.0010667475871741772,
          "train_loss": 0.001065330347046256,
          "validation_loss": 0.0008359675412066281,
          "train_metric": 0.00010729618225013837,
          "validation_metric": 8.419559162575752e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.976615657998991
        },
        {
          "epoch": 1017,
          "updates": 1017,
          "samples_seen": 1041408,
          "online_loss": 0.001065330347046256,
          "train_loss": 0.0010639141546562314,
          "validation_loss": 0.0008310641860589385,
          "train_metric": 0.0001071535371011123,
          "validation_metric": 8.370174327865243e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.982377618998726
        },
        {
          "epoch": 1018,
          "updates": 1018,
          "samples_seen": 1042432,
          "online_loss": 0.0010639141546562314,
          "train_loss": 0.001062507857568562,
          "validation_loss": 0.0008328171097673476,
          "train_metric": 0.00010701190331019461,
          "validation_metric": 8.387827256228775e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.987909822999427
        },
        {
          "epoch": 1019,
          "updates": 1019,
          "samples_seen": 1043456,
          "online_loss": 0.001062507857568562,
          "train_loss": 0.001061108661815524,
          "validation_loss": 0.0008299126056954265,
          "train_metric": 0.00010687098256312311,
          "validation_metric": 8.358576451428235e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.993658821998906
        },
        {
          "epoch": 1020,
          "updates": 1020,
          "samples_seen": 1044480,
          "online_loss": 0.001061108661815524,
          "train_loss": 0.0010597157524898648,
          "validation_loss": 0.0008297190652228892,
          "train_metric": 0.00010673068754840642,
          "validation_metric": 8.356625767191872e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.999516875999689
        },
        {
          "epoch": 1021,
          "updates": 1021,
          "samples_seen": 1045504,
          "online_loss": 0.001059715636074543,
          "train_loss": 0.0010583290131762624,
          "validation_loss": 0.0008287726086564362,
          "train_metric": 0.00010659101826604456,
          "validation_metric": 8.347094990313053e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.005473816998347
        },
        {
          "epoch": 1022,
          "updates": 1022,
          "samples_seen": 1046528,
          "online_loss": 0.0010583290131762624,
          "train_loss": 0.001056944252923131,
          "validation_loss": 0.0008269174722954631,
          "train_metric": 0.00010645153088262305,
          "validation_metric": 8.32841033115983e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.011233810999329
        },
        {
          "epoch": 1023,
          "updates": 1023,
          "samples_seen": 1047552,
          "online_loss": 0.0010569441365078092,
          "train_loss": 0.001055566011928022,
          "validation_loss": 0.0008275028667412698,
          "train_metric": 0.00010631272016325966,
          "validation_metric": 8.334305312018842e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.016497730999617
        },
        {
          "epoch": 1024,
          "updates": 1024,
          "samples_seen": 1048576,
          "online_loss": 0.0010555658955127,
          "train_loss": 0.0010541909141466022,
          "validation_loss": 0.0008243009215220809,
          "train_metric": 0.00010617425868986174,
          "validation_metric": 8.302056812681258e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.022343519998685
        },
        {
          "epoch": 1025,
          "updates": 1025,
          "samples_seen": 1049600,
          "online_loss": 0.001054191030561924,
          "train_loss": 0.001052818144671619,
          "validation_loss": 0.000826011411845684,
          "train_metric": 0.00010603599366731942,
          "validation_metric": 8.31928409752436e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.02768927499892
        },
        {
          "epoch": 1026,
          "updates": 1026,
          "samples_seen": 1050624,
          "online_loss": 0.001052818144671619,
          "train_loss": 0.0010514461901038885,
          "validation_loss": 0.000821778376121074,
          "train_metric": 0.00010589780868031085,
          "validation_metric": 8.276650623884052e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.033608541998547
        },
        {
          "epoch": 1027,
          "updates": 1027,
          "samples_seen": 1051648,
          "online_loss": 0.0010514461901038885,
          "train_loss": 0.001050077029503882,
          "validation_loss": 0.0008243612828664482,
          "train_metric": 0.00010575990017969161,
          "validation_metric": 8.302665082737803e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.039074908998373
        },
        {
          "epoch": 1028,
          "updates": 1028,
          "samples_seen": 1052672,
          "online_loss": 0.00105007691308856,
          "train_loss": 0.0010487104300409555,
          "validation_loss": 0.0008193228859454393,
          "train_metric": 0.00010562227544141933,
          "validation_metric": 8.251919643953443e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.04486751499826
        },
        {
          "epoch": 1029,
          "updates": 1029,
          "samples_seen": 1053696,
          "online_loss": 0.0010487104300409555,
          "train_loss": 0.0010473481379449368,
          "validation_loss": 0.000822733505629003,
          "train_metric": 0.00010548507270868868,
          "validation_metric": 8.286270895041525e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.050282151998545
        },
        {
          "epoch": 1030,
          "updates": 1030,
          "samples_seen": 1054720,
          "online_loss": 0.0010473481379449368,
          "train_loss": 0.001045989221893251,
          "validation_loss": 0.0008169212378561497,
          "train_metric": 0.00010534820467000827,
          "validation_metric": 8.22773072286509e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.05625095799951
        },
        {
          "epoch": 1031,
          "updates": 1031,
          "samples_seen": 1055744,
          "online_loss": 0.001045989221893251,
          "train_loss": 0.0010446365922689438,
          "validation_loss": 0.00082122947787866,
          "train_metric": 0.00010521196236368269,
          "validation_metric": 8.271122351288795e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.061606685998413
        },
        {
          "epoch": 1032,
          "updates": 1032,
          "samples_seen": 1056768,
          "online_loss": 0.0010446365922689438,
          "train_loss": 0.001043285126797855,
          "validation_loss": 0.0008144109742715955,
          "train_metric": 0.000105075872852467,
          "validation_metric": 8.202449680538848e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.067420162999042
        },
        {
          "epoch": 1033,
          "updates": 1033,
          "samples_seen": 1057792,
          "online_loss": 0.001043285126797855,
          "train_loss": 0.001041945768520236,
          "validation_loss": 0.0008199095609597862,
          "train_metric": 0.00010494094749446958,
          "validation_metric": 8.257828449131921e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.072841285998948
        },
        {
          "epoch": 1034,
          "updates": 1034,
          "samples_seen": 1058816,
          "online_loss": 0.0010419456521049142,
          "train_loss": 0.0010406154906377196,
          "validation_loss": 0.0008115966338664293,
          "train_metric": 0.00010480697528691962,
          "validation_metric": 8.174104004865512e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.078611203998662
        },
        {
          "epoch": 1035,
          "updates": 1035,
          "samples_seen": 1059840,
          "online_loss": 0.0010406153742223978,
          "train_loss": 0.0010393027914687991,
          "validation_loss": 0.000818924221675843,
          "train_metric": 0.00010467477841302752,
          "validation_metric": 8.247904770541936e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.08391332499923
        },
        {
          "epoch": 1036,
          "updates": 1036,
          "samples_seen": 1060864,
          "online_loss": 0.0010393027914687991,
          "train_loss": 0.0010380118619650602,
          "validation_loss": 0.0008083602297119796,
          "train_metric": 0.00010454475705046207,
          "validation_metric": 8.141508442349732e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.08960343399849
        },
        {
          "epoch": 1037,
          "updates": 1037,
          "samples_seen": 1061888,
          "online_loss": 0.001038011978380382,
          "train_loss": 0.0010367571376264095,
          "validation_loss": 0.000818643718957901,
          "train_metric": 0.00010441838821861893,
          "validation_metric": 8.24508024379611e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.094787799998812
        },
        {
          "epoch": 1038,
          "updates": 1038,
          "samples_seen": 1062912,
          "online_loss": 0.0010367572540417314,
          "train_loss": 0.0010355535196140409,
          "validation_loss": 0.0008044883725233376,
          "train_metric": 0.0001042971562128514,
          "validation_metric": 8.102512219920754e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.100602485999843
        },
        {
          "epoch": 1039,
          "updates": 1039,
          "samples_seen": 1063936,
          "online_loss": 0.001035553403198719,
          "train_loss": 0.0010344355832785368,
          "validation_loss": 0.0008195575792342424,
          "train_metric": 0.00010418457532068714,
          "validation_metric": 8.254284330178052e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.106003922999662
        },
        {
          "epoch": 1040,
          "updates": 1040,
          "samples_seen": 1064960,
          "online_loss": 0.0010344355832785368,
          "train_loss": 0.0010334511753171682,
          "validation_loss": 0.0007995711057446897,
          "train_metric": 0.0001040854403981939,
          "validation_metric": 8.052986959228292e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.11172577599973
        },
        {
          "epoch": 1041,
          "updates": 1041,
          "samples_seen": 1065984,
          "online_loss": 0.00103345129173249,
          "train_loss": 0.0010326799238100648,
          "validation_loss": 0.0008226088248193264,
          "train_metric": 0.00010400774772278965,
          "validation_metric": 8.285015064757317e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.116889310998886
        },
        {
          "epoch": 1042,
          "updates": 1042,
          "samples_seen": 1067008,
          "online_loss": 0.0010326799238100648,
          "train_loss": 0.0010322824819013476,
          "validation_loss": 0.0007930378778837621,
          "train_metric": 0.00010396772267995402,
          "validation_metric": 7.987188291735947e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.196090336999987
        },
        {
          "epoch": 1043,
          "updates": 1043,
          "samples_seen": 1068032,
          "online_loss": 0.0010322824819013476,
          "train_loss": 0.0010324434842914343,
          "validation_loss": 0.0008297444437630475,
          "train_metric": 0.00010398391168564558,
          "validation_metric": 8.356881880899891e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.20140395299859
        },
        {
          "epoch": 1044,
          "updates": 1044,
          "samples_seen": 1069056,
          "online_loss": 0.0010324433678761125,
          "train_loss": 0.0010337011190131307,
          "validation_loss": 0.000784440606366843,
          "train_metric": 0.0001041105788317509,
          "validation_metric": 7.90059712016955e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.20722487799867
        },
        {
          "epoch": 1045,
          "updates": 1045,
          "samples_seen": 1070080,
          "online_loss": 0.0010337012354284525,
          "train_loss": 0.001036510686390102,
          "validation_loss": 0.0008453326299786568,
          "train_metric": 0.00010439356265123934,
          "validation_metric": 8.513880311511457e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.212342162998539
        },
        {
          "epoch": 1046,
          "updates": 1046,
          "samples_seen": 1071104,
          "online_loss": 0.001036510686390102,
          "train_loss": 0.0010426302906125784,
          "validation_loss": 0.00077442592009902,
          "train_metric": 0.00010500990902073681,
          "validation_metric": 7.799734157742932e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.217895635998502
        },
        {
          "epoch": 1047,
          "updates": 1047,
          "samples_seen": 1072128,
          "online_loss": 0.0010426301741972566,
          "train_loss": 0.001053040730766952,
          "validation_loss": 0.0008793330052867532,
          "train_metric": 0.0001060584036167711,
          "validation_metric": 8.856320346239954e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.22321961599846
        },
        {
          "epoch": 1048,
          "updates": 1048,
          "samples_seen": 1073152,
          "online_loss": 0.0010530406143516302,
          "train_loss": 0.001073591411113739,
          "validation_loss": 0.0007687374600209296,
          "train_metric": 0.00010812819527927786,
          "validation_metric": 7.742442539893091e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.229050010999345
        },
        {
          "epoch": 1049,
          "updates": 1049,
          "samples_seen": 1074176,
          "online_loss": 0.0010735912946984172,
          "train_loss": 0.0011042996775358915,
          "validation_loss": 0.0009538914891891181,
          "train_metric": 0.00011122100841021165,
          "validation_metric": 9.607245738152415e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.234221763999813
        },
        {
          "epoch": 1050,
          "updates": 1050,
          "samples_seen": 1075200,
          "online_loss": 0.0011042995611205697,
          "train_loss": 0.001163286971859634,
          "validation_loss": 0.0007886153180152178,
          "train_metric": 0.00011716201697709039,
          "validation_metric": 7.942644151626155e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.2393666809985
        },
        {
          "epoch": 1051,
          "updates": 1051,
          "samples_seen": 1076224,
          "online_loss": 0.0011632870882749557,
          "train_loss": 0.0012316154316067696,
          "validation_loss": 0.0011021289974451065,
          "train_metric": 0.00012404379958752543,
          "validation_metric": 0.00011100238771177828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.244493863998287
        },
        {
          "epoch": 1052,
          "updates": 1052,
          "samples_seen": 1077248,
          "online_loss": 0.0012316155480220914,
          "train_loss": 0.0013508237898349762,
          "validation_loss": 0.0008672971744090319,
          "train_metric": 0.00013605004642158747,
          "validation_metric": 8.735099254408851e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.249487245999262
        },
        {
          "epoch": 1053,
          "updates": 1053,
          "samples_seen": 1078272,
          "online_loss": 0.0013508237898349762,
          "train_loss": 0.001400851528160274,
          "validation_loss": 0.0012745785061269999,
          "train_metric": 0.00014108863251749426,
          "validation_metric": 0.00012837088434025645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.254788316999111
        },
        {
          "epoch": 1054,
          "updates": 1054,
          "samples_seen": 1079296,
          "online_loss": 0.001400851528160274,
          "train_loss": 0.0014498753007501364,
          "validation_loss": 0.0009137747692875564,
          "train_metric": 0.0001460261264583096,
          "validation_metric": 9.203204535879195e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.259946411999408
        },
        {
          "epoch": 1055,
          "updates": 1055,
          "samples_seen": 1080320,
          "online_loss": 0.0014498753007501364,
          "train_loss": 0.0013004079228267074,
          "validation_loss": 0.0011672988766804338,
          "train_metric": 0.0001309723302256316,
          "validation_metric": 0.00011756607273127884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.265231461999065
        },
        {
          "epoch": 1056,
          "updates": 1056,
          "samples_seen": 1081344,
          "online_loss": 0.0013004078064113855,
          "train_loss": 0.0011398923816159368,
          "validation_loss": 0.000772323866840452,
          "train_metric": 0.00011480577813927084,
          "validation_metric": 7.778562576277182e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.270277210998756
        },
        {
          "epoch": 1057,
          "updates": 1057,
          "samples_seen": 1082368,
          "online_loss": 0.0011398923816159368,
          "train_loss": 0.0010211654007434845,
          "validation_loss": 0.0008281252230517566,
          "train_metric": 0.00010284804011462256,
          "validation_metric": 8.34057354950346e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.275328550998893
        },
        {
          "epoch": 1058,
          "updates": 1058,
          "samples_seen": 1083392,
          "online_loss": 0.0010211654007434845,
          "train_loss": 0.0010326693300157785,
          "validation_loss": 0.0008570919744670391,
          "train_metric": 0.00010400667088106275,
          "validation_metric": 8.63231616676785e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.28022464699825
        },
        {
          "epoch": 1059,
          "updates": 1059,
          "samples_seen": 1084416,
          "online_loss": 0.0010326694464311004,
          "train_loss": 0.0011304705403745174,
          "validation_loss": 0.0007813285337761045,
          "train_metric": 0.00011385685502318665,
          "validation_metric": 7.869255205150694e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.285346043998288
        },
        {
          "epoch": 1060,
          "updates": 1060,
          "samples_seen": 1085440,
          "online_loss": 0.0011304705403745174,
          "train_loss": 0.0011935755610466003,
          "validation_loss": 0.001067234785296023,
          "train_metric": 0.000120212571346201,
          "validation_metric": 0.00010748796921689063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.290601295999295
        },
        {
          "epoch": 1061,
          "updates": 1061,
          "samples_seen": 1086464,
          "online_loss": 0.0011935755610466003,
          "train_loss": 0.0011875529307872057,
          "validation_loss": 0.0007960593793541193,
          "train_metric": 0.00011960598931182176,
          "validation_metric": 8.017618529265746e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.295754385999317
        },
        {
          "epoch": 1062,
          "updates": 1062,
          "samples_seen": 1087488,
          "online_loss": 0.0011875530472025275,
          "train_loss": 0.0010847829980775714,
          "validation_loss": 0.0009313488262705505,
          "train_metric": 0.00010925537208095193,
          "validation_metric": 9.380203846376389e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.30081269099901
        },
        {
          "epoch": 1063,
          "updates": 1063,
          "samples_seen": 1088512,
          "online_loss": 0.0010847831144928932,
          "train_loss": 0.0010111038573086262,
          "validation_loss": 0.0007604066049680114,
          "train_metric": 0.00010183468111790717,
          "validation_metric": 7.65853765187785e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.306456577998688
        },
        {
          "epoch": 1064,
          "updates": 1064,
          "samples_seen": 1089536,
          "online_loss": 0.0010111038573086262,
          "train_loss": 0.0010168482549488544,
          "validation_loss": 0.0007559412624686956,
          "train_metric": 0.00010241323616355658,
          "validation_metric": 7.613562775077298e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.312669442999322
        },
        {
          "epoch": 1065,
          "updates": 1065,
          "samples_seen": 1090560,
          "online_loss": 0.0010168483713641763,
          "train_loss": 0.001072445884346962,
          "validation_loss": 0.0009191544959321618,
          "train_metric": 0.00010801281314343214,
          "validation_metric": 9.257386409444734e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.31775416599885
        },
        {
          "epoch": 1066,
          "updates": 1066,
          "samples_seen": 1091584,
          "online_loss": 0.001072445884346962,
          "train_loss": 0.0011120385024696589,
          "validation_loss": 0.0007718590786680579,
          "train_metric": 0.00011200043809367344,
          "validation_metric": 7.773881225148216e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.32280005499888
        },
        {
          "epoch": 1067,
          "updates": 1067,
          "samples_seen": 1092608,
          "online_loss": 0.0011120382696390152,
          "train_loss": 0.0010797119466587901,
          "validation_loss": 0.0009366384474560618,
          "train_metric": 0.0001087446216843091,
          "validation_metric": 9.433479863218963e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.328025996999713
        },
        {
          "epoch": 1068,
          "updates": 1068,
          "samples_seen": 1093632,
          "online_loss": 0.0010797117138281465,
          "train_loss": 0.0010255356319248676,
          "validation_loss": 0.000758980808313936,
          "train_metric": 0.00010328822099836543,
          "validation_metric": 7.644177094334736e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.333008423998763
        },
        {
          "epoch": 1069,
          "updates": 1069,
          "samples_seen": 1094656,
          "online_loss": 0.0010255358647555113,
          "train_loss": 0.0009965675417333841,
          "validation_loss": 0.0007883929647505283,
          "train_metric": 0.00010037062020273879,
          "validation_metric": 7.94040533946827e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.338232428999618
        },
        {
          "epoch": 1070,
          "updates": 1070,
          "samples_seen": 1095680,
          "online_loss": 0.0009965675417333841,
          "train_loss": 0.0010140048107132316,
          "validation_loss": 0.0008379749488085508,
          "train_metric": 0.0001021268471959047,
          "validation_metric": 8.439776865998283e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.343265193998377
        },
        {
          "epoch": 1071,
          "updates": 1071,
          "samples_seen": 1096704,
          "online_loss": 0.0010140048107132316,
          "train_loss": 0.0010469458065927029,
          "validation_loss": 0.0007482971996068954,
          "train_metric": 0.00010544455290073529,
          "validation_metric": 7.536575867561623e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.348880323999765
        },
        {
          "epoch": 1072,
          "updates": 1072,
          "samples_seen": 1097728,
          "online_loss": 0.0010469458065927029,
          "train_loss": 0.001049601472914219,
          "validation_loss": 0.0008942647255025804,
          "train_metric": 0.00010571201710263267,
          "validation_metric": 9.006707114167511e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.354486956999608
        },
        {
          "epoch": 1073,
          "updates": 1073,
          "samples_seen": 1098752,
          "online_loss": 0.001049601472914219,
          "train_loss": 0.0010248338803648949,
          "validation_loss": 0.0007483666995540261,
          "train_metric": 0.0001032175132422708,
          "validation_metric": 7.537275814684108e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.359595838999667
        },
        {
          "epoch": 1074,
          "updates": 1074,
          "samples_seen": 1099776,
          "online_loss": 0.0010248338803648949,
          "train_loss": 0.0009958763839676976,
          "validation_loss": 0.0008059949614107609,
          "train_metric": 0.0001003010111162439,
          "validation_metric": 8.117686957120895e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.364621069998975
        },
        {
          "epoch": 1075,
          "updates": 1075,
          "samples_seen": 1100800,
          "online_loss": 0.0009958765003830194,
          "train_loss": 0.0009916890412569046,
          "validation_loss": 0.000796181324403733,
          "train_metric": 9.987928933696821e-05,
          "validation_metric": 8.018847438506782e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.369751548998465
        },
        {
          "epoch": 1076,
          "updates": 1076,
          "samples_seen": 1101824,
          "online_loss": 0.0009916889248415828,
          "train_loss": 0.0010081836953759193,
          "validation_loss": 0.0007525021792389452,
          "train_metric": 0.00010154056508326903,
          "validation_metric": 7.578927034046501e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.374856645999898
        },
        {
          "epoch": 1077,
          "updates": 1077,
          "samples_seen": 1102848,
          "online_loss": 0.0010081838117912412,
          "train_loss": 0.0010203539859503508,
          "validation_loss": 0.0008575848769396544,
          "train_metric": 0.00010276632383465767,
          "validation_metric": 8.637280552648008e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.379968604998794
        },
        {
          "epoch": 1078,
          "updates": 1078,
          "samples_seen": 1103872,
          "online_loss": 0.0010203539859503508,
          "train_loss": 0.001015400281175971,
          "validation_loss": 0.0007451463025063276,
          "train_metric": 0.00010226738959318027,
          "validation_metric": 7.50484032323584e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.385822892998476
        },
        {
          "epoch": 1079,
          "updates": 1079,
          "samples_seen": 1104896,
          "online_loss": 0.0010154003975912929,
          "train_loss": 0.000996527262032032,
          "validation_loss": 0.0008133201627060771,
          "train_metric": 0.00010036656749434769,
          "validation_metric": 8.19146225694567e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.391154006998477
        },
        {
          "epoch": 1080,
          "updates": 1080,
          "samples_seen": 1105920,
          "online_loss": 0.000996527262032032,
          "train_loss": 0.0009841189021244645,
          "validation_loss": 0.0007643376011401415,
          "train_metric": 9.911684901453555e-05,
          "validation_metric": 7.698127592448145e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.406408991999342
        },
        {
          "epoch": 1081,
          "updates": 1081,
          "samples_seen": 1106944,
          "online_loss": 0.0009841187857091427,
          "train_loss": 0.0009864232270047069,
          "validation_loss": 0.000754520355258137,
          "train_metric": 9.934894478647038e-05,
          "validation_metric": 7.599253149237484e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.580622035999113
        },
        {
          "epoch": 1082,
          "updates": 1082,
          "samples_seen": 1107968,
          "online_loss": 0.0009864233434200287,
          "train_loss": 0.00099585996940732,
          "validation_loss": 0.0008181051816791296,
          "train_metric": 0.00010029935947386548,
          "validation_metric": 8.239656017394736e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.59264190399881
        },
        {
          "epoch": 1083,
          "updates": 1083,
          "samples_seen": 1108992,
          "online_loss": 0.0009958597365766764,
          "train_loss": 0.0010001453338190913,
          "validation_loss": 0.00074435566784814,
          "train_metric": 0.00010073097655549645,
          "validation_metric": 7.496876787627116e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.605267044998982
        },
        {
          "epoch": 1084,
          "updates": 1084,
          "samples_seen": 1110016,
          "online_loss": 0.0010001453338190913,
          "train_loss": 0.0009930133819580078,
          "validation_loss": 0.0008167334599420428,
          "train_metric": 0.00010001267946790904,
          "validation_metric": 8.225840429076925e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.615135147998444
        },
        {
          "epoch": 1085,
          "updates": 1085,
          "samples_seen": 1111040,
          "online_loss": 0.0009930134983733296,
          "train_loss": 0.0009823436848819256,
          "validation_loss": 0.0007530803559347987,
          "train_metric": 9.893807145999745e-05,
          "validation_metric": 7.58474925532937e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.62059754099937
        },
        {
          "epoch": 1086,
          "updates": 1086,
          "samples_seen": 1112064,
          "online_loss": 0.0009823436848819256,
          "train_loss": 0.0009765827562659979,
          "validation_loss": 0.0007686395547352731,
          "train_metric": 9.835783566813916e-05,
          "validation_metric": 7.741455192444846e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.625711844999387
        },
        {
          "epoch": 1087,
          "updates": 1087,
          "samples_seen": 1113088,
          "online_loss": 0.0009765827562659979,
          "train_loss": 0.0009785296861082315,
          "validation_loss": 0.0007846408989280462,
          "train_metric": 9.855393727775663e-05,
          "validation_metric": 7.902615470811725e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.63087735299996
        },
        {
          "epoch": 1088,
          "updates": 1088,
          "samples_seen": 1114112,
          "online_loss": 0.0009785298025235534,
          "train_loss": 0.0009832482319325209,
          "validation_loss": 0.0007429469260387123,
          "train_metric": 9.902915917336941e-05,
          "validation_metric": 7.482690125470981e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.636661597998682
        },
        {
          "epoch": 1089,
          "updates": 1089,
          "samples_seen": 1115136,
          "online_loss": 0.0009832482319325209,
          "train_loss": 0.0009839392732828856,
          "validation_loss": 0.0008036107756197453,
          "train_metric": 9.909875370794907e-05,
          "validation_metric": 8.093672659015283e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.64193240699933
        },
        {
          "epoch": 1090,
          "updates": 1090,
          "samples_seen": 1116160,
          "online_loss": 0.0009839392732828856,
          "train_loss": 0.000979586155153811,
          "validation_loss": 0.0007441177731379867,
          "train_metric": 9.866033360594884e-05,
          "validation_metric": 7.494482269976288e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.647072013998695
        },
        {
          "epoch": 1091,
          "updates": 1091,
          "samples_seen": 1117184,
          "online_loss": 0.000979586155153811,
          "train_loss": 0.0009732242906466126,
          "validation_loss": 0.0007811984396539629,
          "train_metric": 9.8019590950571e-05,
          "validation_metric": 7.867944077588618e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.652257606998319
        },
        {
          "epoch": 1092,
          "updates": 1092,
          "samples_seen": 1118208,
          "online_loss": 0.0009732242324389517,
          "train_loss": 0.0009696089546196163,
          "validation_loss": 0.0007630594773218036,
          "train_metric": 9.765546565176919e-05,
          "validation_metric": 7.685254968237132e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.657423947999632
        },
        {
          "epoch": 1093,
          "updates": 1093,
          "samples_seen": 1119232,
          "online_loss": 0.0009696088382042944,
          "train_loss": 0.00096992525504902,
          "validation_loss": 0.000752790947444737,
          "train_metric": 9.76873270701617e-05,
          "validation_metric": 7.581833779113367e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.662517119999393
        },
        {
          "epoch": 1094,
          "updates": 1094,
          "samples_seen": 1120256,
          "online_loss": 0.0009699253714643419,
          "train_loss": 0.0009718179935589433,
          "validation_loss": 0.0007850578404031694,
          "train_metric": 9.787794988369569e-05,
          "validation_metric": 7.90681442595087e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.667599967999195
        },
        {
          "epoch": 1095,
          "updates": 1095,
          "samples_seen": 1121280,
          "online_loss": 0.0009718179935589433,
          "train_loss": 0.0009722347604110837,
          "validation_loss": 0.0007412866107188165,
          "train_metric": 9.79199176072143e-05,
          "validation_metric": 7.465967064490542e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.673400201998447
        },
        {
          "epoch": 1096,
          "updates": 1096,
          "samples_seen": 1122304,
          "online_loss": 0.0009722347604110837,
          "train_loss": 0.0009697283967398107,
          "validation_loss": 0.0007824277272447944,
          "train_metric": 9.766749280970544e-05,
          "validation_metric": 7.880324847064912e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.678629409998393
        },
        {
          "epoch": 1097,
          "updates": 1097,
          "samples_seen": 1123328,
          "online_loss": 0.0009697284549474716,
          "train_loss": 0.0009658868657425046,
          "validation_loss": 0.0007456337916664779,
          "train_metric": 9.7280579211656e-05,
          "validation_metric": 7.50975013943389e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.683497082998656
        },
        {
          "epoch": 1098,
          "updates": 1098,
          "samples_seen": 1124352,
          "online_loss": 0.0009658868075348437,
          "train_loss": 0.000962828635238111,
          "validation_loss": 0.0007615529466420412,
          "train_metric": 9.697257337393239e-05,
          "validation_metric": 7.670080958632752e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.688678585998787
        },
        {
          "epoch": 1099,
          "updates": 1099,
          "samples_seen": 1125376,
          "online_loss": 0.0009628287516534328,
          "train_loss": 0.0009617647156119347,
          "validation_loss": 0.0007622464327141643,
          "train_metric": 9.686542034614831e-05,
          "validation_metric": 7.677068060729653e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.693724405999092
        },
        {
          "epoch": 1100,
          "updates": 1100,
          "samples_seen": 1126400,
          "online_loss": 0.0009617647156119347,
          "train_loss": 0.000962079269811511,
          "validation_loss": 0.0007455560844391584,
          "train_metric": 9.689710714155808e-05,
          "validation_metric": 7.508967973990366e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.698776824998276
        },
        {
          "epoch": 1101,
          "updates": 1101,
          "samples_seen": 1127424,
          "online_loss": 0.000962079269811511,
          "train_loss": 0.0009622141369618475,
          "validation_loss": 0.0007743029855191708,
          "train_metric": 9.691067680250853e-05,
          "validation_metric": 7.798495789756998e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.703900141999839
        },
        {
          "epoch": 1102,
          "updates": 1102,
          "samples_seen": 1128448,
          "online_loss": 0.0009622141951695085,
          "train_loss": 0.0009611716377548873,
          "validation_loss": 0.0007404795615002513,
          "train_metric": 9.680569201009348e-05,
          "validation_metric": 7.457838364643976e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.709488440999849
        },
        {
          "epoch": 1103,
          "updates": 1103,
          "samples_seen": 1129472,
          "online_loss": 0.0009611716959625483,
          "train_loss": 0.0009589139954186976,
          "validation_loss": 0.0007678879192098975,
          "train_metric": 9.657830378273502e-05,
          "validation_metric": 7.733886013738811e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.714534166998419
        },
        {
          "epoch": 1104,
          "updates": 1104,
          "samples_seen": 1130496,
          "online_loss": 0.0009589139372110367,
          "train_loss": 0.0009564274223521352,
          "validation_loss": 0.0007443432114087045,
          "train_metric": 9.63278507697396e-05,
          "validation_metric": 7.496752368751913e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.719837748998543
        },
        {
          "epoch": 1105,
          "updates": 1105,
          "samples_seen": 1131520,
          "online_loss": 0.0009564273641444743,
          "train_loss": 0.0009545786306262016,
          "validation_loss": 0.0007522574160248041,
          "train_metric": 9.614166629035026e-05,
          "validation_metric": 7.576461939606816e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.724953753999216
        },
        {
          "epoch": 1106,
          "updates": 1106,
          "samples_seen": 1132544,
          "online_loss": 0.0009545787470415235,
          "train_loss": 0.00095365074230358,
          "validation_loss": 0.0007551340386271477,
          "train_metric": 9.604821389075369e-05,
          "validation_metric": 7.605434802826494e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.730038024999885
        },
        {
          "epoch": 1107,
          "updates": 1107,
          "samples_seen": 1133568,
          "online_loss": 0.0009536505676805973,
          "train_loss": 0.000953255279455334,
          "validation_loss": 0.0007415266591124237,
          "train_metric": 9.600837802281603e-05,
          "validation_metric": 7.468384865205735e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.735445496999091
        },
        {
          "epoch": 1108,
          "updates": 1108,
          "samples_seen": 1134592,
          "online_loss": 0.0009532551048323512,
          "train_loss": 0.0009527084766887128,
          "validation_loss": 0.0007624315330758691,
          "train_metric": 9.59533208515495e-05,
          "validation_metric": 7.678930705878884e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.740649040999415
        },
        {
          "epoch": 1109,
          "updates": 1109,
          "samples_seen": 1135616,
          "online_loss": 0.0009527084766887128,
          "train_loss": 0.0009516266291029751,
          "validation_loss": 0.0007382403709925711,
          "train_metric": 9.584434155840427e-05,
          "validation_metric": 7.435286534018815e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.746247692999532
        },
        {
          "epoch": 1110,
          "updates": 1110,
          "samples_seen": 1136640,
          "online_loss": 0.0009516266291029751,
          "train_loss": 0.000949987443163991,
          "validation_loss": 0.0007587011205032468,
          "train_metric": 9.567925735609606e-05,
          "validation_metric": 7.641360571142286e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.751523668999653
        },
        {
          "epoch": 1111,
          "updates": 1111,
          "samples_seen": 1137664,
          "online_loss": 0.000949987443163991,
          "train_loss": 0.0009481878369115293,
          "validation_loss": 0.0007403697236441076,
          "train_metric": 9.549799870001152e-05,
          "validation_metric": 7.456733874278143e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.756709532999594
        },
        {
          "epoch": 1112,
          "updates": 1112,
          "samples_seen": 1138688,
          "online_loss": 0.0009481877787038684,
          "train_loss": 0.000946601910982281,
          "validation_loss": 0.0007479824125766754,
          "train_metric": 9.533826960250735e-05,
          "validation_metric": 7.53340355004184e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.761812920000011
        },
        {
          "epoch": 1113,
          "updates": 1113,
          "samples_seen": 1139712,
          "online_loss": 0.0009466017945669591,
          "train_loss": 0.0009454145329073071,
          "validation_loss": 0.0007460747729055583,
          "train_metric": 9.521867468720302e-05,
          "validation_metric": 7.514192839153111e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.767013189999489
        },
        {
          "epoch": 1114,
          "updates": 1114,
          "samples_seen": 1140736,
          "online_loss": 0.0009454145329073071,
          "train_loss": 0.000944544852245599,
          "validation_loss": 0.0007390081882476807,
          "train_metric": 9.513108670944348e-05,
          "validation_metric": 7.44302014936693e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.772242273998927
        },
        {
          "epoch": 1115,
          "updates": 1115,
          "samples_seen": 1141760,
          "online_loss": 0.0009445449104532599,
          "train_loss": 0.0009437466505914927,
          "validation_loss": 0.0007510256255045533,
          "train_metric": 9.505069465376437e-05,
          "validation_metric": 7.564056431874633e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.777360644999135
        },
        {
          "epoch": 1116,
          "updates": 1116,
          "samples_seen": 1142784,
          "online_loss": 0.0009437465341761708,
          "train_loss": 0.0009428078774362803,
          "validation_loss": 0.0007348806830123067,
          "train_metric": 9.495615086052567e-05,
          "validation_metric": 7.401448965538293e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.783613893998336
        },
        {
          "epoch": 1117,
          "updates": 1117,
          "samples_seen": 1143808,
          "online_loss": 0.0009428078192286193,
          "train_loss": 0.000941617414355278,
          "validation_loss": 0.0007505462854169309,
          "train_metric": 9.483626490691677e-05,
          "validation_metric": 7.559227378806099e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.788878348999788
        },
        {
          "epoch": 1118,
          "updates": 1118,
          "samples_seen": 1144832,
          "online_loss": 0.000941617414355278,
          "train_loss": 0.0009402481955476105,
          "validation_loss": 0.0007345142657868564,
          "train_metric": 9.469835640629753e-05,
          "validation_metric": 7.397758599836379e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.794525784998768
        },
        {
          "epoch": 1119,
          "updates": 1119,
          "samples_seen": 1145856,
          "online_loss": 0.0009402481373399496,
          "train_loss": 0.0009388207690790296,
          "validation_loss": 0.0007446369854733348,
          "train_metric": 9.455456893192604e-05,
          "validation_metric": 7.499712228309363e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.90540779299954
        },
        {
          "epoch": 1120,
          "updates": 1120,
          "samples_seen": 1146880,
          "online_loss": 0.0009388207108713686,
          "train_loss": 0.0009374820510856807,
          "validation_loss": 0.0007364820339716971,
          "train_metric": 9.441975271329284e-05,
          "validation_metric": 7.417577580781654e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.913836998999614
        },
        {
          "epoch": 1121,
          "updates": 1121,
          "samples_seen": 1147904,
          "online_loss": 0.0009374821092933416,
          "train_loss": 0.0009362918208353221,
          "validation_loss": 0.0007374462438747287,
          "train_metric": 9.429988858755678e-05,
          "validation_metric": 7.427288801409304e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.92285974099832
        },
        {
          "epoch": 1122,
          "updates": 1122,
          "samples_seen": 1148928,
          "online_loss": 0.0009362918790429831,
          "train_loss": 0.0009352356428280473,
          "validation_loss": 0.000739552837330848,
          "train_metric": 9.41935068112798e-05,
          "validation_metric": 7.448504766216502e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.932215963999624
        },
        {
          "epoch": 1123,
          "updates": 1123,
          "samples_seen": 1149952,
          "online_loss": 0.0009352355846203864,
          "train_loss": 0.0009342454723082483,
          "validation_loss": 0.0007322643650695682,
          "train_metric": 9.409378981217742e-05,
          "validation_metric": 7.375099085038528e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.94206029399902
        },
        {
          "epoch": 1124,
          "updates": 1124,
          "samples_seen": 1150976,
          "online_loss": 0.0009342455305159092,
          "train_loss": 0.0009332407498732209,
          "validation_loss": 0.0007411585538648069,
          "train_metric": 9.399259579367936e-05,
          "validation_metric": 7.464677037205547e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.951734798998586
        },
        {
          "epoch": 1125,
          "updates": 1125,
          "samples_seen": 1152000,
          "online_loss": 0.0009332407498732209,
          "train_loss": 0.0009321683319285512,
          "validation_loss": 0.0007294653332792222,
          "train_metric": 9.388459147885442e-05,
          "validation_metric": 7.346906932070851e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.962207089000003
        },
        {
          "epoch": 1126,
          "updates": 1126,
          "samples_seen": 1153024,
          "online_loss": 0.0009321683901362121,
          "train_loss": 0.0009310101740993559,
          "validation_loss": 0.0007397018489427865,
          "train_metric": 9.376794332638383e-05,
          "validation_metric": 7.450005796272308e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.97107651299848
        },
        {
          "epoch": 1127,
          "updates": 1127,
          "samples_seen": 1154048,
          "online_loss": 0.0009310102323070168,
          "train_loss": 0.0009297958458773792,
          "validation_loss": 0.0007284120074473321,
          "train_metric": 9.36456344788894e-05,
          "validation_metric": 7.336299313465133e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.980858675999116
        },
        {
          "epoch": 1128,
          "updates": 1128,
          "samples_seen": 1155072,
          "online_loss": 0.0009297958458773792,
          "train_loss": 0.0009285609703511,
          "validation_loss": 0.0007357989670708776,
          "train_metric": 9.352125925943255e-05,
          "validation_metric": 7.410698162857443e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.989122299999508
        },
        {
          "epoch": 1129,
          "updates": 1129,
          "samples_seen": 1156096,
          "online_loss": 0.0009285609703511,
          "train_loss": 0.0009273383766412735,
          "validation_loss": 0.0007288086926564574,
          "train_metric": 9.339812822872773e-05,
          "validation_metric": 7.34029381419532e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.997595726999862
        },
        {
          "epoch": 1130,
          "updates": 1130,
          "samples_seen": 1157120,
          "online_loss": 0.0009273384348489344,
          "train_loss": 0.000926161534152925,
          "validation_loss": 0.0007314248941838741,
          "train_metric": 9.327959560323507e-05,
          "validation_metric": 7.366643694695085e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.006618252999033
        },
        {
          "epoch": 1131,
          "updates": 1131,
          "samples_seen": 1158144,
          "online_loss": 0.000926161534152925,
          "train_loss": 0.0009250273578800261,
          "validation_loss": 0.0007300101569853723,
          "train_metric": 9.316537034465e-05,
          "validation_metric": 7.352395914494991e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.01542240299932
        },
        {
          "epoch": 1132,
          "updates": 1132,
          "samples_seen": 1159168,
          "online_loss": 0.000925027416087687,
          "train_loss": 0.000923929619602859,
          "validation_loss": 0.000727589416783303,
          "train_metric": 9.305481216870248e-05,
          "validation_metric": 7.328014180529863e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.024936776999311
        },
        {
          "epoch": 1133,
          "updates": 1133,
          "samples_seen": 1160192,
          "online_loss": 0.0009239296778105199,
          "train_loss": 0.000922858074773103,
          "validation_loss": 0.0007307118503376842,
          "train_metric": 9.294687333749607e-05,
          "validation_metric": 7.359462324529886e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.033538611998665
        },
        {
          "epoch": 1134,
          "updates": 1134,
          "samples_seen": 1161216,
          "online_loss": 0.0009228580165654421,
          "train_loss": 0.000921787868719548,
          "validation_loss": 0.0007243974832817912,
          "train_metric": 9.283909457735717e-05,
          "validation_metric": 7.295866089407355e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.043422653998277
        },
        {
          "epoch": 1135,
          "updates": 1135,
          "samples_seen": 1162240,
          "online_loss": 0.0009217877523042262,
          "train_loss": 0.0009207032853737473,
          "validation_loss": 0.0007301164441742003,
          "train_metric": 9.272986062569544e-05,
          "validation_metric": 7.353466207860038e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.051831022999977
        },
        {
          "epoch": 1136,
          "updates": 1136,
          "samples_seen": 1163264,
          "online_loss": 0.0009207032853737473,
          "train_loss": 0.0009196095634251833,
          "validation_loss": 0.0007219576509669423,
          "train_metric": 9.261970990337431e-05,
          "validation_metric": 7.271292997756973e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.066172463999465
        },
        {
          "epoch": 1137,
          "updates": 1137,
          "samples_seen": 1164288,
          "online_loss": 0.0009196095634251833,
          "train_loss": 0.0009184916852973402,
          "validation_loss": 0.000728567480109632,
          "train_metric": 9.250711445929483e-05,
          "validation_metric": 7.337865099543706e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.075738422998256
        },
        {
          "epoch": 1138,
          "updates": 1138,
          "samples_seen": 1165312,
          "online_loss": 0.0009184916270896792,
          "train_loss": 0.0009173614671453834,
          "validation_loss": 0.0007205461733974516,
          "train_metric": 9.239329665433615e-05,
          "validation_metric": 7.257077959366143e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.084372776998862
        },
        {
          "epoch": 1139,
          "updates": 1139,
          "samples_seen": 1166336,
          "online_loss": 0.0009173615253530443,
          "train_loss": 0.000916222226805985,
          "validation_loss": 0.0007265409221872687,
          "train_metric": 9.227854025084525e-05,
          "validation_metric": 7.317453855648637e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.109143831999972
        },
        {
          "epoch": 1140,
          "updates": 1140,
          "samples_seen": 1167360,
          "online_loss": 0.000916222226805985,
          "train_loss": 0.0009150836849585176,
          "validation_loss": 0.0007197707309387624,
          "train_metric": 9.216387115884572e-05,
          "validation_metric": 7.249267218867317e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.13009559099919
        },
        {
          "epoch": 1141,
          "updates": 1141,
          "samples_seen": 1168384,
          "online_loss": 0.0009150836849585176,
          "train_loss": 0.000913940486498177,
          "validation_loss": 0.0007241257699206471,
          "train_metric": 9.204873640555888e-05,
          "validation_metric": 7.293128874152899e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.147931345998586
        },
        {
          "epoch": 1142,
          "updates": 1142,
          "samples_seen": 1169408,
          "online_loss": 0.000913940486498177,
          "train_loss": 0.0009128074161708355,
          "validation_loss": 0.0007190475007519126,
          "train_metric": 9.193462756229565e-05,
          "validation_metric": 7.241982530103996e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.168426937998447
        },
        {
          "epoch": 1143,
          "updates": 1143,
          "samples_seen": 1170432,
          "online_loss": 0.0009128074161708355,
          "train_loss": 0.0009116779547184706,
          "validation_loss": 0.0007213930948637426,
          "train_metric": 9.182086068904027e-05,
          "validation_metric": 7.26560756447725e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.190828077998958
        },
        {
          "epoch": 1144,
          "updates": 1144,
          "samples_seen": 1171456,
          "online_loss": 0.0009116780129261315,
          "train_loss": 0.0009105599019676447,
          "validation_loss": 0.0007181731052696705,
          "train_metric": 9.170825069304556e-05,
          "validation_metric": 7.23317643860355e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.21012492599948
        },
        {
          "epoch": 1145,
          "updates": 1145,
          "samples_seen": 1172480,
          "online_loss": 0.0009105599019676447,
          "train_loss": 0.0009094458655454218,
          "validation_loss": 0.0007187743904069066,
          "train_metric": 9.159606270259246e-05,
          "validation_metric": 7.239234400913119e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.229673645999355
        },
        {
          "epoch": 1146,
          "updates": 1146,
          "samples_seen": 1173504,
          "online_loss": 0.0009094463312067091,
          "train_loss": 0.0009083372424356639,
          "validation_loss": 0.0007173471967689693,
          "train_metric": 9.148439130512998e-05,
          "validation_metric": 7.224859291454777e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.25009383899851
        },
        {
          "epoch": 1147,
          "updates": 1147,
          "samples_seen": 1174528,
          "online_loss": 0.000908337184228003,
          "train_loss": 0.0009072343818843365,
          "validation_loss": 0.0007165014394558966,
          "train_metric": 9.137333108810708e-05,
          "validation_metric": 7.216340600280091e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.270741423999425
        },
        {
          "epoch": 1148,
          "updates": 1148,
          "samples_seen": 1175552,
          "online_loss": 0.0009072345565073192,
          "train_loss": 0.0009061340242624283,
          "validation_loss": 0.0007165170973166823,
          "train_metric": 9.126251097768545e-05,
          "validation_metric": 7.216497760964558e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.288122191999719
        },
        {
          "epoch": 1149,
          "updates": 1149,
          "samples_seen": 1176576,
          "online_loss": 0.0009061339660547674,
          "train_loss": 0.0009050396038219333,
          "validation_loss": 0.0007143809925764799,
          "train_metric": 9.115228021983057e-05,
          "validation_metric": 7.19498420949094e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.297811976999583
        },
        {
          "epoch": 1150,
          "updates": 1150,
          "samples_seen": 1177600,
          "online_loss": 0.0009050394874066114,
          "train_loss": 0.0009039455326274037,
          "validation_loss": 0.0007154698250815272,
          "train_metric": 9.104207128984854e-05,
          "validation_metric": 7.205949077615514e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.30562951399952
        },
        {
          "epoch": 1151,
          "updates": 1151,
          "samples_seen": 1178624,
          "online_loss": 0.0009039455326274037,
          "train_loss": 0.0009028527420014143,
          "validation_loss": 0.0007121904054656625,
          "train_metric": 9.09320151549764e-05,
          "validation_metric": 7.172921323217452e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.31400538899834
        },
        {
          "epoch": 1152,
          "updates": 1152,
          "samples_seen": 1179648,
          "online_loss": 0.0009028527420014143,
          "train_loss": 0.0009017638512887061,
          "validation_loss": 0.0007142710965126753,
          "train_metric": 9.082234464585781e-05,
          "validation_metric": 7.193876808742061e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.322110169998268
        },
        {
          "epoch": 1153,
          "updates": 1153,
          "samples_seen": 1180672,
          "online_loss": 0.0009017639677040279,
          "train_loss": 0.0009006800246424973,
          "validation_loss": 0.0007100271759554744,
          "train_metric": 9.071319800568745e-05,
          "validation_metric": 7.151133468141779e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.330392625999593
        },
        {
          "epoch": 1154,
          "updates": 1154,
          "samples_seen": 1181696,
          "online_loss": 0.0009006799664348364,
          "train_loss": 0.0008995984680950642,
          "validation_loss": 0.0007132159662432969,
          "train_metric": 9.06042696442455e-05,
          "validation_metric": 7.183250272646546e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.338465517999794
        },
        {
          "epoch": 1155,
          "updates": 1155,
          "samples_seen": 1182720,
          "online_loss": 0.0008995984680950642,
          "train_loss": 0.0008985215099528432,
          "validation_loss": 0.0007079026545397937,
          "train_metric": 9.049579239217564e-05,
          "validation_metric": 7.129736331989989e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.347321465998903
        },
        {
          "epoch": 1156,
          "updates": 1156,
          "samples_seen": 1183744,
          "online_loss": 0.0008985213935375214,
          "train_loss": 0.0008974524098448455,
          "validation_loss": 0.0007124035037122667,
          "train_metric": 9.038813004735857e-05,
          "validation_metric": 7.175066275522113e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.354900436999742
        },
        {
          "epoch": 1157,
          "updates": 1157,
          "samples_seen": 1184768,
          "online_loss": 0.0008974525262601674,
          "train_loss": 0.0008963960572145879,
          "validation_loss": 0.0007055828464217484,
          "train_metric": 9.028172644320875e-05,
          "validation_metric": 7.106371776899323e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.373144364999462
        },
        {
          "epoch": 1158,
          "updates": 1158,
          "samples_seen": 1185792,
          "online_loss": 0.0008963961154222488,
          "train_loss": 0.0008953530341386795,
          "validation_loss": 0.0007118523353710771,
          "train_metric": 9.017666889121756e-05,
          "validation_metric": 7.169515447458252e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.391019652999603
        },
        {
          "epoch": 1159,
          "updates": 1159,
          "samples_seen": 1186816,
          "online_loss": 0.0008953530341386795,
          "train_loss": 0.000894336320925504,
          "validation_loss": 0.0007027907413430512,
          "train_metric": 9.007427433971316e-05,
          "validation_metric": 7.078251655912027e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.608213112998783
        },
        {
          "epoch": 1160,
          "updates": 1160,
          "samples_seen": 1187840,
          "online_loss": 0.0008943362627178431,
          "train_loss": 0.0008933587232604623,
          "validation_loss": 0.0007118418579921126,
          "train_metric": 8.997581608127803e-05,
          "validation_metric": 7.16941140126437e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.626984460999665
        },
        {
          "epoch": 1161,
          "updates": 1161,
          "samples_seen": 1188864,
          "online_loss": 0.0008933587232604623,
          "train_loss": 0.0008924372377805412,
          "validation_loss": 0.0006994220893830061,
          "train_metric": 8.988301124190912e-05,
          "validation_metric": 7.044323137961328e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.641564004999964
        },
        {
          "epoch": 1162,
          "updates": 1162,
          "samples_seen": 1189888,
          "online_loss": 0.0008924371213652194,
          "train_loss": 0.000891607254743576,
          "validation_loss": 0.0007129892474040389,
          "train_metric": 8.979941048892215e-05,
          "validation_metric": 7.180967077147216e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.654607661999762
        },
        {
          "epoch": 1163,
          "updates": 1163,
          "samples_seen": 1190912,
          "online_loss": 0.0008916071965359151,
          "train_loss": 0.000890929892193526,
          "validation_loss": 0.0006952111143618822,
          "train_metric": 8.973119838628918e-05,
          "validation_metric": 7.001913036219776e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.668030064998675
        },
        {
          "epoch": 1164,
          "updates": 1164,
          "samples_seen": 1191936,
          "online_loss": 0.0008909299504011869,
          "train_loss": 0.0008904855931177735,
          "validation_loss": 0.0007162174442782998,
          "train_metric": 8.968645852291957e-05,
          "validation_metric": 7.213480421341956e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.679432125998574
        },
        {
          "epoch": 1165,
          "updates": 1165,
          "samples_seen": 1192960,
          "online_loss": 0.0008904856513254344,
          "train_loss": 0.0008904649876058102,
          "validation_loss": 0.0006896844715811312,
          "train_metric": 8.968436304712668e-05,
          "validation_metric": 6.946249050088227e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.689377030999822
        },
        {
          "epoch": 1166,
          "updates": 1166,
          "samples_seen": 1193984,
          "online_loss": 0.0008904648711904883,
          "train_loss": 0.0008910702890716493,
          "validation_loss": 0.0007234045770019293,
          "train_metric": 8.974535012384877e-05,
          "validation_metric": 7.285866740858182e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.703473202998794
        },
        {
          "epoch": 1167,
          "updates": 1167,
          "samples_seen": 1195008,
          "online_loss": 0.0008910702890716493,
          "train_loss": 0.000892902840860188,
          "validation_loss": 0.0006825856398791075,
          "train_metric": 8.99298902368173e-05,
          "validation_metric": 6.874751852592453e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.71716109099907
        },
        {
          "epoch": 1168,
          "updates": 1168,
          "samples_seen": 1196032,
          "online_loss": 0.0008929026080295444,
          "train_loss": 0.0008964769076555967,
          "validation_loss": 0.0007389214006252587,
          "train_metric": 9.028986096382141e-05,
          "validation_metric": 7.442146306857467e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.728771929998402
        },
        {
          "epoch": 1169,
          "updates": 1169,
          "samples_seen": 1197056,
          "online_loss": 0.0008964768494479358,
          "train_loss": 0.0009037843556143343,
          "validation_loss": 0.0006753000780008733,
          "train_metric": 9.102585318032652e-05,
          "validation_metric": 6.801374547649175e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.739133405999382
        },
        {
          "epoch": 1170,
          "updates": 1170,
          "samples_seen": 1198080,
          "online_loss": 0.0009037844138219953,
          "train_loss": 0.0009158637840300798,
          "validation_loss": 0.0007731664227321744,
          "train_metric": 9.22424515010789e-05,
          "validation_metric": 7.787048525642604e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.748485063999397
        },
        {
          "epoch": 1171,
          "updates": 1171,
          "samples_seen": 1199104,
          "online_loss": 0.0009158637840300798,
          "train_loss": 0.0009394361404702067,
          "validation_loss": 0.0006753205670975149,
          "train_metric": 9.46165673667565e-05,
          "validation_metric": 6.801581912441179e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.757194342999355
        },
        {
          "epoch": 1172,
          "updates": 1172,
          "samples_seen": 1200128,
          "online_loss": 0.0009394362568855286,
          "train_loss": 0.0009743171976879239,
          "validation_loss": 0.000849872943945229,
          "train_metric": 9.812965436140075e-05,
          "validation_metric": 8.559608977520838e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.76581291399998
        },
        {
          "epoch": 1173,
          "updates": 1173,
          "samples_seen": 1201152,
          "online_loss": 0.0009743171976879239,
          "train_loss": 0.0010418922174721956,
          "validation_loss": 0.0007079417700879276,
          "train_metric": 0.00010493558511370793,
          "validation_metric": 7.130129233701155e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.774632368998937
        },
        {
          "epoch": 1174,
          "updates": 1174,
          "samples_seen": 1202176,
          "online_loss": 0.0010418922174721956,
          "train_loss": 0.0011205755872651935,
          "validation_loss": 0.0010077779879793525,
          "train_metric": 0.00011286025983281434,
          "validation_metric": 0.00010149971058126539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.782888707999518
        },
        {
          "epoch": 1175,
          "updates": 1175,
          "samples_seen": 1203200,
          "online_loss": 0.0011205755872651935,
          "train_loss": 0.0012631784193217754,
          "validation_loss": 0.0008138756384141743,
          "train_metric": 0.00012722270912490785,
          "validation_metric": 8.197056013159454e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.792196167998554
        },
        {
          "epoch": 1176,
          "updates": 1176,
          "samples_seen": 1204224,
          "online_loss": 0.0012631785357370973,
          "train_loss": 0.00133003038354218,
          "validation_loss": 0.0012085504131391644,
          "train_metric": 0.0001339557784376666,
          "validation_metric": 0.00012172076822025701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.800524612999652
        },
        {
          "epoch": 1177,
          "updates": 1177,
          "samples_seen": 1205248,
          "online_loss": 0.0013300301507115364,
          "train_loss": 0.0014130265917629004,
          "validation_loss": 0.0008924348512664437,
          "train_metric": 0.00014231489330995828,
          "validation_metric": 8.988277113530785e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.809907707998718
        },
        {
          "epoch": 1178,
          "updates": 1178,
          "samples_seen": 1206272,
          "online_loss": 0.0014130267081782222,
          "train_loss": 0.0012482100864872336,
          "validation_loss": 0.0011237083235755563,
          "train_metric": 0.0001257151598110795,
          "validation_metric": 0.00011317578901071101,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.818721952999113
        },
        {
          "epoch": 1179,
          "updates": 1179,
          "samples_seen": 1207296,
          "online_loss": 0.0012482103193178773,
          "train_loss": 0.0010642759734764695,
          "validation_loss": 0.0007075656321831048,
          "train_metric": 0.00010718997509684414,
          "validation_metric": 7.126342825358734e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.827905075999297
        },
        {
          "epoch": 1180,
          "updates": 1180,
          "samples_seen": 1208320,
          "online_loss": 0.0010642759734764695,
          "train_loss": 0.0008986341999843717,
          "validation_loss": 0.0007466614479199052,
          "train_metric": 9.050713561009616e-05,
          "validation_metric": 7.52010164433159e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.83658752699921
        },
        {
          "epoch": 1181,
          "updates": 1181,
          "samples_seen": 1209344,
          "online_loss": 0.0008986340835690498,
          "train_loss": 0.0008864534320309758,
          "validation_loss": 0.0007340588490478694,
          "train_metric": 8.92803436727263e-05,
          "validation_metric": 7.393171108560637e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.8453115049997
        },
        {
          "epoch": 1182,
          "updates": 1182,
          "samples_seen": 1210368,
          "online_loss": 0.0008864534902386367,
          "train_loss": 0.0009936232818290591,
          "validation_loss": 0.0006963027990423143,
          "train_metric": 0.00010007410310208797,
          "validation_metric": 7.012907008174807e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.85418371799824
        },
        {
          "epoch": 1183,
          "updates": 1183,
          "samples_seen": 1211392,
          "online_loss": 0.000993623398244381,
          "train_loss": 0.0010857671732082963,
          "validation_loss": 0.000980461947619915,
          "train_metric": 0.00010935448517557234,
          "validation_metric": 9.874853276414797e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.863036679998913
        },
        {
          "epoch": 1184,
          "updates": 1184,
          "samples_seen": 1212416,
          "online_loss": 0.0010857670567929745,
          "train_loss": 0.001107290736399591,
          "validation_loss": 0.0007391975959762931,
          "train_metric": 0.00011152226943522692,
          "validation_metric": 7.444927177857608e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.871872987998358
        },
        {
          "epoch": 1185,
          "updates": 1185,
          "samples_seen": 1213440,
          "online_loss": 0.0011072908528149128,
          "train_loss": 0.000991345732472837,
          "validation_loss": 0.0008676581783220172,
          "train_metric": 9.9844706710428e-05,
          "validation_metric": 8.738735050428659e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.881379237998772
        },
        {
          "epoch": 1186,
          "updates": 1186,
          "samples_seen": 1214464,
          "online_loss": 0.0009913456160575151,
          "train_loss": 0.0008891180623322725,
          "validation_loss": 0.0006540277972817421,
          "train_metric": 8.954871736932546e-05,
          "validation_metric": 6.587128154933453e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.891032373998314
        },
        {
          "epoch": 1187,
          "updates": 1187,
          "samples_seen": 1215488,
          "online_loss": 0.0008891180623322725,
          "train_loss": 0.0008731157286092639,
          "validation_loss": 0.0006615745951421559,
          "train_metric": 8.793700544629246e-05,
          "validation_metric": 6.66313644615002e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.899964971998997
        },
        {
          "epoch": 1188,
          "updates": 1188,
          "samples_seen": 1216512,
          "online_loss": 0.0008731154957786202,
          "train_loss": 0.000934625742956996,
          "validation_loss": 0.0008045295253396034,
          "train_metric": 9.413209045305848e-05,
          "validation_metric": 8.102926949504763e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.909395565999148
        },
        {
          "epoch": 1189,
          "updates": 1189,
          "samples_seen": 1217536,
          "online_loss": 0.000934625742956996,
          "train_loss": 0.000997108407318592,
          "validation_loss": 0.0006955364951863885,
          "train_metric": 0.00010042510984931141,
          "validation_metric": 7.005188672337681e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.9179858589996
        },
        {
          "epoch": 1190,
          "updates": 1190,
          "samples_seen": 1218560,
          "online_loss": 0.000997108523733914,
          "train_loss": 0.0009757805964909494,
          "validation_loss": 0.0008643504115752876,
          "train_metric": 9.827704343479127e-05,
          "validation_metric": 8.705420623300597e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.927138160999675
        },
        {
          "epoch": 1191,
          "updates": 1191,
          "samples_seen": 1219584,
          "online_loss": 0.0009757805382832885,
          "train_loss": 0.0009145655203610659,
          "validation_loss": 0.0006696566706523299,
          "train_metric": 9.211169526679441e-05,
          "validation_metric": 6.744536949554458e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.93567186999826
        },
        {
          "epoch": 1192,
          "updates": 1192,
          "samples_seen": 1220608,
          "online_loss": 0.0009145655785687268,
          "train_loss": 0.0008647784707136452,
          "validation_loss": 0.0007041929638944566,
          "train_metric": 8.709731628187001e-05,
          "validation_metric": 7.092373562045395e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.944623371999114
        },
        {
          "epoch": 1193,
          "updates": 1193,
          "samples_seen": 1221632,
          "online_loss": 0.0008647784125059843,
          "train_loss": 0.0008722341153770685,
          "validation_loss": 0.0007188410963863134,
          "train_metric": 8.78482242114842e-05,
          "validation_metric": 7.239904516609386e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.953383272999417
        },
        {
          "epoch": 1194,
          "updates": 1194,
          "samples_seen": 1222656,
          "online_loss": 0.0008722341153770685,
          "train_loss": 0.000912727671675384,
          "validation_loss": 0.0006560054607689381,
          "train_metric": 9.192658035317436e-05,
          "validation_metric": 6.60704608890228e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.967880964999495
        },
        {
          "epoch": 1195,
          "updates": 1195,
          "samples_seen": 1223680,
          "online_loss": 0.0009127276134677231,
          "train_loss": 0.0009286038111895323,
          "validation_loss": 0.000801952788606286,
          "train_metric": 9.352558117825538e-05,
          "validation_metric": 8.076976519078016e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.98505444699913
        },
        {
          "epoch": 1196,
          "updates": 1196,
          "samples_seen": 1224704,
          "online_loss": 0.0009286038111895323,
          "train_loss": 0.0009086664649657905,
          "validation_loss": 0.0006587921525351703,
          "train_metric": 9.151755511993542e-05,
          "validation_metric": 6.635113095398992e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.006564171999344
        },
        {
          "epoch": 1197,
          "updates": 1197,
          "samples_seen": 1225728,
          "online_loss": 0.0009086663485504687,
          "train_loss": 0.0008703253697603941,
          "validation_loss": 0.0007231680792756379,
          "train_metric": 8.765598613535985e-05,
          "validation_metric": 7.283484592335299e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.024684871999852
        },
        {
          "epoch": 1198,
          "updates": 1198,
          "samples_seen": 1226752,
          "online_loss": 0.0008703253697603941,
          "train_loss": 0.0008556732209399343,
          "validation_loss": 0.0006859329296275973,
          "train_metric": 8.618026186013594e-05,
          "validation_metric": 6.908465729793534e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.202348484999675
        },
        {
          "epoch": 1199,
          "updates": 1199,
          "samples_seen": 1227776,
          "online_loss": 0.0008556731627322733,
          "train_loss": 0.0008705466752871871,
          "validation_loss": 0.000660610559862107,
          "train_metric": 8.767828694544733e-05,
          "validation_metric": 6.653428135905415e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.210041599999386
        },
        {
          "epoch": 1200,
          "updates": 1200,
          "samples_seen": 1228800,
          "online_loss": 0.000870546733494848,
          "train_loss": 0.0008899137610569596,
          "validation_loss": 0.0007553933537565172,
          "train_metric": 8.962886204244569e-05,
          "validation_metric": 7.608044688822702e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.217674923998857
        },
        {
          "epoch": 1201,
          "updates": 1201,
          "samples_seen": 1229824,
          "online_loss": 0.0008899138192646205,
          "train_loss": 0.0008921158732846379,
          "validation_loss": 0.0006525854114443064,
          "train_metric": 8.985065505839884e-05,
          "validation_metric": 6.572600977960974e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.226348545998917
        },
        {
          "epoch": 1202,
          "updates": 1202,
          "samples_seen": 1230848,
          "online_loss": 0.0008921159314922988,
          "train_loss": 0.0008721833000890911,
          "validation_loss": 0.000724825426004827,
          "train_metric": 8.78430946613662e-05,
          "validation_metric": 7.300176366697997e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.233860859998458
        },
        {
          "epoch": 1203,
          "updates": 1203,
          "samples_seen": 1231872,
          "online_loss": 0.0008721832418814301,
          "train_loss": 0.0008536794921383262,
          "validation_loss": 0.0006593631114810705,
          "train_metric": 8.597946725785732e-05,
          "validation_metric": 6.640864739893004e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.241285677999258
        },
        {
          "epoch": 1204,
          "updates": 1204,
          "samples_seen": 1232896,
          "online_loss": 0.0008536794921383262,
          "train_loss": 0.0008514797082170844,
          "validation_loss": 0.000662732170894742,
          "train_metric": 8.575792162446305e-05,
          "validation_metric": 6.674795440630987e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.248927286998878
        },
        {
          "epoch": 1205,
          "updates": 1205,
          "samples_seen": 1233920,
          "online_loss": 0.0008514796500094235,
          "train_loss": 0.0008623431203886867,
          "validation_loss": 0.000714058696758002,
          "train_metric": 8.685202919878066e-05,
          "validation_metric": 7.191737677203491e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.256647378999332
        },
        {
          "epoch": 1206,
          "updates": 1206,
          "samples_seen": 1234944,
          "online_loss": 0.0008623431203886867,
          "train_loss": 0.0008714146679267287,
          "validation_loss": 0.0006532832630909979,
          "train_metric": 8.776570030022413e-05,
          "validation_metric": 6.579629553016275e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.264615384998251
        },
        {
          "epoch": 1207,
          "updates": 1207,
          "samples_seen": 1235968,
          "online_loss": 0.0008714147843420506,
          "train_loss": 0.0008669500821270049,
          "validation_loss": 0.0007250954513438046,
          "train_metric": 8.731603156775236e-05,
          "validation_metric": 7.302896119654179e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.272487113999887
        },
        {
          "epoch": 1208,
          "updates": 1208,
          "samples_seen": 1236992,
          "online_loss": 0.0008669500821270049,
          "train_loss": 0.0008549949852749705,
          "validation_loss": 0.0006541803013533354,
          "train_metric": 8.611195517005399e-05,
          "validation_metric": 6.588664837181568e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.280010438998943
        },
        {
          "epoch": 1209,
          "updates": 1209,
          "samples_seen": 1238016,
          "online_loss": 0.0008549949270673096,
          "train_loss": 0.0008456670911982656,
          "validation_loss": 0.0006767214508727193,
          "train_metric": 8.517248352291062e-05,
          "validation_metric": 6.815690721850842e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.287696441999287
        },
        {
          "epoch": 1210,
          "updates": 1210,
          "samples_seen": 1239040,
          "online_loss": 0.0008456670329906046,
          "train_loss": 0.0008457585354335606,
          "validation_loss": 0.0006796361412853003,
          "train_metric": 8.518169488525018e-05,
          "validation_metric": 6.845046300441027e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.295269912998265
        },
        {
          "epoch": 1211,
          "updates": 1211,
          "samples_seen": 1240064,
          "online_loss": 0.0008457585354335606,
          "train_loss": 0.0008516908856108785,
          "validation_loss": 0.0006495282286778092,
          "train_metric": 8.577919652452692e-05,
          "validation_metric": 6.541810580529273e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.303591313999277
        },
        {
          "epoch": 1212,
          "updates": 1212,
          "samples_seen": 1241088,
          "online_loss": 0.0008516909438185394,
          "train_loss": 0.0008550037164241076,
          "validation_loss": 0.0007073128945194185,
          "train_metric": 8.611285011284053e-05,
          "validation_metric": 7.123795512598008e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.311824475998947
        },
        {
          "epoch": 1213,
          "updates": 1213,
          "samples_seen": 1242112,
          "online_loss": 0.0008550037164241076,
          "train_loss": 0.0008521984564140439,
          "validation_loss": 0.0006500492454506457,
          "train_metric": 8.583030285080895e-05,
          "validation_metric": 6.547058001160622e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.32996033499876
        },
        {
          "epoch": 1214,
          "updates": 1214,
          "samples_seen": 1243136,
          "online_loss": 0.0008521983399987221,
          "train_loss": 0.000845051952637732,
          "validation_loss": 0.0006908056093379855,
          "train_metric": 8.511054329574108e-05,
          "validation_metric": 6.957540608709678e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.347955003999232
        },
        {
          "epoch": 1215,
          "updates": 1215,
          "samples_seen": 1244160,
          "online_loss": 0.0008450520108453929,
          "train_loss": 0.000839682063087821,
          "validation_loss": 0.0006632032454945147,
          "train_metric": 8.45697068143636e-05,
          "validation_metric": 6.679539364995435e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.365073589999156
        },
        {
          "epoch": 1216,
          "updates": 1216,
          "samples_seen": 1245184,
          "online_loss": 0.000839682063087821,
          "train_loss": 0.0008390345610678196,
          "validation_loss": 0.000660087913274765,
          "train_metric": 8.450447785435244e-05,
          "validation_metric": 6.648163252975792e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.379743976998725
        },
        {
          "epoch": 1217,
          "updates": 1217,
          "samples_seen": 1246208,
          "online_loss": 0.0008390345610678196,
          "train_loss": 0.000841547385789454,
          "validation_loss": 0.0006846108590252697,
          "train_metric": 8.475755748804659e-05,
          "validation_metric": 6.895149999763817e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.387566316998345
        },
        {
          "epoch": 1218,
          "updates": 1218,
          "samples_seen": 1247232,
          "online_loss": 0.0008415475022047758,
          "train_loss": 0.0008434106130152941,
          "validation_loss": 0.0006468630162999034,
          "train_metric": 8.494522626278922e-05,
          "validation_metric": 6.514967390103266e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.39605800499885
        },
        {
          "epoch": 1219,
          "updates": 1219,
          "samples_seen": 1248256,
          "online_loss": 0.000843410671222955,
          "train_loss": 0.000841860834043473,
          "validation_loss": 0.0006877664127387106,
          "train_metric": 8.478913514409214e-05,
          "validation_metric": 6.926932110218331e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.403754471999491
        },
        {
          "epoch": 1220,
          "updates": 1220,
          "samples_seen": 1249280,
          "online_loss": 0.000841860834043473,
          "train_loss": 0.0008379773935303092,
          "validation_loss": 0.000649754423648119,
          "train_metric": 8.43980087665841e-05,
          "validation_metric": 6.544088682858273e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.411377295999046
        },
        {
          "epoch": 1221,
          "updates": 1221,
          "samples_seen": 1250304,
          "online_loss": 0.0008379773935303092,
          "train_loss": 0.0008342103683389723,
          "validation_loss": 0.0006699203513562679,
          "train_metric": 8.401861850870773e-05,
          "validation_metric": 6.747192674083635e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.418880957999136
        },
        {
          "epoch": 1222,
          "updates": 1222,
          "samples_seen": 1251328,
          "online_loss": 0.0008342103101313114,
          "train_loss": 0.0008326416718773544,
          "validation_loss": 0.000664910301566124,
          "train_metric": 8.386061381315812e-05,
          "validation_metric": 6.696733180433512e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.426507640999262
        },
        {
          "epoch": 1223,
          "updates": 1223,
          "samples_seen": 1252352,
          "online_loss": 0.0008326416718773544,
          "train_loss": 0.0008331193821504712,
          "validation_loss": 0.0006532707484439015,
          "train_metric": 8.390874427277595e-05,
          "validation_metric": 6.57950367894955e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.434238447998723
        },
        {
          "epoch": 1224,
          "updates": 1224,
          "samples_seen": 1253376,
          "online_loss": 0.0008331193821504712,
          "train_loss": 0.0008339408086612821,
          "validation_loss": 0.0006777647649869323,
          "train_metric": 8.399145735893399e-05,
          "validation_metric": 6.826198659837246e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.44206922399826
        },
        {
          "epoch": 1225,
          "updates": 1225,
          "samples_seen": 1254400,
          "online_loss": 0.0008339408086612821,
          "train_loss": 0.0008336058235727251,
          "validation_loss": 0.0006462559686042368,
          "train_metric": 8.395773329539225e-05,
          "validation_metric": 6.508854130515829e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.450651275999917
        },
        {
          "epoch": 1226,
          "updates": 1226,
          "samples_seen": 1255424,
          "online_loss": 0.0008336057653650641,
          "train_loss": 0.0008316428866237402,
          "validation_loss": 0.0006735580391250551,
          "train_metric": 8.376003097509965e-05,
          "validation_metric": 6.783830031054094e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.458812222999768
        },
        {
          "epoch": 1227,
          "updates": 1227,
          "samples_seen": 1256448,
          "online_loss": 0.0008316428866237402,
          "train_loss": 0.0008290598052553833,
          "validation_loss": 0.0006489384104497731,
          "train_metric": 8.349987183464691e-05,
          "validation_metric": 6.535869761137292e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.466461446998437
        },
        {
          "epoch": 1228,
          "updates": 1228,
          "samples_seen": 1257472,
          "online_loss": 0.0008290598634630442,
          "train_loss": 0.0008269999525509775,
          "validation_loss": 0.0006597029860131443,
          "train_metric": 8.32924124551937e-05,
          "validation_metric": 6.644288077950478e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.474614161999853
        },
        {
          "epoch": 1229,
          "updates": 1229,
          "samples_seen": 1258496,
          "online_loss": 0.0008269998943433166,
          "train_loss": 0.0008260742761194706,
          "validation_loss": 0.0006600675405934453,
          "train_metric": 8.319917833432555e-05,
          "validation_metric": 6.647958798566833e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.482576280999638
        },
        {
          "epoch": 1230,
          "updates": 1230,
          "samples_seen": 1259520,
          "online_loss": 0.0008260743925347924,
          "train_loss": 0.000825961004011333,
          "validation_loss": 0.0006496028508991003,
          "train_metric": 8.31877623568289e-05,
          "validation_metric": 6.542562186950818e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.490835849999712
        },
        {
          "epoch": 1231,
          "updates": 1231,
          "samples_seen": 1260544,
          "online_loss": 0.0008259609458036721,
          "train_loss": 0.0008258442394435406,
          "validation_loss": 0.0006684011314064264,
          "train_metric": 8.317599713336676e-05,
          "validation_metric": 6.731891335221007e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.498972334999053
        },
        {
          "epoch": 1232,
          "updates": 1232,
          "samples_seen": 1261568,
          "online_loss": 0.0008258440648205578,
          "train_loss": 0.0008251249673776329,
          "validation_loss": 0.000645417720079422,
          "train_metric": 8.310356497531757e-05,
          "validation_metric": 6.50041110930033e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.50780668599873
        },
        {
          "epoch": 1233,
          "updates": 1233,
          "samples_seen": 1262592,
          "online_loss": 0.0008251249673776329,
          "train_loss": 0.000823677284643054,
          "validation_loss": 0.0006650623399764299,
          "train_metric": 8.295775478472933e-05,
          "validation_metric": 6.698264041915536e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.516125352000017
        },
        {
          "epoch": 1234,
          "updates": 1234,
          "samples_seen": 1263616,
          "online_loss": 0.0008236772264353931,
          "train_loss": 0.000821944500785321,
          "validation_loss": 0.0006464611506089568,
          "train_metric": 8.278324094135314e-05,
          "validation_metric": 6.510919774882495e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.524386691999098
        },
        {
          "epoch": 1235,
          "updates": 1235,
          "samples_seen": 1264640,
          "online_loss": 0.0008219445589929819,
          "train_loss": 0.0008204036857932806,
          "validation_loss": 0.000654995848890394,
          "train_metric": 8.262805204140022e-05,
          "validation_metric": 6.596877938136458e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.532260389998555
        },
        {
          "epoch": 1236,
          "updates": 1236,
          "samples_seen": 1265664,
          "online_loss": 0.0008204036857932806,
          "train_loss": 0.0008193376124836504,
          "validation_loss": 0.0006525550852529705,
          "train_metric": 8.252068073488772e-05,
          "validation_metric": 6.57229611533694e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.640107063998585
        },
        {
          "epoch": 1237,
          "updates": 1237,
          "samples_seen": 1266688,
          "online_loss": 0.0008193376124836504,
          "train_loss": 0.0008186963968910277,
          "validation_loss": 0.0006470618536695838,
          "train_metric": 8.245609933510423e-05,
          "validation_metric": 6.516969733638689e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.648041673999614
        },
        {
          "epoch": 1238,
          "updates": 1238,
          "samples_seen": 1267712,
          "online_loss": 0.0008186963386833668,
          "train_loss": 0.0008181845187209547,
          "validation_loss": 0.0006583635695278645,
          "train_metric": 8.240455645136535e-05,
          "validation_metric": 6.630796997342259e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.655839660999845
        },
        {
          "epoch": 1239,
          "updates": 1239,
          "samples_seen": 1268736,
          "online_loss": 0.0008181845769286156,
          "train_loss": 0.000817508960608393,
          "validation_loss": 0.0006431639194488525,
          "train_metric": 8.23365044197999e-05,
          "validation_metric": 6.47771084913984e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.664657418999923
        },
        {
          "epoch": 1240,
          "updates": 1240,
          "samples_seen": 1269760,
          "online_loss": 0.000817508902400732,
          "train_loss": 0.0008165163453668356,
          "validation_loss": 0.0006578762549906969,
          "train_metric": 8.223654003813863e-05,
          "validation_metric": 6.62588863633573e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.672757875998286
        },
        {
          "epoch": 1241,
          "updates": 1241,
          "samples_seen": 1270784,
          "online_loss": 0.0008165164035744965,
          "train_loss": 0.000815290492027998,
          "validation_loss": 0.0006423193262889981,
          "train_metric": 8.211306703742594e-05,
          "validation_metric": 6.469205254688859e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.682002664998436
        },
        {
          "epoch": 1242,
          "updates": 1242,
          "samples_seen": 1271808,
          "online_loss": 0.0008152905502356589,
          "train_loss": 0.000813992228358984,
          "validation_loss": 0.0006518722511827946,
          "train_metric": 8.198231080314144e-05,
          "validation_metric": 6.565418152604252e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.69047774899991
        },
        {
          "epoch": 1243,
          "updates": 1243,
          "samples_seen": 1272832,
          "online_loss": 0.0008139921701513231,
          "train_loss": 0.0008128071203827858,
          "validation_loss": 0.0006445493781939149,
          "train_metric": 8.186295599443838e-05,
          "validation_metric": 6.491665408248082e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.698236164998889
        },
        {
          "epoch": 1244,
          "updates": 1244,
          "samples_seen": 1273856,
          "online_loss": 0.0008128071203827858,
          "train_loss": 0.0008118153782561421,
          "validation_loss": 0.0006454089307226241,
          "train_metric": 8.176307892426848e-05,
          "validation_metric": 6.500322342617437e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.705745574998218
        },
        {
          "epoch": 1245,
          "updates": 1245,
          "samples_seen": 1274880,
          "online_loss": 0.000811815436463803,
          "train_loss": 0.0008109888294711709,
          "validation_loss": 0.0006483267643488944,
          "train_metric": 8.167981286533177e-05,
          "validation_metric": 6.529709935421124e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.713374313998429
        },
        {
          "epoch": 1246,
          "updates": 1246,
          "samples_seen": 1275904,
          "online_loss": 0.00081098877126351,
          "train_loss": 0.0008102352730929852,
          "validation_loss": 0.0006410553469322622,
          "train_metric": 8.160392462741584e-05,
          "validation_metric": 6.456473784055561e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.721752738998475
        },
        {
          "epoch": 1247,
          "updates": 1247,
          "samples_seen": 1276928,
          "online_loss": 0.0008102353313006461,
          "train_loss": 0.0008094464428722858,
          "validation_loss": 0.0006500107701867819,
          "train_metric": 8.15244929981418e-05,
          "validation_metric": 6.546670920215547e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.72987238299902
        },
        {
          "epoch": 1248,
          "updates": 1248,
          "samples_seen": 1277952,
          "online_loss": 0.0008094466174952686,
          "train_loss": 0.0008085542940534651,
          "validation_loss": 0.0006385380402207375,
          "train_metric": 8.143463492160663e-05,
          "validation_metric": 6.431120709748939e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.73854554799982
        },
        {
          "epoch": 1249,
          "updates": 1249,
          "samples_seen": 1278976,
          "online_loss": 0.000808554352261126,
          "train_loss": 0.0008075535297393799,
          "validation_loss": 0.0006480894517153502,
          "train_metric": 8.133383380481973e-05,
          "validation_metric": 6.527319783344865e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.746227209998324
        },
        {
          "epoch": 1250,
          "updates": 1250,
          "samples_seen": 1280000,
          "online_loss": 0.0008075537625700235,
          "train_loss": 0.0008064906578511,
          "validation_loss": 0.0006378735997714102,
          "train_metric": 8.122678991639987e-05,
          "validation_metric": 6.424429011531174e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.754760601999806
        },
        {
          "epoch": 1251,
          "updates": 1251,
          "samples_seen": 1281024,
          "online_loss": 0.0008064905996434391,
          "train_loss": 0.0008054167265072465,
          "validation_loss": 0.0006442207377403975,
          "train_metric": 8.111862553050742e-05,
          "validation_metric": 6.48835557512939e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.762797030998627
        },
        {
          "epoch": 1252,
          "updates": 1252,
          "samples_seen": 1282048,
          "online_loss": 0.0008054166100919247,
          "train_loss": 0.0008043832494877279,
          "validation_loss": 0.0006391337374225259,
          "train_metric": 8.10145284049213e-05,
          "validation_metric": 6.437121191993356e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.770409603999724
        },
        {
          "epoch": 1253,
          "updates": 1253,
          "samples_seen": 1283072,
          "online_loss": 0.0008043833076953888,
          "train_loss": 0.0008034130441956222,
          "validation_loss": 0.0006403823499567807,
          "train_metric": 8.091681229416281e-05,
          "validation_metric": 6.449696229537949e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.778212765999342
        },
        {
          "epoch": 1254,
          "updates": 1254,
          "samples_seen": 1284096,
          "online_loss": 0.0008034131024032831,
          "train_loss": 0.0008025062270462513,
          "validation_loss": 0.0006408564513549209,
          "train_metric": 8.082547719823197e-05,
          "validation_metric": 6.454472168115899e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.785526407998987
        },
        {
          "epoch": 1255,
          "updates": 1255,
          "samples_seen": 1285120,
          "online_loss": 0.0008025061688385904,
          "train_loss": 0.0008016321226023138,
          "validation_loss": 0.000636960961855948,
          "train_metric": 8.073745266301557e-05,
          "validation_metric": 6.415237294277176e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.794062753999242
        },
        {
          "epoch": 1256,
          "updates": 1256,
          "samples_seen": 1286144,
          "online_loss": 0.0008016321808099747,
          "train_loss": 0.0008007672731764615,
          "validation_loss": 0.00064132921397686,
          "train_metric": 8.065035217441618e-05,
          "validation_metric": 6.459232827182859e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.801257026998428
        },
        {
          "epoch": 1257,
          "updates": 1257,
          "samples_seen": 1287168,
          "online_loss": 0.0008007673313841224,
          "train_loss": 0.0007998888613656163,
          "validation_loss": 0.0006341805565170944,
          "train_metric": 8.056187652982771e-05,
          "validation_metric": 6.387233588611707e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.81968739399963
        },
        {
          "epoch": 1258,
          "updates": 1258,
          "samples_seen": 1288192,
          "online_loss": 0.0007998888613656163,
          "train_loss": 0.0007989825680851936,
          "validation_loss": 0.0006403935258276761,
          "train_metric": 8.04706069175154e-05,
          "validation_metric": 6.449808279285207e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.827916942998854
        },
        {
          "epoch": 1259,
          "updates": 1259,
          "samples_seen": 1289216,
          "online_loss": 0.0007989824516698718,
          "train_loss": 0.000798048626165837,
          "validation_loss": 0.000632651848718524,
          "train_metric": 8.037654333747923e-05,
          "validation_metric": 6.371838389895856e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.83657250899887
        },
        {
          "epoch": 1260,
          "updates": 1260,
          "samples_seen": 1290240,
          "online_loss": 0.000798048626165837,
          "train_loss": 0.0007970880251377821,
          "validation_loss": 0.0006387719186022878,
          "train_metric": 8.027980220504105e-05,
          "validation_metric": 6.433476664824411e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.844614096999067
        },
        {
          "epoch": 1261,
          "updates": 1261,
          "samples_seen": 1291264,
          "online_loss": 0.000797088083345443,
          "train_loss": 0.0007961196824908257,
          "validation_loss": 0.0006322909030131996,
          "train_metric": 8.018226799322292e-05,
          "validation_metric": 6.368202593876049e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.85274674899847
        },
        {
          "epoch": 1262,
          "updates": 1262,
          "samples_seen": 1292288,
          "online_loss": 0.0007961197406984866,
          "train_loss": 0.00079514947719872,
          "validation_loss": 0.000636582903098315,
          "train_metric": 8.008454460650682e-05,
          "validation_metric": 6.411429785657674e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.860811263999494
        },
        {
          "epoch": 1263,
          "updates": 1263,
          "samples_seen": 1293312,
          "online_loss": 0.00079514947719872,
          "train_loss": 0.0007941891090013087,
          "validation_loss": 0.0006321465480141342,
          "train_metric": 7.998782530194148e-05,
          "validation_metric": 6.366748129948974e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.869778907999716
        },
        {
          "epoch": 1264,
          "updates": 1264,
          "samples_seen": 1294336,
          "online_loss": 0.0007941891090013087,
          "train_loss": 0.0007932402077130973,
          "validation_loss": 0.0006338332896120846,
          "train_metric": 7.989224832272157e-05,
          "validation_metric": 6.383736035786569e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.877778735999527
        },
        {
          "epoch": 1265,
          "updates": 1265,
          "samples_seen": 1295360,
          "online_loss": 0.0007932401495054364,
          "train_loss": 0.0007923044031485915,
          "validation_loss": 0.0006317725055851042,
          "train_metric": 7.979799556778744e-05,
          "validation_metric": 6.362982094287872e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.886890499999936
        },
        {
          "epoch": 1266,
          "updates": 1266,
          "samples_seen": 1296384,
          "online_loss": 0.0007923044031485915,
          "train_loss": 0.0007913835579529405,
          "validation_loss": 0.0006311112083494663,
          "train_metric": 7.970525621203706e-05,
          "validation_metric": 6.356320955092087e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.89608977899843
        },
        {
          "epoch": 1267,
          "updates": 1267,
          "samples_seen": 1297408,
          "online_loss": 0.0007913835579529405,
          "train_loss": 0.0007904701051302254,
          "validation_loss": 0.0006314226775430143,
          "train_metric": 7.961325172800571e-05,
          "validation_metric": 6.359458348015323e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.904210891998446
        },
        {
          "epoch": 1268,
          "updates": 1268,
          "samples_seen": 1298432,
          "online_loss": 0.0007904701633378863,
          "train_loss": 0.00078956427751109,
          "validation_loss": 0.0006289744051173329,
          "train_metric": 7.952202577143908e-05,
          "validation_metric": 6.334799400065094e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.913005341999451
        },
        {
          "epoch": 1269,
          "updates": 1269,
          "samples_seen": 1299456,
          "online_loss": 0.0007895642193034291,
          "train_loss": 0.0007886596140451729,
          "validation_loss": 0.0006310830358415842,
          "train_metric": 7.943090167827904e-05,
          "validation_metric": 6.356037192745134e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.921777445999396
        },
        {
          "epoch": 1270,
          "updates": 1270,
          "samples_seen": 1300480,
          "online_loss": 0.0007886596140451729,
          "train_loss": 0.0007877578027546406,
          "validation_loss": 0.000627186382189393,
          "train_metric": 7.93400831753388e-05,
          "validation_metric": 6.31679140496999e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.930835546998424
        },
        {
          "epoch": 1271,
          "updates": 1271,
          "samples_seen": 1301504,
          "online_loss": 0.0007877577445469797,
          "train_loss": 0.0007868576212786138,
          "validation_loss": 0.0006303447880782187,
          "train_metric": 7.924941746750847e-05,
          "validation_metric": 6.3486018916592e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.9391465839999
        },
        {
          "epoch": 1272,
          "updates": 1272,
          "samples_seen": 1302528,
          "online_loss": 0.0007868576794862747,
          "train_loss": 0.0007859505712985992,
          "validation_loss": 0.0006253631436266005,
          "train_metric": 7.915806781966239e-05,
          "validation_metric": 6.298428343143314e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.948199520998969
        },
        {
          "epoch": 1273,
          "updates": 1273,
          "samples_seen": 1303552,
          "online_loss": 0.0007859506295062602,
          "train_loss": 0.0007850447436794639,
          "validation_loss": 0.0006292102043516934,
          "train_metric": 7.906682731118053e-05,
          "validation_metric": 6.337175000226125e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.956456491998324
        },
        {
          "epoch": 1274,
          "updates": 1274,
          "samples_seen": 1304576,
          "online_loss": 0.000785044627264142,
          "train_loss": 0.0007841443875804543,
          "validation_loss": 0.0006236385670490563,
          "train_metric": 7.897614705143496e-05,
          "validation_metric": 6.281059177126735e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.965063754998482
        },
        {
          "epoch": 1275,
          "updates": 1275,
          "samples_seen": 1305600,
          "online_loss": 0.0007841443875804543,
          "train_loss": 0.0007832366973161697,
          "validation_loss": 0.0006281105452217162,
          "train_metric": 7.888473919592798e-05,
          "validation_metric": 6.326098809950054e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.97290899699874
        },
        {
          "epoch": 1276,
          "updates": 1276,
          "samples_seen": 1306624,
          "online_loss": 0.0007832367555238307,
          "train_loss": 0.00078233313979581,
          "validation_loss": 0.0006221733056008816,
          "train_metric": 7.879372424213216e-05,
          "validation_metric": 6.266302079893649e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.07960599299986
        },
        {
          "epoch": 1277,
          "updates": 1277,
          "samples_seen": 1307648,
          "online_loss": 0.0007823330233804882,
          "train_loss": 0.0007814319105818868,
          "validation_loss": 0.0006271881866268814,
          "train_metric": 7.870295667089522e-05,
          "validation_metric": 6.316809594864026e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.088182599998618
        },
        {
          "epoch": 1278,
          "updates": 1278,
          "samples_seen": 1308672,
          "online_loss": 0.0007814319687895477,
          "train_loss": 0.0007805359200574458,
          "validation_loss": 0.0006206753314472735,
          "train_metric": 7.861272752052173e-05,
          "validation_metric": 6.251214654184878e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.096782717999304
        },
        {
          "epoch": 1279,
          "updates": 1279,
          "samples_seen": 1309696,
          "online_loss": 0.0007805359782651067,
          "train_loss": 0.0007796455756761134,
          "validation_loss": 0.0006262584356591105,
          "train_metric": 7.852305134292692e-05,
          "validation_metric": 6.307445437414572e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.104586762998224
        },
        {
          "epoch": 1280,
          "updates": 1280,
          "samples_seen": 1310720,
          "online_loss": 0.0007796456338837743,
          "train_loss": 0.000778762623667717,
          "validation_loss": 0.0006188521510921419,
          "train_metric": 7.843411731300876e-05,
          "validation_metric": 6.232853047549725e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.112835585998255
        },
        {
          "epoch": 1281,
          "updates": 1281,
          "samples_seen": 1311744,
          "online_loss": 0.0007787625654600561,
          "train_loss": 0.0007778922445140779,
          "validation_loss": 0.0006254406762309372,
          "train_metric": 7.834646385163069e-05,
          "validation_metric": 6.299209780991077e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.121241998998812
        },
        {
          "epoch": 1282,
          "updates": 1282,
          "samples_seen": 1312768,
          "online_loss": 0.000777892186306417,
          "train_loss": 0.0007770408992655575,
          "validation_loss": 0.0006167718092910945,
          "train_metric": 7.826070941518992e-05,
          "validation_metric": 6.211899744812399e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.129652088999137
        },
        {
          "epoch": 1283,
          "updates": 1283,
          "samples_seen": 1313792,
          "online_loss": 0.0007770409574732184,
          "train_loss": 0.0007762102177366614,
          "validation_loss": 0.0006251207669265568,
          "train_metric": 7.817704317858443e-05,
          "validation_metric": 6.295987986959517e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.137808514999051
        },
        {
          "epoch": 1284,
          "updates": 1284,
          "samples_seen": 1314816,
          "online_loss": 0.0007762102177366614,
          "train_loss": 0.0007754188845865428,
          "validation_loss": 0.0006144638173282146,
          "train_metric": 7.809734233887866e-05,
          "validation_metric": 6.188655243022367e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.14625351199902
        },
        {
          "epoch": 1285,
          "updates": 1285,
          "samples_seen": 1315840,
          "online_loss": 0.0007754188845865428,
          "train_loss": 0.0007746752817183733,
          "validation_loss": 0.0006255630287341774,
          "train_metric": 7.802245090715587e-05,
          "validation_metric": 6.30044232821092e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.15404111799944
        },
        {
          "epoch": 1286,
          "updates": 1286,
          "samples_seen": 1316864,
          "online_loss": 0.0007746749324724078,
          "train_loss": 0.0007740203873254359,
          "validation_loss": 0.0006116176955401897,
          "train_metric": 7.79564943513833e-05,
          "validation_metric": 6.159989425214007e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.162508877998334
        },
        {
          "epoch": 1287,
          "updates": 1287,
          "samples_seen": 1317888,
          "online_loss": 0.0007740202127024531,
          "train_loss": 0.0007734789978712797,
          "validation_loss": 0.000627059256657958,
          "train_metric": 7.790197560098022e-05,
          "validation_metric": 6.315510836429894e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.170492965999074
        },
        {
          "epoch": 1288,
          "updates": 1288,
          "samples_seen": 1318912,
          "online_loss": 0.0007734790560789406,
          "train_loss": 0.0007731453515589237,
          "validation_loss": 0.0006079014856368303,
          "train_metric": 7.786836795276031e-05,
          "validation_metric": 6.122560444055125e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.179361472999517
        },
        {
          "epoch": 1289,
          "updates": 1289,
          "samples_seen": 1319936,
          "online_loss": 0.0007731454097665846,
          "train_loss": 0.0007730944780632854,
          "validation_loss": 0.0006305532879196107,
          "train_metric": 7.786325295455754e-05,
          "validation_metric": 6.350701733026654e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.197889822999059
        },
        {
          "epoch": 1290,
          "updates": 1290,
          "samples_seen": 1320960,
          "online_loss": 0.0007730945362709463,
          "train_loss": 0.000773566251154989,
          "validation_loss": 0.0006032537785358727,
          "train_metric": 7.791076495777816e-05,
          "validation_metric": 6.0757512983400375e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.216733602999739
        },
        {
          "epoch": 1291,
          "updates": 1291,
          "samples_seen": 1321984,
          "online_loss": 0.0007735663675703108,
          "train_loss": 0.0007747426279820502,
          "validation_loss": 0.0006380522390827537,
          "train_metric": 7.802924665156752e-05,
          "validation_metric": 6.426228355849162e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.231240754999817
        },
        {
          "epoch": 1292,
          "updates": 1292,
          "samples_seen": 1323008,
          "online_loss": 0.0007747426861897111,
          "train_loss": 0.0007773425313644111,
          "validation_loss": 0.0005979794077575207,
          "train_metric": 7.829108653822914e-05,
          "validation_metric": 6.022629895596765e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.253102154998487
        },
        {
          "epoch": 1293,
          "updates": 1293,
          "samples_seen": 1324032,
          "online_loss": 0.0007773424731567502,
          "train_loss": 0.0007817496079951525,
          "validation_loss": 0.0006536368746310472,
          "train_metric": 7.873496360844001e-05,
          "validation_metric": 6.58319186186418e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.268058193998513
        },
        {
          "epoch": 1294,
          "updates": 1294,
          "samples_seen": 1325056,
          "online_loss": 0.0007817496662028134,
          "train_loss": 0.0007901619537733495,
          "validation_loss": 0.000593935779761523,
          "train_metric": 7.95822124928236e-05,
          "validation_metric": 5.981904178042896e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.277166789999683
        },
        {
          "epoch": 1295,
          "updates": 1295,
          "samples_seen": 1326080,
          "online_loss": 0.0007901618955656886,
          "train_loss": 0.0008030908065848053,
          "validation_loss": 0.0006865570321679115,
          "train_metric": 8.088436152320355e-05,
          "validation_metric": 6.914749974384904e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.285243069998614
        },
        {
          "epoch": 1296,
          "updates": 1296,
          "samples_seen": 1327104,
          "online_loss": 0.0008030907483771443,
          "train_loss": 0.0008274510619230568,
          "validation_loss": 0.0005989279597997665,
          "train_metric": 8.333782898262143e-05,
          "validation_metric": 6.032182864146307e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.293044710999311
        },
        {
          "epoch": 1297,
          "updates": 1297,
          "samples_seen": 1328128,
          "online_loss": 0.0008274510037153959,
          "train_loss": 0.0008613371755927801,
          "validation_loss": 0.000757209025323391,
          "train_metric": 8.6750733316876e-05,
          "validation_metric": 7.626332808285952e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.300593924999703
        },
        {
          "epoch": 1298,
          "updates": 1298,
          "samples_seen": 1329152,
          "online_loss": 0.0008613371755927801,
          "train_loss": 0.0009256357443518937,
          "validation_loss": 0.000635804608464241,
          "train_metric": 9.322664845967665e-05,
          "validation_metric": 6.403591396519914e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.308275718000004
        },
        {
          "epoch": 1299,
          "updates": 1299,
          "samples_seen": 1330176,
          "online_loss": 0.0009256356861442327,
          "train_loss": 0.0009970655664801598,
          "validation_loss": 0.0008972232462838292,
          "train_metric": 0.00010042078793048859,
          "validation_metric": 9.036503615789115e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.31584158899932
        },
        {
          "epoch": 1300,
          "updates": 1300,
          "samples_seen": 1331200,
          "online_loss": 0.000997065450064838,
          "train_loss": 0.0011277925223112106,
          "validation_loss": 0.000737981463316828,
          "train_metric": 0.0001135871498263441,
          "validation_metric": 7.432678830809891e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.323396432999289
        },
        {
          "epoch": 1301,
          "updates": 1301,
          "samples_seen": 1332224,
          "online_loss": 0.0011277926387265325,
          "train_loss": 0.0011951372725889087,
          "validation_loss": 0.0010814964771270752,
          "train_metric": 0.00012036983389407396,
          "validation_metric": 0.0001089243742171675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.331604166998659
        },
        {
          "epoch": 1302,
          "updates": 1302,
          "samples_seen": 1333248,
          "online_loss": 0.0011951371561735868,
          "train_loss": 0.0012960012536495924,
          "validation_loss": 0.0008303148206323385,
          "train_metric": 0.00013052848225925118,
          "validation_metric": 8.362626249436289e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.339262091998535
        },
        {
          "epoch": 1303,
          "updates": 1303,
          "samples_seen": 1334272,
          "online_loss": 0.0012960010208189487,
          "train_loss": 0.0011761115165427327,
          "validation_loss": 0.001057697692885995,
          "train_metric": 0.00011845363769680262,
          "validation_metric": 0.00010652743367245421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.347390957998869
        },
        {
          "epoch": 1304,
          "updates": 1304,
          "samples_seen": 1335296,
          "online_loss": 0.0011761115165427327,
          "train_loss": 0.0010342831956222653,
          "validation_loss": 0.0006797658279538155,
          "train_metric": 0.00010416922305012122,
          "validation_metric": 6.846351607237011e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.355508313999962
        },
        {
          "epoch": 1305,
          "updates": 1305,
          "samples_seen": 1336320,
          "online_loss": 0.0010342831956222653,
          "train_loss": 0.000838356208987534,
          "validation_loss": 0.0007239844999276102,
          "train_metric": 8.443617116427049e-05,
          "validation_metric": 7.291707152035087e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.363185998998233
        },
        {
          "epoch": 1306,
          "updates": 1306,
          "samples_seen": 1337344,
          "online_loss": 0.000838356267195195,
          "train_loss": 0.0007569033186882734,
          "validation_loss": 0.0005988722550682724,
          "train_metric": 7.623252167832106e-05,
          "validation_metric": 6.0316218878142536e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.372386143999393
        },
        {
          "epoch": 1307,
          "updates": 1307,
          "samples_seen": 1338368,
          "online_loss": 0.0007569033186882734,
          "train_loss": 0.0008022654219530523,
          "validation_loss": 0.00059590331511572,
          "train_metric": 8.080124098341912e-05,
          "validation_metric": 6.0017202486051247e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.385877433998758
        },
        {
          "epoch": 1308,
          "updates": 1308,
          "samples_seen": 1339392,
          "online_loss": 0.0008022654219530523,
          "train_loss": 0.0009044929756782949,
          "validation_loss": 0.0008130563655868173,
          "train_metric": 9.109720849664882e-05,
          "validation_metric": 8.188805804820731e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.398065056999258
        },
        {
          "epoch": 1309,
          "updates": 1309,
          "samples_seen": 1340416,
          "online_loss": 0.0009044928592629731,
          "train_loss": 0.0009918666910380125,
          "validation_loss": 0.0006741239922121167,
          "train_metric": 9.989718091674149e-05,
          "validation_metric": 6.789528561057523e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.412485894999918
        },
        {
          "epoch": 1310,
          "updates": 1310,
          "samples_seen": 1341440,
          "online_loss": 0.0009918666910380125,
          "train_loss": 0.0009511650423519313,
          "validation_loss": 0.0008512783679179847,
          "train_metric": 9.579786274116486e-05,
          "validation_metric": 8.573764353059232e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.429266005998215
        },
        {
          "epoch": 1311,
          "updates": 1311,
          "samples_seen": 1342464,
          "online_loss": 0.0009511650423519313,
          "train_loss": 0.0008629229851067066,
          "validation_loss": 0.000598336395341903,
          "train_metric": 8.69104333105497e-05,
          "validation_metric": 6.026225310051814e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.442666559998543
        },
        {
          "epoch": 1312,
          "updates": 1312,
          "samples_seen": 1343488,
          "online_loss": 0.0008629228686913848,
          "train_loss": 0.0007705661118961871,
          "validation_loss": 0.0006360209081321955,
          "train_metric": 7.76085871621035e-05,
          "validation_metric": 6.40576909063384e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.459260706998975
        },
        {
          "epoch": 1313,
          "updates": 1313,
          "samples_seen": 1344512,
          "online_loss": 0.0007705660536885262,
          "train_loss": 0.0007565154228359461,
          "validation_loss": 0.0006174047011882067,
          "train_metric": 7.619346433784813e-05,
          "validation_metric": 6.218274211278185e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.47103647899894
        },
        {
          "epoch": 1314,
          "updates": 1314,
          "samples_seen": 1345536,
          "online_loss": 0.0007565154228359461,
          "train_loss": 0.0008101073908619583,
          "validation_loss": 0.0005947960889898241,
          "train_metric": 8.159104618243873e-05,
          "validation_metric": 5.990568024571985e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.478947223999057
        }
      ],
      "best": {
        "epoch": 1294,
        "updates": 1294
      },
      "best_validation_metric": 5.981904178042896e-05,
      "config": {
        "label": "Full batch",
        "batch": 1024,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.007734277069023902
    },
    "7a196ed2cdbd_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3071288261562586,
          "train_loss": 0.11203804612159729,
          "validation_loss": 0.07620196044445038,
          "train_metric": 0.011284060776233673,
          "validation_metric": 0.007674782536923885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04553349599882495
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.074585021706298,
          "train_loss": 0.05086174234747887,
          "validation_loss": 0.04995014891028404,
          "train_metric": 0.005122608505189419,
          "validation_metric": 0.005030796863138676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0985282069996174
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.04231863818131387,
          "train_loss": 0.032734062522649765,
          "validation_loss": 0.024340001866221428,
          "train_metric": 0.0032968551386147738,
          "validation_metric": 0.0024514361284673214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1313384839995706
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.028199329390190542,
          "train_loss": 0.02129916287958622,
          "validation_loss": 0.020750556141138077,
          "train_metric": 0.0021451739594340324,
          "validation_metric": 0.0020899202208966017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15431109499877493
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.02146986947627738,
          "train_loss": 0.02125341072678566,
          "validation_loss": 0.033918291330337524,
          "train_metric": 0.002140565775334835,
          "validation_metric": 0.0034161261282861233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18367860999933328
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.019407856394536793,
          "train_loss": 0.012668851763010025,
          "validation_loss": 0.009883912280201912,
          "train_metric": 0.001275960123166442,
          "validation_metric": 0.0009954714914783835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21548120899933565
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.024012377456529066,
          "train_loss": 0.02390378527343273,
          "validation_loss": 0.021115930750966072,
          "train_metric": 0.002407501917332411,
          "validation_metric": 0.0021267193369567394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24676636399999552
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.025285185838583857,
          "train_loss": 0.01872871071100235,
          "validation_loss": 0.012351857498288155,
          "train_metric": 0.0018862871220335364,
          "validation_metric": 0.001244033919647336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28670007700020506
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.012040380737744272,
          "train_loss": 0.011232025921344757,
          "validation_loss": 0.006161433178931475,
          "train_metric": 0.0011312485439702868,
          "validation_metric": 0.0006205569952726364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31499210400033917
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.009275261327275075,
          "train_loss": 0.009033228270709515,
          "validation_loss": 0.006427751388400793,
          "train_metric": 0.0009097935981117189,
          "validation_metric": 0.0006473796092905104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32870426299996325
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.011373822577297688,
          "train_loss": 0.0066519249230623245,
          "validation_loss": 0.005146905779838562,
          "train_metric": 0.0006699575460515916,
          "validation_metric": 0.0005183775210753083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34271165399877646
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.008848477293213364,
          "train_loss": 0.016483120620250702,
          "validation_loss": 0.012250147759914398,
          "train_metric": 0.0016601196257397532,
          "validation_metric": 0.0012337899534031749,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.356108881000182
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.01770462401327677,
          "train_loss": 0.0076902518048882484,
          "validation_loss": 0.0066915531642735004,
          "train_metric": 0.0007745339535176754,
          "validation_metric": 0.0006739486125297844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36958356799914327
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.01383943113614805,
          "train_loss": 0.007693617604672909,
          "validation_loss": 0.008185774087905884,
          "train_metric": 0.0007748730131424963,
          "validation_metric": 0.0008244412601925433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38358281000000716
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.013440196213196032,
          "train_loss": 0.020549535751342773,
          "validation_loss": 0.015472847037017345,
          "train_metric": 0.002069673966616392,
          "validation_metric": 0.00155836867634207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3969296849991224
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.015208673314191401,
          "train_loss": 0.02309454418718815,
          "validation_loss": 0.015796467661857605,
          "train_metric": 0.002325997920706868,
          "validation_metric": 0.0015909625217318535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41047016399897984
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.01357882475713268,
          "train_loss": 0.024356551468372345,
          "validation_loss": 0.019021105021238327,
          "train_metric": 0.002453102497383952,
          "validation_metric": 0.0019157358910888433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4238435629995365
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.01771878165891394,
          "train_loss": 0.016155371442437172,
          "validation_loss": 0.01659294404089451,
          "train_metric": 0.0016271100612357259,
          "validation_metric": 0.0016711805947124958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.436712175998764
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.01291148022573907,
          "train_loss": 0.009475353173911572,
          "validation_loss": 0.00587684940546751,
          "train_metric": 0.0009543228079564869,
          "validation_metric": 0.0005918947863392532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4500318899990816
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.009440544323297217,
          "train_loss": 0.012063981965184212,
          "validation_loss": 0.010029615834355354,
          "train_metric": 0.001215039985254407,
          "validation_metric": 0.001010146108455956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4633910639986425
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.006840501737315208,
          "train_loss": 0.006719559896737337,
          "validation_loss": 0.0064447373151779175,
          "train_metric": 0.0006767694721929729,
          "validation_metric": 0.0006490902742370963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4767102320001868
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.006070428840757813,
          "train_loss": 0.004238975699990988,
          "validation_loss": 0.0047693317756056786,
          "train_metric": 0.00042693412979133427,
          "validation_metric": 0.0004803496412932873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4910117839990562
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.009627404615457635,
          "train_loss": 0.009205630980432034,
          "validation_loss": 0.009717017412185669,
          "train_metric": 0.0009271574672311544,
          "validation_metric": 0.0009786623995751143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5043552839997574
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.014047909789951518,
          "train_loss": 0.013953428715467453,
          "validation_loss": 0.015782037749886513,
          "train_metric": 0.001405338174663484,
          "validation_metric": 0.0015895090764388442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5179856159993506
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.010443313098221552,
          "train_loss": 0.006825929507613182,
          "validation_loss": 0.006799179129302502,
          "train_metric": 0.0006874826503917575,
          "validation_metric": 0.0006847883923910558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5314287490000424
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.011437548062531278,
          "train_loss": 0.011943532153964043,
          "validation_loss": 0.008627168834209442,
          "train_metric": 0.0012029088102281094,
          "validation_metric": 0.0008688968955539167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6824136809991614
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.006769984611310065,
          "train_loss": 0.004456540569663048,
          "validation_loss": 0.004082114901393652,
          "train_metric": 0.00044884643284603953,
          "validation_metric": 0.0004111356975045055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7142829080003139
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01576909028517548,
          "train_loss": 0.01666870154440403,
          "validation_loss": 0.014270413666963577,
          "train_metric": 0.0016788106877356768,
          "validation_metric": 0.001437263679690659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.74516892399879
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.014986050591687672,
          "train_loss": 0.01137054618448019,
          "validation_loss": 0.009255947545170784,
          "train_metric": 0.0011451997561380267,
          "validation_metric": 0.0009322250261902809,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7762038939999911
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.013578396625234745,
          "train_loss": 0.006296770181506872,
          "validation_loss": 0.0036617377772927284,
          "train_metric": 0.0006341876578517258,
          "validation_metric": 0.00036879684193991125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8149122609993356
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.009990488950279541,
          "train_loss": 0.019505172967910767,
          "validation_loss": 0.015952683985233307,
          "train_metric": 0.001964489696547389,
          "validation_metric": 0.001606695936061442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8445498250002856
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01945007365429774,
          "train_loss": 0.009888243861496449,
          "validation_loss": 0.006357372738420963,
          "train_metric": 0.0009959075832739472,
          "validation_metric": 0.000640291313175112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8801719290004257
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.010607568343402818,
          "train_loss": 0.013945555314421654,
          "validation_loss": 0.012985952198505402,
          "train_metric": 0.0014045452699065208,
          "validation_metric": 0.0013078978518024087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9108981959998346
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.011888013665156905,
          "train_loss": 0.007090152706950903,
          "validation_loss": 0.006639047991484404,
          "train_metric": 0.0007140942034311593,
          "validation_metric": 0.0006686606211587787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9415893089990277
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.0071046433949959464,
          "train_loss": 0.004070295486599207,
          "validation_loss": 0.0030450946651399136,
          "train_metric": 0.0004099452344235033,
          "validation_metric": 0.0003066907811444253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9751920199996675
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.008023191941902041,
          "train_loss": 0.004413176327943802,
          "validation_loss": 0.0034315665252506733,
          "train_metric": 0.00044447893742471933,
          "validation_metric": 0.00034561482607387006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0065741010002967
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.006907705130288377,
          "train_loss": 0.005530179478228092,
          "validation_loss": 0.003692805767059326,
          "train_metric": 0.000556979444809258,
          "validation_metric": 0.0003719258529599756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0375607749992923
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.0040482903641532175,
          "train_loss": 0.002839048858731985,
          "validation_loss": 0.0027677796315401793,
          "train_metric": 0.0002859385858755559,
          "validation_metric": 0.00027876062085852027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0722364929988544
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01328999012184795,
          "train_loss": 0.035411953926086426,
          "validation_loss": 0.02749413251876831,
          "train_metric": 0.0035665626637637615,
          "validation_metric": 0.0027691081631928682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1027916290004214
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.022517298988532275,
          "train_loss": 0.011345796287059784,
          "validation_loss": 0.013017391785979271,
          "train_metric": 0.0011427070712670684,
          "validation_metric": 0.0013110639993101358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1163253969989455
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.008080109604634345,
          "train_loss": 0.00588337192311883,
          "validation_loss": 0.004020294640213251,
          "train_metric": 0.0005925517180003226,
          "validation_metric": 0.00040490939863957465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1301825560003635
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.005963767820503563,
          "train_loss": 0.004166283644735813,
          "validation_loss": 0.003553282469511032,
          "train_metric": 0.0004196127993054688,
          "validation_metric": 0.0003578736213967204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.148314729998674
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.009335822789580561,
          "train_loss": 0.01667610928416252,
          "validation_loss": 0.014476744458079338,
          "train_metric": 0.0016795566771179438,
          "validation_metric": 0.0014580447459593415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1722660630002792
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.01296637166524306,
          "train_loss": 0.008167150430381298,
          "validation_loss": 0.007495296187698841,
          "train_metric": 0.0008225654601119459,
          "validation_metric": 0.0007548988214693964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1872474550000334
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.006371240058797412,
          "train_loss": 0.007321102544665337,
          "validation_loss": 0.007356789894402027,
          "train_metric": 0.0007373546250164509,
          "validation_metric": 0.0007409490062855184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2016777949993411
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.005271240741421934,
          "train_loss": 0.004778060130774975,
          "validation_loss": 0.005544597283005714,
          "train_metric": 0.00048122869338840246,
          "validation_metric": 0.0005584315513260663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.215044439999474
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.00498176013934426,
          "train_loss": 0.002443666337057948,
          "validation_loss": 0.0022033751010894775,
          "train_metric": 0.00024611709523014724,
          "validation_metric": 0.00022191590687725693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2294914319991221
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0035049753241764847,
          "train_loss": 0.004942765925079584,
          "validation_loss": 0.004673020448535681,
          "train_metric": 0.0004978172364644706,
          "validation_metric": 0.0004706494801212102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2427500569992844
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.004495720146223903,
          "train_loss": 0.004035751335322857,
          "validation_loss": 0.0031157019548118114,
          "train_metric": 0.0004064661043230444,
          "validation_metric": 0.00031380209838971496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.255898473000343
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.004526940640062094,
          "train_loss": 0.003896507667377591,
          "validation_loss": 0.003825979772955179,
          "train_metric": 0.00039244198706001043,
          "validation_metric": 0.0003853386442642659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.269216372998926
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.005048604900366627,
          "train_loss": 0.008761128410696983,
          "validation_loss": 0.00579631794244051,
          "train_metric": 0.0008823888492770493,
          "validation_metric": 0.0005837838398292661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2828137389988115
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008277216817077715,
          "train_loss": 0.005358617752790451,
          "validation_loss": 0.0038503266405314207,
          "train_metric": 0.0005397003842517734,
          "validation_metric": 0.00038779081660322845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.29641606199948
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.008937666658312082,
          "train_loss": 0.004297289066016674,
          "validation_loss": 0.003845198079943657,
          "train_metric": 0.0004328072536736727,
          "validation_metric": 0.00038727428182028234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3105121070002497
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.004864150418143254,
          "train_loss": 0.003194466233253479,
          "validation_loss": 0.003566374769434333,
          "train_metric": 0.00032173492945730686,
          "validation_metric": 0.0003591921995393932,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.324044385999514
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.004782680254720617,
          "train_loss": 0.004385479260236025,
          "validation_loss": 0.004400455392897129,
          "train_metric": 0.00044168936437927186,
          "validation_metric": 0.000443197728600353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3377201759994932
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.003914387518307194,
          "train_loss": 0.0026528919115662575,
          "validation_loss": 0.002810490084812045,
          "train_metric": 0.00026718954904936254,
          "validation_metric": 0.0002830622543115169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3513126189991453
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.004540577207080787,
          "train_loss": 0.002056040335446596,
          "validation_loss": 0.0017707201186567545,
          "train_metric": 0.00020707685325760394,
          "validation_metric": 0.00017834047321230173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3658842829991045
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.018147993556340225,
          "train_loss": 0.02214161306619644,
          "validation_loss": 0.015251225791871548,
          "train_metric": 0.0022300221025943756,
          "validation_metric": 0.001536047668196261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.379590989999997
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.01538866848568432,
          "train_loss": 0.012630714103579521,
          "validation_loss": 0.01062769629061222,
          "train_metric": 0.001272119116038084,
          "validation_metric": 0.0010703826555982232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.393315478999284
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0071873173146741465,
          "train_loss": 0.006293626967817545,
          "validation_loss": 0.0052931103855371475,
          "train_metric": 0.0006338709499686956,
          "validation_metric": 0.0005331026623025537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4065996909994283
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004570935227093287,
          "train_loss": 0.00463627465069294,
          "validation_loss": 0.0032044495455920696,
          "train_metric": 0.00046694857883267105,
          "validation_metric": 0.0003227404085919261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.419726781999998
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0039688462420599535,
          "train_loss": 0.006133649498224258,
          "validation_loss": 0.0034299264661967754,
          "train_metric": 0.0006177587201818824,
          "validation_metric": 0.0003454496036283672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.510533217999182
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.010329144039133098,
          "train_loss": 0.009874947369098663,
          "validation_loss": 0.007222643122076988,
          "train_metric": 0.0009945685742422938,
          "validation_metric": 0.0007274382514879107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.524697152999579
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.007473938385373913,
          "train_loss": 0.003380092326551676,
          "validation_loss": 0.002962950151413679,
          "train_metric": 0.00034043053165078163,
          "validation_metric": 0.00029841752257198095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5380219130001933
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.007203828259662259,
          "train_loss": 0.00945940613746643,
          "validation_loss": 0.006850644946098328,
          "train_metric": 0.0009527167421765625,
          "validation_metric": 0.000689971842803061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5515364410002803
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.006102527371695032,
          "train_loss": 0.004661147017031908,
          "validation_loss": 0.0043709771707654,
          "train_metric": 0.00046945360372774303,
          "validation_metric": 0.0004402288468554616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5652187569994567
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0037031408755865414,
          "train_loss": 0.0022908838000148535,
          "validation_loss": 0.0015214508166536689,
          "train_metric": 0.00023072943440638483,
          "validation_metric": 0.0001532349269837141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.57935680799892
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0028230119569343515,
          "train_loss": 0.0017397275660187006,
          "validation_loss": 0.002219205955043435,
          "train_metric": 0.00017521898553241044,
          "validation_metric": 0.0002235103165730834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5930320799998299
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.008306211986564449,
          "train_loss": 0.0058769723400473595,
          "validation_loss": 0.004003970883786678,
          "train_metric": 0.0005919070681557059,
          "validation_metric": 0.00040326532325707376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6077890549986478
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.010768441410618834,
          "train_loss": 0.005165383219718933,
          "validation_loss": 0.004679038655012846,
          "train_metric": 0.0005202384199947119,
          "validation_metric": 0.0004712556255981326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6211363889997301
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0041675820320961066,
          "train_loss": 0.0027609083335846663,
          "validation_loss": 0.00321635277941823,
          "train_metric": 0.0002780685608740896,
          "validation_metric": 0.00032393925357609987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6346216139991157
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.005793672764411895,
          "train_loss": 0.009301627986133099,
          "validation_loss": 0.005919034592807293,
          "train_metric": 0.0009368258761242032,
          "validation_metric": 0.0005961434217169881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6476951539989386
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.01884086456266232,
          "train_loss": 0.008877714164555073,
          "validation_loss": 0.010837715119123459,
          "train_metric": 0.0008941309060901403,
          "validation_metric": 0.0010915349703282118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.661279072999605
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.008154400333296508,
          "train_loss": 0.0052547999657690525,
          "validation_loss": 0.006944365333765745,
          "train_metric": 0.0005292442510835826,
          "validation_metric": 0.0006994110881350935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6746135989997128
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.006463957513915375,
          "train_loss": 0.014513191767036915,
          "validation_loss": 0.014055795967578888,
          "train_metric": 0.0014617154374718666,
          "validation_metric": 0.0014156483812257648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6880282889997034
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.01248252679943107,
          "train_loss": 0.022043246775865555,
          "validation_loss": 0.021901432424783707,
          "train_metric": 0.0022201151587069035,
          "validation_metric": 0.002205832162871957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.701326248999976
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.014458285921136849,
          "train_loss": 0.007923133671283722,
          "validation_loss": 0.008137684315443039,
          "train_metric": 0.0007979889633134007,
          "validation_metric": 0.0008195977425202727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.714701384998989
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.006276465603150427,
          "train_loss": 0.004931956063956022,
          "validation_loss": 0.0049920761957764626,
          "train_metric": 0.000496728578582406,
          "validation_metric": 0.0005027836887165904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7281405510002514
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.005267220200039446,
          "train_loss": 0.0036401061806827784,
          "validation_loss": 0.0032883307430893183,
          "train_metric": 0.0003666181582957506,
          "validation_metric": 0.00033118861028924584,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.74136881800041
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0030278763588285074,
          "train_loss": 0.0021883423905819654,
          "validation_loss": 0.0020286240614950657,
          "train_metric": 0.0002204018528573215,
          "validation_metric": 0.00020431558368727565,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7546607469994342
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.002528374385292409,
          "train_loss": 0.002161098178476095,
          "validation_loss": 0.0019721882417798042,
          "train_metric": 0.0002176579146180302,
          "validation_metric": 0.00019863159104716033,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.768130876998839
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0025953458480216796,
          "train_loss": 0.0016554875764995813,
          "validation_loss": 0.0018826116574928164,
          "train_metric": 0.00016673465142957866,
          "validation_metric": 0.00018960975285153836,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7815110809988255
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0024079941722447984,
          "train_loss": 0.0013693112414330244,
          "validation_loss": 0.0017088877502828836,
          "train_metric": 0.00013791202218271792,
          "validation_metric": 0.00017211293743457645,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7950711960002081
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.002988389904203359,
          "train_loss": 0.004422005731612444,
          "validation_loss": 0.005552405025810003,
          "train_metric": 0.0004453682340681553,
          "validation_metric": 0.0005592179368250072,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8084561569994548
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.005503419684828259,
          "train_loss": 0.020778298377990723,
          "validation_loss": 0.012193625792860985,
          "train_metric": 0.002092714188620448,
          "validation_metric": 0.0012280972441658378,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8219124569986889
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.009110441140364856,
          "train_loss": 0.015589948743581772,
          "validation_loss": 0.01742377132177353,
          "train_metric": 0.0015701628290116787,
          "validation_metric": 0.0017548584146425128,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.836789355998917
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.011906222731340677,
          "train_loss": 0.009792032651603222,
          "validation_loss": 0.008311251178383827,
          "train_metric": 0.0009862176375463605,
          "validation_metric": 0.0008370787254534662,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8504025370002637
        }
      ],
      "best": {
        "epoch": 67,
        "updates": 1072
      },
      "best_validation_metric": 0.0001532349269837141,
      "config": {
        "label": "LayerNorm + relu + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012378809594775829
    },
    "d5c218b8a06d_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 1.0353699550032616,
          "train_loss": 0.6616478562355042,
          "validation_loss": 0.15127693116664886,
          "train_metric": 0.0666387602686882,
          "validation_metric": 0.01523605827242136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.012213568999868585
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.6407988909631968,
          "train_loss": 0.5816639065742493,
          "validation_loss": 0.1613098382949829,
          "train_metric": 0.058583058416843414,
          "validation_metric": 0.016246533021330833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.023571843999889097
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.5717756627127528,
          "train_loss": 0.5513314008712769,
          "validation_loss": 0.1295091062784195,
          "train_metric": 0.055528078228235245,
          "validation_metric": 0.01304368395358324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03562683700147318
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.5522993565537035,
          "train_loss": 0.5369275212287903,
          "validation_loss": 0.10265173763036728,
          "train_metric": 0.05407737195491791,
          "validation_metric": 0.010338706895709038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04804629000136629
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.535174772143364,
          "train_loss": 0.5222739577293396,
          "validation_loss": 0.10566778481006622,
          "train_metric": 0.052601516246795654,
          "validation_metric": 0.01064247079193592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.059683904000849
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.519182694144547,
          "train_loss": 0.5074906945228577,
          "validation_loss": 0.09390892833471298,
          "train_metric": 0.05111260339617729,
          "validation_metric": 0.009458162821829319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0717628220008919
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.5094327344559133,
          "train_loss": 0.49732738733291626,
          "validation_loss": 0.08204476535320282,
          "train_metric": 0.05008899047970772,
          "validation_metric": 0.008263248950242996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08406997800011595
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.49998185131698847,
          "train_loss": 0.49542731046676636,
          "validation_loss": 0.06494864076375961,
          "train_metric": 0.04989761859178543,
          "validation_metric": 0.006541389040648937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09895877900089545
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.4992927308194339,
          "train_loss": 0.49247777462005615,
          "validation_loss": 0.0863134041428566,
          "train_metric": 0.04960056021809578,
          "validation_metric": 0.008693169802427292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11188411300099688
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.49509616754949093,
          "train_loss": 0.487638920545578,
          "validation_loss": 0.07519876956939697,
          "train_metric": 0.049113210290670395,
          "validation_metric": 0.007573744282126427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1253504050000629
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.4903661129064858,
          "train_loss": 0.48657703399658203,
          "validation_loss": 0.064376100897789,
          "train_metric": 0.04900626093149185,
          "validation_metric": 0.0064837257377803326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1393449650004186
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.49242843873798847,
          "train_loss": 0.4839695394039154,
          "validation_loss": 0.07430604100227356,
          "train_metric": 0.04874364286661148,
          "validation_metric": 0.007483832538127899,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15263032199982263
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.4873566683381796,
          "train_loss": 0.48235568404197693,
          "validation_loss": 0.0620291642844677,
          "train_metric": 0.04858110100030899,
          "validation_metric": 0.006247350480407476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16675674399994023
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.4886193424463272,
          "train_loss": 0.48054391145706177,
          "validation_loss": 0.07828045636415482,
          "train_metric": 0.048398613929748535,
          "validation_metric": 0.00788412056863308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17977085799975612
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.48502733930945396,
          "train_loss": 0.4803396463394165,
          "validation_loss": 0.06485161185264587,
          "train_metric": 0.048378050327301025,
          "validation_metric": 0.00653161620721221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1940407440015406
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.4787627160549164,
          "train_loss": 0.47911354899406433,
          "validation_loss": 0.06501246243715286,
          "train_metric": 0.04825455695390701,
          "validation_metric": 0.0065478174947202206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20821288200022536
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.4790285797789693,
          "train_loss": 0.4731758236885071,
          "validation_loss": 0.07103712111711502,
          "train_metric": 0.04765653610229492,
          "validation_metric": 0.0071545979008078575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22136226000111492
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.4792795339599252,
          "train_loss": 0.47354623675346375,
          "validation_loss": 0.04455650970339775,
          "train_metric": 0.04769383743405342,
          "validation_metric": 0.004487568512558937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23549961500066274
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.47602085722610354,
          "train_loss": 0.4771367013454437,
          "validation_loss": 0.06354783475399017,
          "train_metric": 0.048055458813905716,
          "validation_metric": 0.0064003043808043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24897189300099853
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.48254165798425674,
          "train_loss": 0.4772379994392395,
          "validation_loss": 0.0780072808265686,
          "train_metric": 0.0480656623840332,
          "validation_metric": 0.007856608368456364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2629516040015005
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.4813127941451967,
          "train_loss": 0.47334831953048706,
          "validation_loss": 0.07286420464515686,
          "train_metric": 0.04767390340566635,
          "validation_metric": 0.007338615134358406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27557258600063506
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.47227866016328335,
          "train_loss": 0.4636358916759491,
          "validation_loss": 0.05443054065108299,
          "train_metric": 0.04669570550322533,
          "validation_metric": 0.005482045002281666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2887249070008693
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.47379430290311575,
          "train_loss": 0.4612402319908142,
          "validation_loss": 0.05303630605340004,
          "train_metric": 0.046454425901174545,
          "validation_metric": 0.005341622047126293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3011608350007009
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.4818386137485504,
          "train_loss": 0.46348708868026733,
          "validation_loss": 0.04295629635453224,
          "train_metric": 0.04668072611093521,
          "validation_metric": 0.004326401278376579,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.315792549001344
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.47507835319265723,
          "train_loss": 0.48637092113494873,
          "validation_loss": 0.04923360422253609,
          "train_metric": 0.048985496163368225,
          "validation_metric": 0.004958628211170435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3284527430005255
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.4704994782805443,
          "train_loss": 0.46492984890937805,
          "validation_loss": 0.05777253583073616,
          "train_metric": 0.04682603478431702,
          "validation_metric": 0.005818638019263744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34112865600036457
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.4649177510291338,
          "train_loss": 0.46791544556617737,
          "validation_loss": 0.07896246016025543,
          "train_metric": 0.04712672904133797,
          "validation_metric": 0.007952810265123844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35361474199999066
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.47049538884311914,
          "train_loss": 0.4559851586818695,
          "validation_loss": 0.051537223160266876,
          "train_metric": 0.04592515155673027,
          "validation_metric": 0.00519063975661993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3671150840000337
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.46564103197306395,
          "train_loss": 0.45946016907691956,
          "validation_loss": 0.057996660470962524,
          "train_metric": 0.046275146305561066,
          "validation_metric": 0.005841211415827274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3812403820011241
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.4598056273534894,
          "train_loss": 0.4615652859210968,
          "validation_loss": 0.03355047479271889,
          "train_metric": 0.04648716375231743,
          "validation_metric": 0.0033790809102356434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39449836000130745
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.4660371863283217,
          "train_loss": 0.46245336532592773,
          "validation_loss": 0.030446941033005714,
          "train_metric": 0.0465766079723835,
          "validation_metric": 0.0030665043741464615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4095392240014917
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.46344073908403516,
          "train_loss": 0.45461222529411316,
          "validation_loss": 0.04936487600207329,
          "train_metric": 0.045786868780851364,
          "validation_metric": 0.004971849732100964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42242156000065734
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.46396748069673777,
          "train_loss": 0.4540259540081024,
          "validation_loss": 0.05142803490161896,
          "train_metric": 0.04572782665491104,
          "validation_metric": 0.005179643630981445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43844701800117036
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.45731757394969463,
          "train_loss": 0.45554202795028687,
          "validation_loss": 0.027565453201532364,
          "train_metric": 0.045880526304244995,
          "validation_metric": 0.0027762914542108774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4516201690003072
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.4570326553657651,
          "train_loss": 0.45226216316223145,
          "validation_loss": 0.05119622126221657,
          "train_metric": 0.04555018991231918,
          "validation_metric": 0.005156295374035835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46397177100152476
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.4670932376757264,
          "train_loss": 0.46009474992752075,
          "validation_loss": 0.029939865693449974,
          "train_metric": 0.04633905738592148,
          "validation_metric": 0.0030154334381222725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.476494234000711
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.4550467061344534,
          "train_loss": 0.44997450709342957,
          "validation_loss": 0.042140841484069824,
          "train_metric": 0.045319780707359314,
          "validation_metric": 0.004244270734488964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4882127729997592
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.45891597494482994,
          "train_loss": 0.46029165387153625,
          "validation_loss": 0.05527597665786743,
          "train_metric": 0.04635889083147049,
          "validation_metric": 0.0055671934969723225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5001565260008647
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.462711033411324,
          "train_loss": 0.45298707485198975,
          "validation_loss": 0.05509708821773529,
          "train_metric": 0.04562320187687874,
          "validation_metric": 0.005549176596105099,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5140107580009499
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.461613105610013,
          "train_loss": 0.4605438709259033,
          "validation_loss": 0.055018145591020584,
          "train_metric": 0.046384297311306,
          "validation_metric": 0.005541226360946894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5261519720006618
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.45986065175384283,
          "train_loss": 0.44919830560684204,
          "validation_loss": 0.045472390949726105,
          "train_metric": 0.04524161294102669,
          "validation_metric": 0.004579812753945589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5383381820010982
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.4546426245942712,
          "train_loss": 0.45495471358299255,
          "validation_loss": 0.04777459427714348,
          "train_metric": 0.04582137241959572,
          "validation_metric": 0.0048116822727024555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5506942730007722
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.45067225117236376,
          "train_loss": 0.448247492313385,
          "validation_loss": 0.04502889886498451,
          "train_metric": 0.04514584317803383,
          "validation_metric": 0.004535145591944456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5633908720010368
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.4578054128214717,
          "train_loss": 0.4497813880443573,
          "validation_loss": 0.022640923038125038,
          "train_metric": 0.04530033469200134,
          "validation_metric": 0.0022803107276558876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5768870110005082
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.45466248970478773,
          "train_loss": 0.44620800018310547,
          "validation_loss": 0.039916276931762695,
          "train_metric": 0.04494043439626694,
          "validation_metric": 0.0040202210657298565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5898174480007583
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.44875627930741757,
          "train_loss": 0.4486849009990692,
          "validation_loss": 0.045239124447107315,
          "train_metric": 0.04518989846110344,
          "validation_metric": 0.004556318745017052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6019304140008899
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.45506242010742426,
          "train_loss": 0.4593687355518341,
          "validation_loss": 0.0665869489312172,
          "train_metric": 0.046265941113233566,
          "validation_metric": 0.006706393323838711,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6144440620009846
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.4564667334780097,
          "train_loss": 0.4457947015762329,
          "validation_loss": 0.03135795146226883,
          "train_metric": 0.044898808002471924,
          "validation_metric": 0.003158257808536291,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6263728590001847
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.46093868650496006,
          "train_loss": 0.45640018582344055,
          "validation_loss": 0.026277367025613785,
          "train_metric": 0.04596695676445961,
          "validation_metric": 0.002646559849381447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6383548860012525
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.4602956287562847,
          "train_loss": 0.45489564538002014,
          "validation_loss": 0.035917963832616806,
          "train_metric": 0.045815423130989075,
          "validation_metric": 0.003617525566369295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6507210409999971
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.4567730398848653,
          "train_loss": 0.4491254389286041,
          "validation_loss": 0.027994636446237564,
          "train_metric": 0.045234259217977524,
          "validation_metric": 0.0028195171616971493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6624226969997835
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.4600798471365124,
          "train_loss": 0.44552186131477356,
          "validation_loss": 0.036951109766960144,
          "train_metric": 0.04487133026123047,
          "validation_metric": 0.003721579909324646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6845369980001124
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.4582110242918134,
          "train_loss": 0.45083722472190857,
          "validation_loss": 0.04066529870033264,
          "train_metric": 0.04540667310357094,
          "validation_metric": 0.004095659591257572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7110393889997795
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.44755968544632196,
          "train_loss": 0.45257875323295593,
          "validation_loss": 0.03459788113832474,
          "train_metric": 0.04558207467198372,
          "validation_metric": 0.003484572283923626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7404195510007412
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.45624739420600235,
          "train_loss": 0.45292559266090393,
          "validation_loss": 0.05981434881687164,
          "train_metric": 0.04561701416969299,
          "validation_metric": 0.006024281494319439,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7676148800001101
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.45226658321917057,
          "train_loss": 0.44595324993133545,
          "validation_loss": 0.03254307433962822,
          "train_metric": 0.04491477832198143,
          "validation_metric": 0.003277619369328022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7938779880005313
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.4487843941897154,
          "train_loss": 0.44519415497779846,
          "validation_loss": 0.0440913587808609,
          "train_metric": 0.044838327914476395,
          "validation_metric": 0.004440720658749342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.817524329000662
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.44964618887752295,
          "train_loss": 0.4454033076763153,
          "validation_loss": 0.03324204310774803,
          "train_metric": 0.0448593907058239,
          "validation_metric": 0.003348016180098057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.831331997000234
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.45117867086082697,
          "train_loss": 0.4468100070953369,
          "validation_loss": 0.036789942532777786,
          "train_metric": 0.0450010672211647,
          "validation_metric": 0.003705347655341029,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8442239129999507
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.4514299649745226,
          "train_loss": 0.44863221049308777,
          "validation_loss": 0.020992068573832512,
          "train_metric": 0.04518459364771843,
          "validation_metric": 0.002114244271069765,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8573266550010885
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.4499249327927828,
          "train_loss": 0.4434334933757782,
          "validation_loss": 0.03723693639039993,
          "train_metric": 0.04466099664568901,
          "validation_metric": 0.003750368021428585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8705696610013547
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.4456436764448881,
          "train_loss": 0.4607108533382416,
          "validation_loss": 0.028266891837120056,
          "train_metric": 0.04640110954642296,
          "validation_metric": 0.002846937859430909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8839775810010906
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.4543207325041294,
          "train_loss": 0.443825900554657,
          "validation_loss": 0.04322343319654465,
          "train_metric": 0.04470052570104599,
          "validation_metric": 0.004353306256234646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8973139740010083
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.45230016438290477,
          "train_loss": 0.4476718306541443,
          "validation_loss": 0.04844824969768524,
          "train_metric": 0.045087866485118866,
          "validation_metric": 0.004879530519247055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9101660220003396
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.45832579024136066,
          "train_loss": 0.46341755986213684,
          "validation_loss": 0.05514805391430855,
          "train_metric": 0.0466737262904644,
          "validation_metric": 0.005554309580475092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.922131459999946
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.4550164146348834,
          "train_loss": 0.4447096586227417,
          "validation_loss": 0.04249831661581993,
          "train_metric": 0.044789526611566544,
          "validation_metric": 0.0042802742682397366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9344279120014107
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.4573918841779232,
          "train_loss": 0.46633458137512207,
          "validation_loss": 0.04197141155600548,
          "train_metric": 0.046967506408691406,
          "validation_metric": 0.004227207042276859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9470662800013088
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.45568590983748436,
          "train_loss": 0.4477614462375641,
          "validation_loss": 0.03357980027794838,
          "train_metric": 0.04509689658880234,
          "validation_metric": 0.0033820345997810364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9601720750015375
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.4498795915860683,
          "train_loss": 0.46561533212661743,
          "validation_loss": 0.08556371182203293,
          "train_metric": 0.046895068138837814,
          "validation_metric": 0.008617663756012917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9721373000011226
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.45767307467758656,
          "train_loss": 0.4471234381198883,
          "validation_loss": 0.019085068255662918,
          "train_metric": 0.04503263533115387,
          "validation_metric": 0.001922178198583424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9850164090003091
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.44701090711168945,
          "train_loss": 0.44442689418792725,
          "validation_loss": 0.045573923736810684,
          "train_metric": 0.0447610504925251,
          "validation_metric": 0.004590038675814867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9978945920011029
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.4470449477666989,
          "train_loss": 0.4421254098415375,
          "validation_loss": 0.035837844014167786,
          "train_metric": 0.044529251754283905,
          "validation_metric": 0.003609456354752183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0101014140000188
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.45312035363167524,
          "train_loss": 0.447226881980896,
          "validation_loss": 0.025150932371616364,
          "train_metric": 0.045043062418699265,
          "validation_metric": 0.0025331098586320877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.022247282000535
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.45613514445722103,
          "train_loss": 0.45527589321136475,
          "validation_loss": 0.03257346898317337,
          "train_metric": 0.04585372284054756,
          "validation_metric": 0.0032806803938001394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0354998239999986
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.4552249498665333,
          "train_loss": 0.44262292981147766,
          "validation_loss": 0.042065054178237915,
          "train_metric": 0.044579364359378815,
          "validation_metric": 0.00423663854598999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.050018498001009
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.45048741390928626,
          "train_loss": 0.44754281640052795,
          "validation_loss": 0.020735640078783035,
          "train_metric": 0.04507487639784813,
          "validation_metric": 0.0020884175319224596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0625168950009538
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.45119931804947555,
          "train_loss": 0.44219326972961426,
          "validation_loss": 0.033274129033088684,
          "train_metric": 0.04453608766198158,
          "validation_metric": 0.0033512485679239035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.074435425000047
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.4448435890953988,
          "train_loss": 0.4421235918998718,
          "validation_loss": 0.03808201849460602,
          "train_metric": 0.044529061764478683,
          "validation_metric": 0.0038354818243533373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0861739629999647
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.4476005891337991,
          "train_loss": 0.44729524850845337,
          "validation_loss": 0.05271569639444351,
          "train_metric": 0.045049939304590225,
          "validation_metric": 0.005309331696480513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.098673508000502
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.45265840482898057,
          "train_loss": 0.4419906735420227,
          "validation_loss": 0.03034554421901703,
          "train_metric": 0.04451568424701691,
          "validation_metric": 0.0030562926549464464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1107659130011598
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.4479385642334819,
          "train_loss": 0.4428245425224304,
          "validation_loss": 0.03359167277812958,
          "train_metric": 0.04459966719150543,
          "validation_metric": 0.0033832299523055553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1226614470015193
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.4484796413453296,
          "train_loss": 0.4572812616825104,
          "validation_loss": 0.06644178181886673,
          "train_metric": 0.046055685728788376,
          "validation_metric": 0.00669177295640111,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1344457820014213
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.45350265013985336,
          "train_loss": 0.4431055784225464,
          "validation_loss": 0.030999265611171722,
          "train_metric": 0.04462796449661255,
          "validation_metric": 0.003122132271528244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1469295070000953
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.44793989788740873,
          "train_loss": 0.4422091543674469,
          "validation_loss": 0.02754894271492958,
          "train_metric": 0.04453768953680992,
          "validation_metric": 0.002774628810584545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1674199670014787
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.44367310125380754,
          "train_loss": 0.4593409299850464,
          "validation_loss": 0.06205276399850845,
          "train_metric": 0.046263132244348526,
          "validation_metric": 0.00624972814694047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1849923340014357
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.4650671852286905,
          "train_loss": 0.4478142261505127,
          "validation_loss": 0.02052300237119198,
          "train_metric": 0.045102208852767944,
          "validation_metric": 0.002067001536488533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2139074610004172
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.4596380051225424,
          "train_loss": 0.44475626945495605,
          "validation_loss": 0.025068430230021477,
          "train_metric": 0.04479421675205231,
          "validation_metric": 0.0025248003657907248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2350166960004572
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.451009432785213,
          "train_loss": 0.4459236264228821,
          "validation_loss": 0.047567982226610184,
          "train_metric": 0.04491179436445236,
          "validation_metric": 0.004790872801095247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2653203260015289
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.44617454416584224,
          "train_loss": 0.44194066524505615,
          "validation_loss": 0.023828912526369095,
          "train_metric": 0.044510647654533386,
          "validation_metric": 0.002399961231276393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2879099620004126
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.4467276372015476,
          "train_loss": 0.44305866956710815,
          "validation_loss": 0.0339503213763237,
          "train_metric": 0.044623248279094696,
          "validation_metric": 0.0034193519968539476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.305672141001196
        }
      ],
      "best": {
        "epoch": 70,
        "updates": 1120
      },
      "best_validation_metric": 0.001922178198583424,
      "config": {
        "label": "MSE + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": true,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.04384265273205334
    },
    "7a196ed2cdbd_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.3273510686121881,
          "train_loss": 0.11705735325813293,
          "validation_loss": 0.07973263412714005,
          "train_metric": 0.011789587326347828,
          "validation_metric": 0.008030379191040993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02096395599983225
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.08486356446519494,
          "train_loss": 0.05732471123337746,
          "validation_loss": 0.07251747697591782,
          "train_metric": 0.005773534998297691,
          "validation_metric": 0.007303694728761911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04443592700044974
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.04616492288187146,
          "train_loss": 0.032562948763370514,
          "validation_loss": 0.03916358947753906,
          "train_metric": 0.0032796207815408707,
          "validation_metric": 0.003944413270801306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06561307499941904
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.029842033341992646,
          "train_loss": 0.025069084018468857,
          "validation_loss": 0.03333912417292595,
          "train_metric": 0.002524866256862879,
          "validation_metric": 0.003357794601470232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0862153109992505
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.0214119884185493,
          "train_loss": 0.017583012580871582,
          "validation_loss": 0.024185124784708023,
          "train_metric": 0.0017708966042846441,
          "validation_metric": 0.0024358374066650867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10669878299995617
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.016405746253440157,
          "train_loss": 0.013314454816281796,
          "validation_loss": 0.017470011487603188,
          "train_metric": 0.0013409830862656236,
          "validation_metric": 0.0017595157260075212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12800912000056996
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.014689415314933285,
          "train_loss": 0.015650659799575806,
          "validation_loss": 0.026589423418045044,
          "train_metric": 0.001576277194544673,
          "validation_metric": 0.0026779891923069954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1478075289996923
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.01493760384619236,
          "train_loss": 0.018550381064414978,
          "validation_loss": 0.023763317614793777,
          "train_metric": 0.001868326566182077,
          "validation_metric": 0.002393354196101427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1674417190006352
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.015906014014035463,
          "train_loss": 0.008983106352388859,
          "validation_loss": 0.01192385982722044,
          "train_metric": 0.0009047455387189984,
          "validation_metric": 0.001200927421450615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18810527099958563
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.009969578837626614,
          "train_loss": 0.011814270168542862,
          "validation_loss": 0.011308618821203709,
          "train_metric": 0.0011898900847882032,
          "validation_metric": 0.0011389626888558269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2089575820009486
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.01114819009671919,
          "train_loss": 0.011215937323868275,
          "validation_loss": 0.015419055707752705,
          "train_metric": 0.0011296281591057777,
          "validation_metric": 0.0015529508236795664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2291823569994449
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.011344271610141732,
          "train_loss": 0.00578324357047677,
          "validation_loss": 0.007838734425604343,
          "train_metric": 0.0005824671825394034,
          "validation_metric": 0.0007894885493442416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.251707057999738
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.007560007594292983,
          "train_loss": 0.005091303959488869,
          "validation_loss": 0.009888500906527042,
          "train_metric": 0.0005127774784341455,
          "validation_metric": 0.0009959336603060365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2719943409992993
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.007257838544319384,
          "train_loss": 0.005637935362756252,
          "validation_loss": 0.008928872644901276,
          "train_metric": 0.0005678322049789131,
          "validation_metric": 0.0008992831571958959,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29338495900083217
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.00623394575814018,
          "train_loss": 0.012971029616892338,
          "validation_loss": 0.0078086103312671185,
          "train_metric": 0.0013063945807516575,
          "validation_metric": 0.000786454533226788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3143698999992921
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.013430104008875787,
          "train_loss": 0.007163480389863253,
          "validation_loss": 0.009570866823196411,
          "train_metric": 0.0007214794750325382,
          "validation_metric": 0.0009639424388296902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33418628800063743
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.008007690135855228,
          "train_loss": 0.006668552756309509,
          "validation_loss": 0.007338789291679859,
          "train_metric": 0.0006716322968713939,
          "validation_metric": 0.0007391360122710466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35551116699934937
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.007334137088037096,
          "train_loss": 0.011812794022262096,
          "validation_loss": 0.011058768257498741,
          "train_metric": 0.0011897413060069084,
          "validation_metric": 0.001113798702135682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37559772200074804
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.006115007839980535,
          "train_loss": 0.0056520020589232445,
          "validation_loss": 0.010433555580675602,
          "train_metric": 0.0005692489794455469,
          "validation_metric": 0.0010508294217288494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.395116028999837
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.004616073609213345,
          "train_loss": 0.003704607952386141,
          "validation_loss": 0.0050687710754573345,
          "train_metric": 0.00037311456981115043,
          "validation_metric": 0.0005105080199427903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4156688570001279
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.0033525712133268826,
          "train_loss": 0.005505973938852549,
          "validation_loss": 0.005003985017538071,
          "train_metric": 0.0005545414751395583,
          "validation_metric": 0.0005039830575697124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43691268200018385
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.0048169265355682,
          "train_loss": 0.0031576580367982388,
          "validation_loss": 0.004121873993426561,
          "train_metric": 0.0003180277126375586,
          "validation_metric": 0.00041514006443321705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4616393150008662
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.004369251444586553,
          "train_loss": 0.006998151540756226,
          "validation_loss": 0.005888920743018389,
          "train_metric": 0.0007048281840980053,
          "validation_metric": 0.000593110453337431,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48187072099972283
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.006612358818529174,
          "train_loss": 0.004325545392930508,
          "validation_loss": 0.006397956516593695,
          "train_metric": 0.0004356531717348844,
          "validation_metric": 0.0006443787715397775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5023767580005369
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.005372359715693165,
          "train_loss": 0.01272423006594181,
          "validation_loss": 0.004804378841072321,
          "train_metric": 0.0012815379304811358,
          "validation_metric": 0.00048387947026640177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5218237560002308
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.013801360037177801,
          "train_loss": 0.010914361104369164,
          "validation_loss": 0.009418616071343422,
          "train_metric": 0.0010992544703185558,
          "validation_metric": 0.0009486085036769509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5420058700001391
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.023850818717619404,
          "train_loss": 0.017270682379603386,
          "validation_loss": 0.013394763693213463,
          "train_metric": 0.001739440020173788,
          "validation_metric": 0.0013490712735801935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5622072959995421
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.011587520159082487,
          "train_loss": 0.006311653647571802,
          "validation_loss": 0.006933359429240227,
          "train_metric": 0.0006356866215355694,
          "validation_metric": 0.0006983026396483183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5821066700009396
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.007428047931171022,
          "train_loss": 0.008943111635744572,
          "validation_loss": 0.010714882984757423,
          "train_metric": 0.0009007173939608037,
          "validation_metric": 0.001079163863323629,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6022911960008059
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.008485349782858975,
          "train_loss": 0.004578865133225918,
          "validation_loss": 0.006117227487266064,
          "train_metric": 0.00046116652083583176,
          "validation_metric": 0.0006161048077046871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6222045970007457
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.005771150841610506,
          "train_loss": 0.008173983544111252,
          "validation_loss": 0.00479541951790452,
          "train_metric": 0.0008232537657022476,
          "validation_metric": 0.0004829770477954298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6417634429999453
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.007144150244130287,
          "train_loss": 0.012213042005896568,
          "validation_loss": 0.01698092184960842,
          "train_metric": 0.0012300527887418866,
          "validation_metric": 0.0017102562123909593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6664108340009989
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.005899935073102824,
          "train_loss": 0.0051960451528429985,
          "validation_loss": 0.003369505750015378,
          "train_metric": 0.0005233266274444759,
          "validation_metric": 0.0003393642255105078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6896921349998593
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.005917149548622547,
          "train_loss": 0.007233455311506987,
          "validation_loss": 0.007564662955701351,
          "train_metric": 0.000728527142200619,
          "validation_metric": 0.0007618850795552135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7089088810007524
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.004864623650064459,
          "train_loss": 0.0051497905515134335,
          "validation_loss": 0.00282531906850636,
          "train_metric": 0.0005186679773032665,
          "validation_metric": 0.0002845557755790651,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7299274510005489
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.00409684213809669,
          "train_loss": 0.006532116327434778,
          "validation_loss": 0.007584122009575367,
          "train_metric": 0.0006578908069059253,
          "validation_metric": 0.0007638449314981699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7502091130008921
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.009312566791777499,
          "train_loss": 0.005541311576962471,
          "validation_loss": 0.005547116044908762,
          "train_metric": 0.0005581005825661123,
          "validation_metric": 0.0005586851621046662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7700094940009876
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.017059682053513825,
          "train_loss": 0.008513408713042736,
          "validation_loss": 0.006189574953168631,
          "train_metric": 0.0008574392995797098,
          "validation_metric": 0.0006233913591131568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7888351950005017
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.0177556861017365,
          "train_loss": 0.019069751724600792,
          "validation_loss": 0.025645239278674126,
          "train_metric": 0.0019206355791538954,
          "validation_metric": 0.0025828946381807327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8080010459998448
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012423908628989011,
          "train_loss": 0.0077280704863369465,
          "validation_loss": 0.007089395076036453,
          "train_metric": 0.0007783429464325309,
          "validation_metric": 0.0007140179513953626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8278509720003058
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.0052849889907520264,
          "train_loss": 0.003457134822383523,
          "validation_loss": 0.0031268971506506205,
          "train_metric": 0.0003481899912003428,
          "validation_metric": 0.00031492963898926973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8475070250005956
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.003456670143350493,
          "train_loss": 0.0027233094442635775,
          "validation_loss": 0.0024531311355531216,
          "train_metric": 0.0002742817741818726,
          "validation_metric": 0.000247070420300588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8725859460009815
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.0057835834377328865,
          "train_loss": 0.004220916423946619,
          "validation_loss": 0.0033909312915056944,
          "train_metric": 0.0004251152276992798,
          "validation_metric": 0.0003415220999158919,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.894050884000535
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.004961211503541563,
          "train_loss": 0.0027626005467027426,
          "validation_loss": 0.0027505706530064344,
          "train_metric": 0.0002782390220090747,
          "validation_metric": 0.00027702737133949995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9136788769992563
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.004748126368212979,
          "train_loss": 0.00945703499019146,
          "validation_loss": 0.007011713460087776,
          "train_metric": 0.0009524778579361737,
          "validation_metric": 0.0007061940850690007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9338624960000743
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.004740721691632643,
          "train_loss": 0.0028950944542884827,
          "validation_loss": 0.0023027523420751095,
          "train_metric": 0.00029158330289646983,
          "validation_metric": 0.0002319247869309038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9541838250006549
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.002202396521170158,
          "train_loss": 0.001396849867887795,
          "validation_loss": 0.001685641473159194,
          "train_metric": 0.0001406856026733294,
          "validation_metric": 0.00016977165068965405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9751968750006199
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0018137458646378946,
          "train_loss": 0.0035003693774342537,
          "validation_loss": 0.002999718999490142,
          "train_metric": 0.0003525444190017879,
          "validation_metric": 0.0003021206648554653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9951583550009673
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.00825509933201829,
          "train_loss": 0.017938241362571716,
          "validation_loss": 0.021720224991440773,
          "train_metric": 0.0018066740594804287,
          "validation_metric": 0.0021875817328691483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0147993820010015
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.01623914416995831,
          "train_loss": 0.010971436277031898,
          "validation_loss": 0.008731418289244175,
          "train_metric": 0.001105002942495048,
          "validation_metric": 0.0008793964516371489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.034158980999564
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007084169847075827,
          "train_loss": 0.004891781136393547,
          "validation_loss": 0.005322495009750128,
          "train_metric": 0.0004926822730340064,
          "validation_metric": 0.0005360622308216989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0539297430004808
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.004670059774070978,
          "train_loss": 0.0025661683175712824,
          "validation_loss": 0.004384555388242006,
          "train_metric": 0.0002584551111795008,
          "validation_metric": 0.0004415963776409626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0775504930006718
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.0026889196306001395,
          "train_loss": 0.001824769889935851,
          "validation_loss": 0.0024213145952671766,
          "train_metric": 0.00018378414097242057,
          "validation_metric": 0.0002438659139443189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0975595740001154
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.002969987181131728,
          "train_loss": 0.0016826242208480835,
          "validation_loss": 0.002400194527581334,
          "train_metric": 0.00016946776304394007,
          "validation_metric": 0.0002417388022877276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.119520112999453
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.002957138916826807,
          "train_loss": 0.0018714220495894551,
          "validation_loss": 0.001718224142678082,
          "train_metric": 0.00018848275067284703,
          "validation_metric": 0.00017305325309280306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1408270329993684
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.002361739312618738,
          "train_loss": 0.0021901200525462627,
          "validation_loss": 0.0017026650020852685,
          "train_metric": 0.0002205808414146304,
          "validation_metric": 0.0001714861864456907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1606391389996134
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.0040892575270845555,
          "train_loss": 0.004727373830974102,
          "validation_loss": 0.006496707908809185,
          "train_metric": 0.00047612376511096954,
          "validation_metric": 0.0006543245981447399,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.180609478000406
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.002474357552273432,
          "train_loss": 0.001369926263578236,
          "validation_loss": 0.0023739011958241463,
          "train_metric": 0.00013797394058201462,
          "validation_metric": 0.00023909061565063894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2007210869996925
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.0016907036460906966,
          "train_loss": 0.0010246984893456101,
          "validation_loss": 0.002347027650102973,
          "train_metric": 0.00010320388537365943,
          "validation_metric": 0.00023638404672965407,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.220105712000077
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.002186627702030819,
          "train_loss": 0.0018552659312263131,
          "validation_loss": 0.0031346564646810293,
          "train_metric": 0.00018685555551201105,
          "validation_metric": 0.0003157111059408635,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.240023194000969
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0026110352628165856,
          "train_loss": 0.002827586606144905,
          "validation_loss": 0.0014451548922806978,
          "train_metric": 0.00028478421154432,
          "validation_metric": 0.00014555070083588362,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2613929630006169
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.001241714559000684,
          "train_loss": 0.0010449594119563699,
          "validation_loss": 0.001358612789772451,
          "train_metric": 0.00010524448589421809,
          "validation_metric": 0.00013683449651580304,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2879518829995504
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.001329095244727796,
          "train_loss": 0.0009994227439165115,
          "validation_loss": 0.002490328624844551,
          "train_metric": 0.00010065818059956655,
          "validation_metric": 0.00025081675266847014,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3104488600001787
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0011292404706182424,
          "train_loss": 0.0009754827478900552,
          "validation_loss": 0.0012510214000940323,
          "train_metric": 9.82470519375056e-05,
          "validation_metric": 0.0001259982818737626,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3317589380003483
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.002585577301942976,
          "train_loss": 0.004744133912026882,
          "validation_loss": 0.0025751451030373573,
          "train_metric": 0.000477811845485121,
          "validation_metric": 0.00025935916346497834,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3519751769999857
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0021348100490286015,
          "train_loss": 0.001119809108786285,
          "validation_loss": 0.002103478414937854,
          "train_metric": 0.00011278307647444308,
          "validation_metric": 0.00021185463992878795,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3720479130006424
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0011001978709828109,
          "train_loss": 0.0011749510886147618,
          "validation_loss": 0.0023703998886048794,
          "train_metric": 0.00011833676398964599,
          "validation_metric": 0.0002387380227446556,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3917808789992705
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0009743360333231976,
          "train_loss": 0.0009154492290690541,
          "validation_loss": 0.0019180644303560257,
          "train_metric": 9.220070205628872e-05,
          "validation_metric": 0.00019318041449878365,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.41103128100076
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0009681366900622379,
          "train_loss": 0.0013804984046146274,
          "validation_loss": 0.002975519048050046,
          "train_metric": 0.0001390387478750199,
          "validation_metric": 0.00029968336457386613,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4306239600009576
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0010279269354214193,
          "train_loss": 0.0028919046744704247,
          "validation_loss": 0.004589701537042856,
          "train_metric": 0.0002912620548158884,
          "validation_metric": 0.00046225794358178973,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4507876229999965
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0012693619246419985,
          "train_loss": 0.0009471553494222462,
          "validation_loss": 0.0011858876096084714,
          "train_metric": 9.539402526570484e-05,
          "validation_metric": 0.0001194382639368996,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4720026239992876
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0012232228746142937,
          "train_loss": 0.001447721617296338,
          "validation_loss": 0.0012160086771473289,
          "train_metric": 0.00014580923016183078,
          "validation_metric": 0.0001224719308083877,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4931387420001556
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0011924395475944038,
          "train_loss": 0.001938000787049532,
          "validation_loss": 0.0011832054005935788,
          "train_metric": 0.0001951883314177394,
          "validation_metric": 0.00011916812945855781,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5150236640001822
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.002543218513892498,
          "train_loss": 0.0008852295577526093,
          "validation_loss": 0.001490240916609764,
          "train_metric": 8.915708895074204e-05,
          "validation_metric": 0.00015009159687906504,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5354485120005847
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0012029782628815155,
          "train_loss": 0.0010417609009891748,
          "validation_loss": 0.0011980276321992278,
          "train_metric": 0.0001049223355948925,
          "validation_metric": 0.00012066094495821744,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.555818230001023
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0015144028839131352,
          "train_loss": 0.0008142028236761689,
          "validation_loss": 0.0023483149707317352,
          "train_metric": 8.200352021958679e-05,
          "validation_metric": 0.0002365136897424236,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5774377260004258
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.0035857947914337274,
          "train_loss": 0.0038422574289143085,
          "validation_loss": 0.005409252364188433,
          "train_metric": 0.0003869781503453851,
          "validation_metric": 0.000544800132047385,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5970725549996132
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.002301683351106476,
          "train_loss": 0.001555003458634019,
          "validation_loss": 0.0021782456897199154,
          "train_metric": 0.00015661426004953682,
          "validation_metric": 0.00021938493591733277,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.616809529999955
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0013299314014147967,
          "train_loss": 0.0009656188194639981,
          "validation_loss": 0.002070009708404541,
          "train_metric": 9.725357813294977e-05,
          "validation_metric": 0.00020848379062954336,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6368894339993858
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0011596536223805742,
          "train_loss": 0.0016610492020845413,
          "validation_loss": 0.000978475552983582,
          "train_metric": 0.0001672947983024642,
          "validation_metric": 9.854847303358838e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6580335310009104
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0009832189298322191,
          "train_loss": 0.0006549770478159189,
          "validation_loss": 0.0011876774951815605,
          "train_metric": 6.596689490834251e-05,
          "validation_metric": 0.000119618525786791,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6781502480007475
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0008994660329335602,
          "train_loss": 0.002889293944463134,
          "validation_loss": 0.0017420437652617693,
          "train_metric": 0.00029099907260388136,
          "validation_metric": 0.00017545228183735162,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6996397109996906
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0023601777393196244,
          "train_loss": 0.0012349584139883518,
          "validation_loss": 0.0028798857238143682,
          "train_metric": 0.00012438050180207938,
          "validation_metric": 0.00029005156829953194,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.722020410999903
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.001243471630004933,
          "train_loss": 0.0008578114211559296,
          "validation_loss": 0.0011556565295904875,
          "train_metric": 8.639562292955816e-05,
          "validation_metric": 0.00011639349395409226,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.743413010999575
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0007214927536551841,
          "train_loss": 0.0005700846086256206,
          "validation_loss": 0.0014415085315704346,
          "train_metric": 5.7416829804424196e-05,
          "validation_metric": 0.00014518346870318055,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7631180419994052
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.000755388392462919,
          "train_loss": 0.0013481059577316046,
          "validation_loss": 0.0008707201341167092,
          "train_metric": 0.00013577629579231143,
          "validation_metric": 8.769574196776375e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7841946560001816
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0008483338970108889,
          "train_loss": 0.00045457581290975213,
          "validation_loss": 0.0016026702942326665,
          "train_metric": 4.5783213863614947e-05,
          "validation_metric": 0.00016141508240252733,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8049039470006392
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0006797111964260694,
          "train_loss": 0.0005602188175544143,
          "validation_loss": 0.0014226867351680994,
          "train_metric": 5.642318501486443e-05,
          "validation_metric": 0.00014328780525829643,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8259347820003313
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0010003119705288555,
          "train_loss": 0.0010615314822643995,
          "validation_loss": 0.0010130543960258365,
          "train_metric": 0.00010691357601899654,
          "validation_metric": 0.00010203113197349012,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8455364420005935
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0009537421283312142,
          "train_loss": 0.0010391671676188707,
          "validation_loss": 0.0008658661972731352,
          "train_metric": 0.0001046611214405857,
          "validation_metric": 8.720689220353961e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8661190160000842
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0008539658992958721,
          "train_loss": 0.0008721183403395116,
          "validation_loss": 0.000996188260614872,
          "train_metric": 8.783656812738627e-05,
          "validation_metric": 0.00010033244325313717,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8866807560007146
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.003267318123107543,
          "train_loss": 0.009412392973899841,
          "validation_loss": 0.006077543832361698,
          "train_metric": 0.0009479817235842347,
          "validation_metric": 0.0006121079786680639,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.906297421999625
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.004925288107187953,
          "train_loss": 0.006673025898635387,
          "validation_loss": 0.010198971256613731,
          "train_metric": 0.0006720827659592032,
          "validation_metric": 0.0010272030485793948,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9303334590003942
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.003629779770562891,
          "train_loss": 0.001246808678843081,
          "validation_loss": 0.0023394180461764336,
          "train_metric": 0.00012557400623336434,
          "validation_metric": 0.000235617597354576,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9528267050009163
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0019837770269077737,
          "train_loss": 0.0015553507255390286,
          "validation_loss": 0.0011455613421276212,
          "train_metric": 0.00015664921374991536,
          "validation_metric": 0.00011537674436112866,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9728715340006602
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0015487518030568026,
          "train_loss": 0.0014964089496061206,
          "validation_loss": 0.0014994633384048939,
          "train_metric": 0.00015071284724399447,
          "validation_metric": 0.00015102046018000692,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9927358940003614
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0009401635516042006,
          "train_loss": 0.000647250097244978,
          "validation_loss": 0.0016316617839038372,
          "train_metric": 6.518865848192945e-05,
          "validation_metric": 0.00016433498240076005,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 2.012670412999796
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0007099347803887213,
          "train_loss": 0.000858257117215544,
          "validation_loss": 0.0010052421130239964,
          "train_metric": 8.644051558803767e-05,
          "validation_metric": 0.00010124430264113471,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 2.0325797850000527
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.000753646316297818,
          "train_loss": 0.0006416055257432163,
          "validation_loss": 0.0017484031850472093,
          "train_metric": 6.462015880970284e-05,
          "validation_metric": 0.00017609276983421296,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 2.0520583990000887
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.000870500393830298,
          "train_loss": 0.0005054370849393308,
          "validation_loss": 0.001443113200366497,
          "train_metric": 5.0905768148368225e-05,
          "validation_metric": 0.0001453450822737068,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 2.071218498000235
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.000536837938852841,
          "train_loss": 0.0006600471679121256,
          "validation_loss": 0.0017880682135000825,
          "train_metric": 6.647754344157875e-05,
          "validation_metric": 0.00018008767801802605,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 2.0920487160001358
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0006218197595444508,
          "train_loss": 0.0005516476812772453,
          "validation_loss": 0.0010544236283749342,
          "train_metric": 5.555993629968725e-05,
          "validation_metric": 0.0001061976799974218,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.111599228999694
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0006659173614025349,
          "train_loss": 0.0016336627304553986,
          "validation_loss": 0.00351150194182992,
          "train_metric": 0.0001645365118747577,
          "validation_metric": 0.0003536656149663031,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.136295883999992
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0008614836315246066,
          "train_loss": 0.0005910834297537804,
          "validation_loss": 0.0010836105793714523,
          "train_metric": 5.953175786999054e-05,
          "validation_metric": 0.00010913728328887373,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1571751850005967
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0005452323448480456,
          "train_loss": 0.0005029039457440376,
          "validation_loss": 0.0016401938628405333,
          "train_metric": 5.06506439705845e-05,
          "validation_metric": 0.00016519430209882557,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1774724059996515
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0005720770977859502,
          "train_loss": 0.0007531093433499336,
          "validation_loss": 0.0008963328436948359,
          "train_metric": 7.585040293633938e-05,
          "validation_metric": 9.027535270433873e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.198061441000391
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0005463918541863677,
          "train_loss": 0.0003383220755495131,
          "validation_loss": 0.0012939369771629572,
          "train_metric": 3.407456460990943e-05,
          "validation_metric": 0.00013032059359829873,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2179883860007976
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0005097633938930812,
          "train_loss": 0.00038866724935360253,
          "validation_loss": 0.0011109947226941586,
          "train_metric": 3.914513945346698e-05,
          "validation_metric": 0.0001118953077821061,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.236693285000001
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0005125327179484884,
          "train_loss": 0.0003703611728269607,
          "validation_loss": 0.0012035761028528214,
          "train_metric": 3.7301419069990516e-05,
          "validation_metric": 0.0001212197748827748,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.255504609000127
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0004927896943627275,
          "train_loss": 0.00039821717655286193,
          "validation_loss": 0.0011946475133299828,
          "train_metric": 4.010697011835873e-05,
          "validation_metric": 0.00012032052472932264,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2751398210002662
        }
      ],
      "best": {
        "epoch": 90,
        "updates": 1440
      },
      "best_validation_metric": 8.720689220353961e-05,
      "config": {
        "label": "LayerNorm + relu + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "relu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009338463053604678
    },
    "d5c218b8a06d_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 1.2132763303816319,
          "train_loss": 0.8167810440063477,
          "validation_loss": 0.3234926462173462,
          "train_metric": 0.08226318657398224,
          "validation_metric": 0.03258099406957626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.018602788000862347
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.6342314295470715,
          "train_loss": 0.5575058460235596,
          "validation_loss": 0.12935969233512878,
          "train_metric": 0.05614994093775749,
          "validation_metric": 0.013028634712100029,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03764593500090996
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.5561380879953504,
          "train_loss": 0.5301071405410767,
          "validation_loss": 0.11835011094808578,
          "train_metric": 0.05339045077562332,
          "validation_metric": 0.011919789016246796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.056263327000124264
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.5216408148407936,
          "train_loss": 0.5128851532936096,
          "validation_loss": 0.0812135711312294,
          "train_metric": 0.05165592208504677,
          "validation_metric": 0.008179533295333385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07518999400053872
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.510880745947361,
          "train_loss": 0.5032582879066467,
          "validation_loss": 0.09775566309690475,
          "train_metric": 0.05068633332848549,
          "validation_metric": 0.009845593012869358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0924769320008636
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.5049624405801296,
          "train_loss": 0.5041108131408691,
          "validation_loss": 0.05911053717136383,
          "train_metric": 0.05077219009399414,
          "validation_metric": 0.005953396670520306,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11033312800100248
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.5024863481521606,
          "train_loss": 0.4954800009727478,
          "validation_loss": 0.06458129733800888,
          "train_metric": 0.04990293085575104,
          "validation_metric": 0.006504391320049763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12727742500101158
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.49563559773378074,
          "train_loss": 0.490771621465683,
          "validation_loss": 0.0861147865653038,
          "train_metric": 0.04942872002720833,
          "validation_metric": 0.0086731668561697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14432054700046137
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.4911167239770293,
          "train_loss": 0.48621734976768494,
          "validation_loss": 0.06462273001670837,
          "train_metric": 0.04897002503275871,
          "validation_metric": 0.00650856364518404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16123909800080582
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.4957964299246669,
          "train_loss": 0.4888230264186859,
          "validation_loss": 0.09016392379999161,
          "train_metric": 0.049232468008995056,
          "validation_metric": 0.009080979973077774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17806627299978572
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.49728860706090927,
          "train_loss": 0.48660948872566223,
          "validation_loss": 0.084101602435112,
          "train_metric": 0.049009524285793304,
          "validation_metric": 0.008470405824482441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1943288250004116
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.4906944069080055,
          "train_loss": 0.4845162630081177,
          "validation_loss": 0.08440437912940979,
          "train_metric": 0.048798710107803345,
          "validation_metric": 0.008500899188220501,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21752951400048914
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.49048071447759867,
          "train_loss": 0.48175930976867676,
          "validation_loss": 0.05945156514644623,
          "train_metric": 0.04852103441953659,
          "validation_metric": 0.005987744312733412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2578102229999786
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.4825219539925456,
          "train_loss": 0.47857871651649475,
          "validation_loss": 0.06835450232028961,
          "train_metric": 0.04820069670677185,
          "validation_metric": 0.006884415168315172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2746406880014547
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.4865980800241232,
          "train_loss": 0.4799634516239166,
          "validation_loss": 0.07027029246091843,
          "train_metric": 0.048340167850255966,
          "validation_metric": 0.007077366579324007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2909643170005438
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.477096913382411,
          "train_loss": 0.47546783089637756,
          "validation_loss": 0.061021678149700165,
          "train_metric": 0.04788737744092941,
          "validation_metric": 0.006145879626274109,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3079734330003703
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.48021393921226263,
          "train_loss": 0.47280311584472656,
          "validation_loss": 0.058286890387535095,
          "train_metric": 0.04761900007724762,
          "validation_metric": 0.005870442371815443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.32550149300004705
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.4780709780752659,
          "train_loss": 0.47951334714889526,
          "validation_loss": 0.08996700495481491,
          "train_metric": 0.048294827342033386,
          "validation_metric": 0.009061147458851337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34253272200112406
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.485353184863925,
          "train_loss": 0.46945732831954956,
          "validation_loss": 0.05990143120288849,
          "train_metric": 0.04728202149271965,
          "validation_metric": 0.0060330526903271675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35930024800109095
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.47796534188091755,
          "train_loss": 0.46753767132759094,
          "validation_loss": 0.05441322550177574,
          "train_metric": 0.047088686376810074,
          "validation_metric": 0.005480300635099411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3771832720012753
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.4791602718178183,
          "train_loss": 0.4845353960990906,
          "validation_loss": 0.035921961069107056,
          "train_metric": 0.048800624907016754,
          "validation_metric": 0.0036179281305521727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3953512200005207
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.4888176657259464,
          "train_loss": 0.48510029911994934,
          "validation_loss": 0.06871698796749115,
          "train_metric": 0.04885752871632576,
          "validation_metric": 0.006920923944562674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.412351615001171
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.48372018430382013,
          "train_loss": 0.4647457003593445,
          "validation_loss": 0.04795319214463234,
          "train_metric": 0.046807486563920975,
          "validation_metric": 0.004829669836908579,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4286459750001086
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.46974114421755075,
          "train_loss": 0.46312054991722107,
          "validation_loss": 0.042788803577423096,
          "train_metric": 0.04664381220936775,
          "validation_metric": 0.0043095313012599945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45483106299980136
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.47082600416615605,
          "train_loss": 0.4641077518463135,
          "validation_loss": 0.06827972829341888,
          "train_metric": 0.046743232756853104,
          "validation_metric": 0.0068768830969929695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48674497100000735
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.46686482802033424,
          "train_loss": 0.46779459714889526,
          "validation_loss": 0.03132595121860504,
          "train_metric": 0.04711456224322319,
          "validation_metric": 0.0031550349667668343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5045545009998023
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.4657664028927684,
          "train_loss": 0.4644176959991455,
          "validation_loss": 0.07221309840679169,
          "train_metric": 0.046774450689554214,
          "validation_metric": 0.007273039780557156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5215769490005187
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.4647680437192321,
          "train_loss": 0.4730559289455414,
          "validation_loss": 0.04118833318352699,
          "train_metric": 0.04764446243643761,
          "validation_metric": 0.004148338455706835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5386604710001848
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.47570820804685354,
          "train_loss": 0.48130732774734497,
          "validation_loss": 0.03660513833165169,
          "train_metric": 0.048475511372089386,
          "validation_metric": 0.0036867354065179825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5557588020001276
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.47693018964491785,
          "train_loss": 0.46180209517478943,
          "validation_loss": 0.06194908544421196,
          "train_metric": 0.04651101678609848,
          "validation_metric": 0.006239284295588732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.573255354000139
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.46422809385694563,
          "train_loss": 0.45661890506744385,
          "validation_loss": 0.054124802350997925,
          "train_metric": 0.045988984405994415,
          "validation_metric": 0.00545125175267458,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5904705130014918
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.46820839308202267,
          "train_loss": 0.46360456943511963,
          "validation_loss": 0.050795312970876694,
          "train_metric": 0.046692557632923126,
          "validation_metric": 0.005115917418152094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6079432690003159
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.46555255795829,
          "train_loss": 0.4574800729751587,
          "validation_loss": 0.03818688541650772,
          "train_metric": 0.046075716614723206,
          "validation_metric": 0.003846043022349477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6253210419999959
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.4597025252878666,
          "train_loss": 0.45476168394088745,
          "validation_loss": 0.049395132809877396,
          "train_metric": 0.045801930129528046,
          "validation_metric": 0.004974897485226393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6421787740000582
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.4599427720531821,
          "train_loss": 0.4650295376777649,
          "validation_loss": 0.06924092024564743,
          "train_metric": 0.04683607071638107,
          "validation_metric": 0.006973693147301674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6597739440003352
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.46915776608511806,
          "train_loss": 0.45287024974823,
          "validation_loss": 0.04088326916098595,
          "train_metric": 0.0456114336848259,
          "validation_metric": 0.004117613658308983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6771791610008222
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.47419218346476555,
          "train_loss": 0.4690296947956085,
          "validation_loss": 0.045552272349596024,
          "train_metric": 0.04723895341157913,
          "validation_metric": 0.004587858449667692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6937411560011242
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.46977793518453836,
          "train_loss": 0.45485153794288635,
          "validation_loss": 0.027604499831795692,
          "train_metric": 0.04581097885966301,
          "validation_metric": 0.002780224196612835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8078038420007942
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.4583455394022167,
          "train_loss": 0.4549722373485565,
          "validation_loss": 0.04675073176622391,
          "train_metric": 0.04582313820719719,
          "validation_metric": 0.004708562046289444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8254685740012064
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.45624313899315894,
          "train_loss": 0.45054134726524353,
          "validation_loss": 0.034293282777071,
          "train_metric": 0.04537687823176384,
          "validation_metric": 0.003453893819823861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8428567620012473
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.45564698055386543,
          "train_loss": 0.4507003724575043,
          "validation_loss": 0.043089233338832855,
          "train_metric": 0.04539288952946663,
          "validation_metric": 0.004339789506047964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8593703570004436
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.4555254112929106,
          "train_loss": 0.45254382491111755,
          "validation_loss": 0.029514415189623833,
          "train_metric": 0.04557855799794197,
          "validation_metric": 0.0029725837521255016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8762686760001088
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.45586897525936365,
          "train_loss": 0.4503338038921356,
          "validation_loss": 0.04146416485309601,
          "train_metric": 0.04535597190260887,
          "validation_metric": 0.00417611887678504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8927029960013897
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.4520233981311321,
          "train_loss": 0.4510681629180908,
          "validation_loss": 0.027728747576475143,
          "train_metric": 0.045429933816194534,
          "validation_metric": 0.002792737912386656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9197828940014006
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.4560909131541848,
          "train_loss": 0.46200236678123474,
          "validation_loss": 0.07180581241846085,
          "train_metric": 0.04653118550777435,
          "validation_metric": 0.007232018746435642,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.942141225999876
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.4645171072334051,
          "train_loss": 0.45820316672325134,
          "validation_loss": 0.04954813793301582,
          "train_metric": 0.04614855349063873,
          "validation_metric": 0.004990307614207268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9619370010004786
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.46312880609184504,
          "train_loss": 0.451050728559494,
          "validation_loss": 0.0541679672896862,
          "train_metric": 0.04542817920446396,
          "validation_metric": 0.005455599632114172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9801275480003824
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.4536345973610878,
          "train_loss": 0.45732685923576355,
          "validation_loss": 0.030904188752174377,
          "train_metric": 0.046060286462306976,
          "validation_metric": 0.0031125566456466913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9987120910009253
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.4577762736007571,
          "train_loss": 0.45665687322616577,
          "validation_loss": 0.028902431949973106,
          "train_metric": 0.045992810279130936,
          "validation_metric": 0.0029109471943229437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0188373220007634
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.4648487735539675,
          "train_loss": 0.45985615253448486,
          "validation_loss": 0.06493131816387177,
          "train_metric": 0.0463150292634964,
          "validation_metric": 0.006539644207805395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.038322012000208
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.4500563775654882,
          "train_loss": 0.4491775929927826,
          "validation_loss": 0.028535068035125732,
          "train_metric": 0.045239515602588654,
          "validation_metric": 0.0028739478439092636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0572557610012154
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.45671636145561934,
          "train_loss": 0.45260295271873474,
          "validation_loss": 0.033650267869234085,
          "train_metric": 0.045584507286548615,
          "validation_metric": 0.0033891319762915373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0753381110007467
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.45485985837876797,
          "train_loss": 0.44746139645576477,
          "validation_loss": 0.027879439294338226,
          "train_metric": 0.04506668075919151,
          "validation_metric": 0.0028079149778932333,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0933732930006954
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.45386314392089844,
          "train_loss": 0.4522029459476471,
          "validation_loss": 0.02741893008351326,
          "train_metric": 0.04554422199726105,
          "validation_metric": 0.002761533949524164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1135201590004726
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.45322565734386444,
          "train_loss": 0.44827619194984436,
          "validation_loss": 0.04650898277759552,
          "train_metric": 0.045148734003305435,
          "validation_metric": 0.0046842144802212715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1322425350008416
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.45430502388626337,
          "train_loss": 0.4530741572380066,
          "validation_loss": 0.03960351645946503,
          "train_metric": 0.04563196748495102,
          "validation_metric": 0.003988720942288637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1504128369997488
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.4564303494989872,
          "train_loss": 0.4468158781528473,
          "validation_loss": 0.03920767828822136,
          "train_metric": 0.04500165581703186,
          "validation_metric": 0.00394885428249836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1691631020003115
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.45682643190957606,
          "train_loss": 0.4489975571632385,
          "validation_loss": 0.05011985823512077,
          "train_metric": 0.045221395790576935,
          "validation_metric": 0.005047888495028019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2157627420001518
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.45424904860556126,
          "train_loss": 0.45345339179039,
          "validation_loss": 0.021708033978939056,
          "train_metric": 0.04567016661167145,
          "validation_metric": 0.002186353551223874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2630269970013615
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.46067300345748663,
          "train_loss": 0.45210734009742737,
          "validation_loss": 0.03465196490287781,
          "train_metric": 0.045534588396549225,
          "validation_metric": 0.0034900191240012646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2812637750012073
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.4535186958964914,
          "train_loss": 0.4486906826496124,
          "validation_loss": 0.030491380020976067,
          "train_metric": 0.045190487056970596,
          "validation_metric": 0.00307098007760942,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.298531163000007
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.45133756601717323,
          "train_loss": 0.4472711384296417,
          "validation_loss": 0.049828335642814636,
          "train_metric": 0.04504750669002533,
          "validation_metric": 0.005018528085201979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3154862339997635
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.4508542828261852,
          "train_loss": 0.4481826424598694,
          "validation_loss": 0.03100813366472721,
          "train_metric": 0.045139312744140625,
          "validation_metric": 0.0031230258755385876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3321670790010103
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.4522631783038378,
          "train_loss": 0.4467158019542694,
          "validation_loss": 0.03137901425361633,
          "train_metric": 0.04499157890677452,
          "validation_metric": 0.003160379361361265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3494862469997315
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.4512679358012974,
          "train_loss": 0.4638649821281433,
          "validation_loss": 0.029721157625317574,
          "train_metric": 0.0467187836766243,
          "validation_metric": 0.002993405796587467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3666649040005723
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.4533693250268698,
          "train_loss": 0.44934025406837463,
          "validation_loss": 0.05301603674888611,
          "train_metric": 0.045255910605192184,
          "validation_metric": 0.005339581519365311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3838522830010334
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.4552579978480935,
          "train_loss": 0.4518739879131317,
          "validation_loss": 0.021120673045516014,
          "train_metric": 0.04551108926534653,
          "validation_metric": 0.002127197105437517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4092485630008014
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.4585624923929572,
          "train_loss": 0.4515632390975952,
          "validation_loss": 0.034949496388435364,
          "train_metric": 0.04547979682683945,
          "validation_metric": 0.003519985591992736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4351222590012185
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.4620119514875114,
          "train_loss": 0.4549259543418884,
          "validation_loss": 0.023815177381038666,
          "train_metric": 0.04581848159432411,
          "validation_metric": 0.0023985777515918016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.471662305000791
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.4508094284683466,
          "train_loss": 0.44666925072669983,
          "validation_loss": 0.028841406106948853,
          "train_metric": 0.04498688876628876,
          "validation_metric": 0.002904800930991769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.488947805000862
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.4572400525212288,
          "train_loss": 0.4645861089229584,
          "validation_loss": 0.030613092705607414,
          "train_metric": 0.04679140821099281,
          "validation_metric": 0.0030832388438284397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.506106874001489
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.461317197419703,
          "train_loss": 0.45621949434280396,
          "validation_loss": 0.026435155421495438,
          "train_metric": 0.045948758721351624,
          "validation_metric": 0.002662452170625329,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.530351872999745
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.4593809777870774,
          "train_loss": 0.4443557858467102,
          "validation_loss": 0.03680255636572838,
          "train_metric": 0.04475389048457146,
          "validation_metric": 0.003706619143486023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.547715701000925
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.4568900535814464,
          "train_loss": 0.4501539170742035,
          "validation_loss": 0.050557080656290054,
          "train_metric": 0.045337848365306854,
          "validation_metric": 0.005091924220323563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5778058050000254
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.45444350875914097,
          "train_loss": 0.4528997540473938,
          "validation_loss": 0.05098661780357361,
          "train_metric": 0.04561440646648407,
          "validation_metric": 0.0051351855508983135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.604866896999738
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.4478973629884422,
          "train_loss": 0.44877585768699646,
          "validation_loss": 0.021212894469499588,
          "train_metric": 0.045199062675237656,
          "validation_metric": 0.0021364849526435137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.628901793001205
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.4531464781612158,
          "train_loss": 0.44392743706703186,
          "validation_loss": 0.0234550628811121,
          "train_metric": 0.04471074044704437,
          "validation_metric": 0.002362307859584689,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.650932324000678
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.4495970727875829,
          "train_loss": 0.44746723771095276,
          "validation_loss": 0.05045453459024429,
          "train_metric": 0.04506726190447807,
          "validation_metric": 0.005081596318632364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6829995300013252
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.4494123316835612,
          "train_loss": 0.44885435700416565,
          "validation_loss": 0.03757021203637123,
          "train_metric": 0.04520696774125099,
          "validation_metric": 0.003783934284001589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.705987058001483
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.45438051561359316,
          "train_loss": 0.4432876706123352,
          "validation_loss": 0.031484831124544144,
          "train_metric": 0.04464631527662277,
          "validation_metric": 0.0031710369512438774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7224693180014583
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.45517500559799373,
          "train_loss": 0.46361541748046875,
          "validation_loss": 0.030576419085264206,
          "train_metric": 0.046693652868270874,
          "validation_metric": 0.0030795452184975147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7389812809997238
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.44738132506608963,
          "train_loss": 0.47017228603363037,
          "validation_loss": 0.07133234292268753,
          "train_metric": 0.04735402762889862,
          "validation_metric": 0.007184331770986319,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7559938940012216
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.4708770215511322,
          "train_loss": 0.449205219745636,
          "validation_loss": 0.03560194373130798,
          "train_metric": 0.0452423021197319,
          "validation_metric": 0.003585697617381811,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.772671786000501
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.4564231261610985,
          "train_loss": 0.4491090178489685,
          "validation_loss": 0.05670673027634621,
          "train_metric": 0.045232612639665604,
          "validation_metric": 0.005711293779313564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.788835655001094
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.45315060392022133,
          "train_loss": 0.4462970197200775,
          "validation_loss": 0.016596395522356033,
          "train_metric": 0.044949404895305634,
          "validation_metric": 0.001671528210863471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8057672169998114
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.45211621664930135,
          "train_loss": 0.44399768114089966,
          "validation_loss": 0.03378140553832054,
          "train_metric": 0.04471781849861145,
          "validation_metric": 0.0034023390617221594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8221711780006444
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.45101910550147295,
          "train_loss": 0.45813077688217163,
          "validation_loss": 0.029888035729527473,
          "train_metric": 0.046141255646944046,
          "validation_metric": 0.0030102133750915527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8386054930015234
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.449222520692274,
          "train_loss": 0.4538984000682831,
          "validation_loss": 0.03632649779319763,
          "train_metric": 0.04571498557925224,
          "validation_metric": 0.0036586716305464506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8548333079997974
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.4520052643492818,
          "train_loss": 0.4441951513290405,
          "validation_loss": 0.04223034158349037,
          "train_metric": 0.04473770409822464,
          "validation_metric": 0.004253285471349955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8710648779997427
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.4534426871687174,
          "train_loss": 0.465620756149292,
          "validation_loss": 0.021486107259988785,
          "train_metric": 0.046895623207092285,
          "validation_metric": 0.00216400227509439,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8875218800003495
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.45499282516539097,
          "train_loss": 0.44229835271835327,
          "validation_loss": 0.034904301166534424,
          "train_metric": 0.04454667121171951,
          "validation_metric": 0.0035154330544173717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9036480020004092
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.45033813640475273,
          "train_loss": 0.4463084042072296,
          "validation_loss": 0.04382502660155296,
          "train_metric": 0.044950541108846664,
          "validation_metric": 0.00441389624029398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9198344570013433
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.4505853448063135,
          "train_loss": 0.44914013147354126,
          "validation_loss": 0.061274126172065735,
          "train_metric": 0.04523574560880661,
          "validation_metric": 0.006171305663883686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9359870990010677
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.4508105041459203,
          "train_loss": 0.4446415305137634,
          "validation_loss": 0.037464823573827744,
          "train_metric": 0.04478266462683678,
          "validation_metric": 0.0037733197677880526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9520447209997656
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.4482052708044648,
          "train_loss": 0.4467461407184601,
          "validation_loss": 0.027745310217142105,
          "train_metric": 0.04499463737010956,
          "validation_metric": 0.0027944063767790794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9681106180014467
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.44637309294193983,
          "train_loss": 0.4515087902545929,
          "validation_loss": 0.05134117603302002,
          "train_metric": 0.045474305748939514,
          "validation_metric": 0.005170895252376795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9842067820009106
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.4566079108044505,
          "train_loss": 0.44481468200683594,
          "validation_loss": 0.04968519136309624,
          "train_metric": 0.044800106436014175,
          "validation_metric": 0.005004110746085644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0006056210004317
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.459231685847044,
          "train_loss": 0.4559769630432129,
          "validation_loss": 0.012408791109919548,
          "train_metric": 0.04592432454228401,
          "validation_metric": 0.0012497679563239217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0176050950012723
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.45417256955988705,
          "train_loss": 0.4431121349334717,
          "validation_loss": 0.03325655311346054,
          "train_metric": 0.04462863504886627,
          "validation_metric": 0.003349478356540203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.03327670300132
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.44751263689249754,
          "train_loss": 0.4451705515384674,
          "validation_loss": 0.035051118582487106,
          "train_metric": 0.04483594745397568,
          "validation_metric": 0.0035302198957651854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0488931870004308
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.4581496939063072,
          "train_loss": 0.4631498157978058,
          "validation_loss": 0.019893670454621315,
          "train_metric": 0.04664674773812294,
          "validation_metric": 0.002003617584705353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.064687137000874
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.4522576795425266,
          "train_loss": 0.4517659842967987,
          "validation_loss": 0.052553482353687286,
          "train_metric": 0.04550021514296532,
          "validation_metric": 0.005292993970215321,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.080918460000248
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.45784165407530963,
          "train_loss": 0.4505918025970459,
          "validation_loss": 0.050756294280290604,
          "train_metric": 0.04538195952773094,
          "validation_metric": 0.005111987702548504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0972745700000814
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.45544068654999137,
          "train_loss": 0.4459974467754364,
          "validation_loss": 0.03555840253829956,
          "train_metric": 0.044919226318597794,
          "validation_metric": 0.003581312019377947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.113230876000671
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.4511456764303148,
          "train_loss": 0.4414437413215637,
          "validation_loss": 0.03415510803461075,
          "train_metric": 0.04446059837937355,
          "validation_metric": 0.003439977066591382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.129161483000644
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.448213457595557,
          "train_loss": 0.45424073934555054,
          "validation_loss": 0.024614272639155388,
          "train_metric": 0.04574945941567421,
          "validation_metric": 0.00247905938886106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1452128640012234
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.45369876315817237,
          "train_loss": 0.4431411623954773,
          "validation_loss": 0.03874247521162033,
          "train_metric": 0.04463155195116997,
          "validation_metric": 0.0039020003750920296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.161018603001139
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.45050271903164685,
          "train_loss": 0.4441874325275421,
          "validation_loss": 0.04187585785984993,
          "train_metric": 0.04473692923784256,
          "validation_metric": 0.00421758322045207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.176704299999983
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.44760807789862156,
          "train_loss": 0.44265982508659363,
          "validation_loss": 0.019201548770070076,
          "train_metric": 0.04458307847380638,
          "validation_metric": 0.0019339099526405334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1923778820000734
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.4476700499653816,
          "train_loss": 0.4444507956504822,
          "validation_loss": 0.04601467028260231,
          "train_metric": 0.044763460755348206,
          "validation_metric": 0.004634428769350052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2992255940007453
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.4494558656588197,
          "train_loss": 0.4450490176677704,
          "validation_loss": 0.029580997303128242,
          "train_metric": 0.04482370615005493,
          "validation_metric": 0.0029792897403240204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3158323950010526
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.44795296154916286,
          "train_loss": 0.4451199173927307,
          "validation_loss": 0.03488420322537422,
          "train_metric": 0.04483084753155708,
          "validation_metric": 0.0035134092904627323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.331905705999816
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.4522477751597762,
          "train_loss": 0.4445529878139496,
          "validation_loss": 0.030739376321434975,
          "train_metric": 0.04477375000715256,
          "validation_metric": 0.003095957450568676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.347735373001342
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.44804845843464136,
          "train_loss": 0.45138126611709595,
          "validation_loss": 0.04883338883519173,
          "train_metric": 0.045461468398571014,
          "validation_metric": 0.00491832010447979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.36343695800133
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.44909702986478806,
          "train_loss": 0.44540271162986755,
          "validation_loss": 0.01856388710439205,
          "train_metric": 0.044859327375888824,
          "validation_metric": 0.0018696868792176247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3793126239997946
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.444470806280151,
          "train_loss": 0.44155609607696533,
          "validation_loss": 0.038002077490091324,
          "train_metric": 0.04447191581130028,
          "validation_metric": 0.0038274298422038555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.395015133000925
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.4471111479215324,
          "train_loss": 0.4422171413898468,
          "validation_loss": 0.033152345567941666,
          "train_metric": 0.04453849792480469,
          "validation_metric": 0.0033389825839549303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4111383660001593
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.4532542796805501,
          "train_loss": 0.45232105255126953,
          "validation_loss": 0.028597718104720116,
          "train_metric": 0.045556116849184036,
          "validation_metric": 0.0028802575543522835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4270683260001533
        }
      ],
      "best": {
        "epoch": 98,
        "updates": 1568
      },
      "best_validation_metric": 0.0012497679563239217,
      "config": {
        "label": "MSE + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": true,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.03535205731388092
    },
    "14985df6b90f_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.18389458605088294,
          "train_loss": 0.06916050612926483,
          "validation_loss": 0.06209268793463707,
          "train_metric": 0.006965593434870243,
          "validation_metric": 0.0062537482008337975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02333306099899346
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.045738674933090806,
          "train_loss": 0.03489449992775917,
          "validation_loss": 0.032735906541347504,
          "train_metric": 0.0035144463181495667,
          "validation_metric": 0.003297040704637766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04611985600058688
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.02863602398429066,
          "train_loss": 0.02373126894235611,
          "validation_loss": 0.017781876027584076,
          "train_metric": 0.0023901264648884535,
          "validation_metric": 0.0017909255111590028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0690487989995745
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.03047496487852186,
          "train_loss": 0.018814587965607643,
          "validation_loss": 0.01837657205760479,
          "train_metric": 0.001894936547614634,
          "validation_metric": 0.001850820961408317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09156705900022644
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.01888181286631152,
          "train_loss": 0.01885739341378212,
          "validation_loss": 0.016206376254558563,
          "train_metric": 0.001899247639812529,
          "validation_metric": 0.0016322467708960176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11321936499916774
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.01619607725297101,
          "train_loss": 0.011927196756005287,
          "validation_loss": 0.011007381603121758,
          "train_metric": 0.0012012633960694075,
          "validation_metric": 0.001108623226173222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13473909300046216
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.012518655508756638,
          "train_loss": 0.010872814804315567,
          "validation_loss": 0.01140688918530941,
          "train_metric": 0.0010950700379908085,
          "validation_metric": 0.0011488600866869092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15501596499962034
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.009969741659006104,
          "train_loss": 0.009192890487611294,
          "validation_loss": 0.009333253838121891,
          "train_metric": 0.000925874279346317,
          "validation_metric": 0.00094001111574471,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1759318570002506
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.010474069189513102,
          "train_loss": 0.00656439084559679,
          "validation_loss": 0.00658778939396143,
          "train_metric": 0.0006611414137296379,
          "validation_metric": 0.0006634980672970414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19761058800031606
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.013046462088823318,
          "train_loss": 0.023848779499530792,
          "validation_loss": 0.01683362014591694,
          "train_metric": 0.0024019619449973106,
          "validation_metric": 0.0016954205930233002,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2176567949991295
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.0150880494620651,
          "train_loss": 0.01769361086189747,
          "validation_loss": 0.013639772310853004,
          "train_metric": 0.0017820355715230107,
          "validation_metric": 0.0013737480621784925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23796098300044832
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.013406007667072117,
          "train_loss": 0.010042965412139893,
          "validation_loss": 0.01069402415305376,
          "train_metric": 0.0010114905890077353,
          "validation_metric": 0.0010770629160106182,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25779569599944807
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.007156405947171152,
          "train_loss": 0.0063883112743496895,
          "validation_loss": 0.0051514930091798306,
          "train_metric": 0.0006434072856791317,
          "validation_metric": 0.0005188394570723176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3813991469996836
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.006110915433964692,
          "train_loss": 0.005998062901198864,
          "validation_loss": 0.008657868020236492,
          "train_metric": 0.0006041027954779565,
          "validation_metric": 0.0008719888282939792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40239525499964657
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.0062475081358570606,
          "train_loss": 0.01197153702378273,
          "validation_loss": 0.007577677723020315,
          "train_metric": 0.0012057292042300105,
          "validation_metric": 0.0007631958578713238,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42303843799891183
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.0113911721855402,
          "train_loss": 0.008418227545917034,
          "validation_loss": 0.005429287441074848,
          "train_metric": 0.0008478531381115317,
          "validation_metric": 0.0005468179006129503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4430628730005992
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.006947965404833667,
          "train_loss": 0.006220518611371517,
          "validation_loss": 0.004411776550114155,
          "train_metric": 0.0006265078554861248,
          "validation_metric": 0.00044433801667764783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4647173230005137
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.015920912410365418,
          "train_loss": 0.005532862152904272,
          "validation_loss": 0.0060539236292243,
          "train_metric": 0.0005572495865635574,
          "validation_metric": 0.0006097290315665305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.48505106099946715
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.0066804879024857655,
          "train_loss": 0.005352771375328302,
          "validation_loss": 0.006394522730261087,
          "train_metric": 0.0005391114973463118,
          "validation_metric": 0.0006440328434109688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5056490900005883
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.006032283257809468,
          "train_loss": 0.007059706840664148,
          "validation_loss": 0.006349417380988598,
          "train_metric": 0.0007110278238542378,
          "validation_metric": 0.000639490142930299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5259560960003
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.011215290942345746,
          "train_loss": 0.010435744188725948,
          "validation_loss": 0.007207291200757027,
          "train_metric": 0.0010510500287637115,
          "validation_metric": 0.0007258919067680836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5463446659996407
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.006948341120732948,
          "train_loss": 0.004147992469370365,
          "validation_loss": 0.004910673946142197,
          "train_metric": 0.0004177706432528794,
          "validation_metric": 0.0004945851396769285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5668650899988279
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.005095057058497332,
          "train_loss": 0.005464919842779636,
          "validation_loss": 0.00746883312240243,
          "train_metric": 0.000550406810361892,
          "validation_metric": 0.0007522335508838296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5871275240006071
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.005976300351903774,
          "train_loss": 0.006015682592988014,
          "validation_loss": 0.004650956951081753,
          "train_metric": 0.0006058774888515472,
          "validation_metric": 0.00046842731535434723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.60731837899948
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.008279251887870487,
          "train_loss": 0.004652548581361771,
          "validation_loss": 0.004610558971762657,
          "train_metric": 0.0004685877065639943,
          "validation_metric": 0.00046435859985649586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6273249709993252
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.005110172278364189,
          "train_loss": 0.004339386243373156,
          "validation_loss": 0.003442443208768964,
          "train_metric": 0.0004370470705907792,
          "validation_metric": 0.0003467102360446006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6502715539991186
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.004839025321416557,
          "train_loss": 0.004047211725264788,
          "validation_loss": 0.0022406503558158875,
          "train_metric": 0.000407620333135128,
          "validation_metric": 0.0002256701118312776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6724845139997342
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.006638483784627169,
          "train_loss": 0.0031345903407782316,
          "validation_loss": 0.0033052018843591213,
          "train_metric": 0.0003157044993713498,
          "validation_metric": 0.00033288789563812315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6941820840002038
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.004400948935654014,
          "train_loss": 0.007335529197007418,
          "validation_loss": 0.007428395561873913,
          "train_metric": 0.0007388076628558338,
          "validation_metric": 0.0007481608190573752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7160780109988991
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.007314472968573682,
          "train_loss": 0.01119297370314598,
          "validation_loss": 0.009809521026909351,
          "train_metric": 0.001127315335907042,
          "validation_metric": 0.000987979001365602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7379637069989258
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.011946768732741475,
          "train_loss": 0.015263335779309273,
          "validation_loss": 0.009868407621979713,
          "train_metric": 0.00153726723510772,
          "validation_metric": 0.0009939097799360752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7596423659997527
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.01918425314943306,
          "train_loss": 0.008226007223129272,
          "validation_loss": 0.006636414676904678,
          "train_metric": 0.0008284933865070343,
          "validation_metric": 0.0006683953106403351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7806624029999512
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.007922696531750262,
          "train_loss": 0.00950158853083849,
          "validation_loss": 0.00873914361000061,
          "train_metric": 0.0009569652029313147,
          "validation_metric": 0.0008801745716482401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8016491500002303
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.0060403976967791095,
          "train_loss": 0.004429431166499853,
          "validation_loss": 0.0031370145734399557,
          "train_metric": 0.00044611605699174106,
          "validation_metric": 0.0003159486223012209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8236236689990619
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.007136315063689835,
          "train_loss": 0.00483589805662632,
          "validation_loss": 0.004416401963680983,
          "train_metric": 0.00048705397057347,
          "validation_metric": 0.00044480382348410785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8463680040003965
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.006166164930618834,
          "train_loss": 0.008886346593499184,
          "validation_loss": 0.006666672416031361,
          "train_metric": 0.0008950002957135439,
          "validation_metric": 0.0006714428309351206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8678952390000632
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.013328372235264396,
          "train_loss": 0.006340267136693001,
          "validation_loss": 0.005754882004112005,
          "train_metric": 0.0006385684828273952,
          "validation_metric": 0.0005796106997877359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8894811720001599
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.012484577498980798,
          "train_loss": 0.010533424094319344,
          "validation_loss": 0.008602084591984749,
          "train_metric": 0.0010608879383653402,
          "validation_metric": 0.0008663704502396286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9107658270004322
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.005340583942597732,
          "train_loss": 0.004230243153870106,
          "validation_loss": 0.003673012601211667,
          "train_metric": 0.0004260546702425927,
          "validation_metric": 0.0003699323860928416,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9324729540003318
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.003958030189096462,
          "train_loss": 0.0034799654968082905,
          "validation_loss": 0.0027653248980641365,
          "train_metric": 0.0003504893684294075,
          "validation_metric": 0.00027851335471495986,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9585030950001965
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.003779596743697766,
          "train_loss": 0.0027438648976385593,
          "validation_loss": 0.0030321215745061636,
          "train_metric": 0.0002763520460575819,
          "validation_metric": 0.0003053841646760702,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 0.9906086979990505
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.003792784315010067,
          "train_loss": 0.002959120785817504,
          "validation_loss": 0.002368425950407982,
          "train_metric": 0.00029803180950693786,
          "validation_metric": 0.00023853919992689043,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0127113220005413
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.003266840103606228,
          "train_loss": 0.0026186981704086065,
          "validation_loss": 0.0020507124718278646,
          "train_metric": 0.0002637456636875868,
          "validation_metric": 0.00020654025138355792,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0344747959989036
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.0031263779928849544,
          "train_loss": 0.0018942949827760458,
          "validation_loss": 0.0015515806153416634,
          "train_metric": 0.00019078644982073456,
          "validation_metric": 0.00015626955428160727,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0561790720003046
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.002282231584104011,
          "train_loss": 0.002708063693717122,
          "validation_loss": 0.001892526401206851,
          "train_metric": 0.0002727462269831449,
          "validation_metric": 0.00019060831982642412,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.0771126719992026
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.003364461757882964,
          "train_loss": 0.004969216883182526,
          "validation_loss": 0.004811704624444246,
          "train_metric": 0.0005004813428968191,
          "validation_metric": 0.0004846172232646495,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.097665679999409
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0042546289405436255,
          "train_loss": 0.0032602050341665745,
          "validation_loss": 0.002327308990061283,
          "train_metric": 0.00032835587626323104,
          "validation_metric": 0.0002343980740988627,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.118489110000155
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0024624920843052678,
          "train_loss": 0.0020010184962302446,
          "validation_loss": 0.002191733568906784,
          "train_metric": 0.00020153526565991342,
          "validation_metric": 0.00022074338630773127,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.1388082539997413
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.0031925522671372164,
          "train_loss": 0.0037992135621607304,
          "validation_loss": 0.003762789536267519,
          "train_metric": 0.00038264290196821094,
          "validation_metric": 0.00037897436413913965,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.1587180749993422
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.004498012462136103,
          "train_loss": 0.0018756872741505504,
          "validation_loss": 0.0010174153139814734,
          "train_metric": 0.00018891235231421888,
          "validation_metric": 0.00010247033787891269,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.1797242869997717
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.0016706399583199527,
          "train_loss": 0.0014610332436859608,
          "validation_loss": 0.0009999549947679043,
          "train_metric": 0.00014714991266373545,
          "validation_metric": 0.00010071179713122547,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2017020419989421
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0021081952581880614,
          "train_loss": 0.0025933822616934776,
          "validation_loss": 0.001571556436829269,
          "train_metric": 0.0002611959644127637,
          "validation_metric": 0.0001582814147695899,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2223558780005988
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.0020395970896061044,
          "train_loss": 0.0013225739821791649,
          "validation_loss": 0.0012418730184435844,
          "train_metric": 0.00013320479774847627,
          "validation_metric": 0.0001250768982572481,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2430251059995499
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0028531694151752163,
          "train_loss": 0.004799867980182171,
          "validation_loss": 0.005346137098968029,
          "train_metric": 0.0004834251303691417,
          "validation_metric": 0.0005384434480220079,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.2639132419990347
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.004429804837855045,
          "train_loss": 0.004045144654810429,
          "validation_loss": 0.004567306488752365,
          "train_metric": 0.00040741218253970146,
          "validation_metric": 0.0004600024258252233,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.284918714000014
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0031316953027271666,
          "train_loss": 0.0032288390211760998,
          "validation_loss": 0.0019552449230104685,
          "train_metric": 0.00032519683009013534,
          "validation_metric": 0.00019692510250024498,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.30562085100064
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.005380759339459473,
          "train_loss": 0.006660917773842812,
          "validation_loss": 0.0060523864813148975,
          "train_metric": 0.0006708631990477443,
          "validation_metric": 0.0006095741991885006,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3262663099994825
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.003408056953048799,
          "train_loss": 0.002101266523823142,
          "validation_loss": 0.0021351310424506664,
          "train_metric": 0.00021163185010664165,
          "validation_metric": 0.00021504257165361196,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3468740079988493
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.0018077379900205415,
          "train_loss": 0.001777643570676446,
          "validation_loss": 0.0014206002233549953,
          "train_metric": 0.00017903775733429939,
          "validation_metric": 0.00014307766105048358,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.3673931849989458
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.001430878310202388,
          "train_loss": 0.0011914899805560708,
          "validation_loss": 0.0008654165430925786,
          "train_metric": 0.00012000252172583714,
          "validation_metric": 8.716157753951848e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.389192753998941
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.001971792684344109,
          "train_loss": 0.0015797116793692112,
          "validation_loss": 0.0012506500352174044,
          "train_metric": 0.00015910275396890938,
          "validation_metric": 0.0001259608834516257,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4098513409990119
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.005401995935244486,
          "train_loss": 0.003050468862056732,
          "validation_loss": 0.0030622242484241724,
          "train_metric": 0.00030723202507942915,
          "validation_metric": 0.00030841605621390045,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4302440699993895
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.003195493158273166,
          "train_loss": 0.0018956264248117805,
          "validation_loss": 0.0013271392090246081,
          "train_metric": 0.00019092054571956396,
          "validation_metric": 0.00013366462371777743,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4502923469990492
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.001677237029070966,
          "train_loss": 0.0017810814315453172,
          "validation_loss": 0.0013566982233896852,
          "train_metric": 0.00017938400560524315,
          "validation_metric": 0.00013664166908711195,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4704998839988548
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.003182141444995068,
          "train_loss": 0.005172677803784609,
          "validation_loss": 0.005077451933175325,
          "train_metric": 0.0005209732335060835,
          "validation_metric": 0.0005113823572173715,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4908907640001416
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.003169672407238977,
          "train_loss": 0.0015204192604869604,
          "validation_loss": 0.0010483849328011274,
          "train_metric": 0.00015313106996472925,
          "validation_metric": 0.00010558950452832505,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5111760389991105
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0020033447108289693,
          "train_loss": 0.0025511691346764565,
          "validation_loss": 0.0019035914447158575,
          "train_metric": 0.00025694441865198314,
          "validation_metric": 0.00019172276370227337,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5321596670000872
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0016619744819763582,
          "train_loss": 0.0015163269126787782,
          "validation_loss": 0.0011533438228070736,
          "train_metric": 0.00015271891606971622,
          "validation_metric": 0.00011616056872298941,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5540746879996732
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0015179603487922577,
          "train_loss": 0.0010273667285218835,
          "validation_loss": 0.0009914685506373644,
          "train_metric": 0.00010347260831622407,
          "validation_metric": 9.985708311432973e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5750143610002851
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0015253068613674259,
          "train_loss": 0.0022541654761880636,
          "validation_loss": 0.0014503425918519497,
          "train_metric": 0.00022703129798173904,
          "validation_metric": 0.00014607320190407336,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.595644335000543
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0014404388530238066,
          "train_loss": 0.0015009132912382483,
          "validation_loss": 0.0010611419565975666,
          "train_metric": 0.00015116647409740835,
          "validation_metric": 0.00010687434405554086,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6162368679997599
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0014964083711674903,
          "train_loss": 0.000843506189994514,
          "validation_loss": 0.0005511737545020878,
          "train_metric": 8.495486690662801e-05,
          "validation_metric": 5.5512198741780594e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.6383390579994739
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0020285771397539065,
          "train_loss": 0.0011159328278154135,
          "validation_loss": 0.0009043682948686182,
          "train_metric": 0.00011239266314078122,
          "validation_metric": 9.108466474572197e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.6615890649991343
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.00159594154865772,
          "train_loss": 0.002564321970567107,
          "validation_loss": 0.002573132049292326,
          "train_metric": 0.0002582691377028823,
          "validation_metric": 0.00025915642618201673,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.682171758999175
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.001989463178688311,
          "train_loss": 0.0012428787304088473,
          "validation_loss": 0.0012404659064486623,
          "train_metric": 0.000125178208691068,
          "validation_metric": 0.00012493517715483904,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.702571858999363
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.0011931037697650027,
          "train_loss": 0.0009971588151529431,
          "validation_loss": 0.0006950655952095985,
          "train_metric": 0.0001004301811917685,
          "validation_metric": 7.000446203164756e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.7226488950000203
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.0009488010600762209,
          "train_loss": 0.0008507153252139688,
          "validation_loss": 0.0006065382622182369,
          "train_metric": 8.568093471694738e-05,
          "validation_metric": 6.108831439632922e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.7432025449998036
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.0008343484760189313,
          "train_loss": 0.0008572710212320089,
          "validation_loss": 0.0006751189939677715,
          "train_metric": 8.634119149064645e-05,
          "validation_metric": 6.799550465075299e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.7633963959997345
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0009182748162857024,
          "train_loss": 0.0009832093492150307,
          "validation_loss": 0.0007327371276915073,
          "train_metric": 9.902524470817298e-05,
          "validation_metric": 7.37986047170125e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.7838600750001206
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.001212134913657792,
          "train_loss": 0.0017133249202743173,
          "validation_loss": 0.0011205439222976565,
          "train_metric": 0.00017255979764740914,
          "validation_metric": 0.00011285710206720978,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.8043161259993212
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.00227742695096822,
          "train_loss": 0.0024035379756242037,
          "validation_loss": 0.002715857233852148,
          "train_metric": 0.000242075533606112,
          "validation_metric": 0.0002735312154982239,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.82447853299891
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.004072214724146761,
          "train_loss": 0.0035324934870004654,
          "validation_loss": 0.002653245348483324,
          "train_metric": 0.0003557798045221716,
          "validation_metric": 0.00026722511393018067,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.8447327969988692
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.002495613489372772,
          "train_loss": 0.0007845122599974275,
          "validation_loss": 0.0007261664140969515,
          "train_metric": 7.901319622760639e-05,
          "validation_metric": 7.313682726817206e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9609326850004436
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0008701173646841198,
          "train_loss": 0.0006956751458346844,
          "validation_loss": 0.00046837294939905405,
          "train_metric": 7.006584928603843e-05,
          "validation_metric": 4.717280171462335e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 1.983759779999673
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0007068669920045068,
          "train_loss": 0.0006698669749312103,
          "validation_loss": 0.0006065756315365434,
          "train_metric": 6.746654253220186e-05,
          "validation_metric": 6.109207606641576e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0049650549990474
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.0006383104282576824,
          "train_loss": 0.0005993851809762418,
          "validation_loss": 0.0004608859308063984,
          "train_metric": 6.036787817720324e-05,
          "validation_metric": 4.641874329536222e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0263686109992705
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0006607069499295903,
          "train_loss": 0.0006324349087662995,
          "validation_loss": 0.00045339783537201583,
          "train_metric": 6.369652692228556e-05,
          "validation_metric": 4.566457209875807e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.048214205000477
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.000619731581537053,
          "train_loss": 0.0005477811791934073,
          "validation_loss": 0.0004459129413589835,
          "train_metric": 5.517050885828212e-05,
          "validation_metric": 4.491072104428895e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0707554429991433
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0006693748637189856,
          "train_loss": 0.0005220915772952139,
          "validation_loss": 0.00042084953747689724,
          "train_metric": 5.258315650280565e-05,
          "validation_metric": 4.238642213749699e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.0933506569999736
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0006681859322270611,
          "train_loss": 0.0011663727927953005,
          "validation_loss": 0.0008670196402817965,
          "train_metric": 0.000117472794954665,
          "validation_metric": 8.732305286685005e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.1144051709998166
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0008925436559366062,
          "train_loss": 0.0008190940716303885,
          "validation_loss": 0.0007454908918589354,
          "train_metric": 8.249613892985508e-05,
          "validation_metric": 7.508311682613567e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.134473120999246
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0008474731203023111,
          "train_loss": 0.0005371115985326469,
          "validation_loss": 0.0004170157480984926,
          "train_metric": 5.4095915402285755e-05,
          "validation_metric": 4.2000290704891086e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.1559842080005183
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0006023637361067813,
          "train_loss": 0.0005184495821595192,
          "validation_loss": 0.00039884104626253247,
          "train_metric": 5.221634637564421e-05,
          "validation_metric": 4.0169812564272434e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.1777762139990955
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0005925915520492708,
          "train_loss": 0.0005057367961853743,
          "validation_loss": 0.00040048666414804757,
          "train_metric": 5.093595609650947e-05,
          "validation_metric": 4.0335555240744725e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.199022196999067
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0005337636184776784,
          "train_loss": 0.00048719157348386943,
          "validation_loss": 0.00035465986002236605,
          "train_metric": 4.906815229333006e-05,
          "validation_metric": 3.5720044252229854e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.2209026999989874
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0005283006994432071,
          "train_loss": 0.0005378840724006295,
          "validation_loss": 0.00046729634050279856,
          "train_metric": 5.417371357907541e-05,
          "validation_metric": 4.70643826702144e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.241242608999528
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0005313867732184008,
          "train_loss": 0.0004453692235983908,
          "validation_loss": 0.00034873277763836086,
          "train_metric": 4.48559585493058e-05,
          "validation_metric": 3.5123088309774175e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.262701924999419
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0005509997390618082,
          "train_loss": 0.0004511236911639571,
          "validation_loss": 0.0003458083374425769,
          "train_metric": 4.543552495306358e-05,
          "validation_metric": 3.4828550269594416e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.2847871569993004
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0006912787721375935,
          "train_loss": 0.0007766124908812344,
          "validation_loss": 0.000625491316895932,
          "train_metric": 7.82175557105802e-05,
          "validation_metric": 6.299718370428309e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.305155525998998
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0005400173922680551,
          "train_loss": 0.0006971806287765503,
          "validation_loss": 0.0004150439635850489,
          "train_metric": 7.021748024271801e-05,
          "validation_metric": 4.180170799372718e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.325503021998884
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0006987284850765718,
          "train_loss": 0.000843168469145894,
          "validation_loss": 0.0008488795720040798,
          "train_metric": 8.49208518047817e-05,
          "validation_metric": 8.549603808205575e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.346340964999399
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0007000443765718956,
          "train_loss": 0.0005398172652348876,
          "validation_loss": 0.00044017963227815926,
          "train_metric": 5.436841456685215e-05,
          "validation_metric": 4.433327922015451e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.3665916840000136
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.000468253664621443,
          "train_loss": 0.0008504493744112551,
          "validation_loss": 0.000756210065446794,
          "train_metric": 8.565415191696957e-05,
          "validation_metric": 7.616270886501297e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.3874139380004635
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0008260653530669515,
          "train_loss": 0.0004479802446439862,
          "validation_loss": 0.0003722138935700059,
          "train_metric": 4.5118937123334035e-05,
          "validation_metric": 3.748802555492148e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.4082763559999876
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0005999941340633086,
          "train_loss": 0.0004657612007576972,
          "validation_loss": 0.00032526947325095534,
          "train_metric": 4.6909764932934195e-05,
          "validation_metric": 3.275994458817877e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.4290182049990108
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.00045484592374123167,
          "train_loss": 0.0005409568548202515,
          "validation_loss": 0.0004488911072257906,
          "train_metric": 5.448318916023709e-05,
          "validation_metric": 4.521067239693366e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.449990192999394
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0004908250612061238,
          "train_loss": 0.0004001584311481565,
          "validation_loss": 0.0002980621939059347,
          "train_metric": 4.0302493289345875e-05,
          "validation_metric": 3.0019733458175324e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.4718969490004383
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0005983905011817114,
          "train_loss": 0.0005870725144632161,
          "validation_loss": 0.0004464038647711277,
          "train_metric": 5.912779306527227e-05,
          "validation_metric": 4.496015753829852e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.49221235999903
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0004827441807719879,
          "train_loss": 0.0003902669413946569,
          "validation_loss": 0.0002997760602738708,
          "train_metric": 3.930625462089665e-05,
          "validation_metric": 3.0192346457624808e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.5128774999993766
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0005665363687512581,
          "train_loss": 0.0011384909739717841,
          "validation_loss": 0.0008203002507798374,
          "train_metric": 0.00011466463911347091,
          "validation_metric": 8.261763287009671e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.533492266999019
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.0006922082320670597,
          "train_loss": 0.0006019834545440972,
          "validation_loss": 0.0005475224461406469,
          "train_metric": 6.0629568906733766e-05,
          "validation_metric": 5.514446093002334e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.553384143000585
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.0005106444623379502,
          "train_loss": 0.0003945694479625672,
          "validation_loss": 0.000296755664749071,
          "train_metric": 3.9739588828524575e-05,
          "validation_metric": 2.988814412674401e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.574601905998861
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.00048763505856186384,
          "train_loss": 0.001366328913718462,
          "validation_loss": 0.0011787258554250002,
          "train_metric": 0.00013761164154857397,
          "validation_metric": 0.00011871694005094469,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.5947206519995234
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0009404624097442138,
          "train_loss": 0.0003464885812718421,
          "validation_loss": 0.0002557314292062074,
          "train_metric": 3.4897064324468374e-05,
          "validation_metric": 2.575633152446244e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.615813757000069
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0004283728640075424,
          "train_loss": 0.00041304799378849566,
          "validation_loss": 0.00029907116550020874,
          "train_metric": 4.1600676922826096e-05,
          "validation_metric": 3.012135493918322e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.6364472009991005
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0003918726147276175,
          "train_loss": 0.000370957946870476,
          "validation_loss": 0.0003360460978001356,
          "train_metric": 3.736152939382009e-05,
          "validation_metric": 3.3845331927295774e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.656615731999409
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.00037772477480757516,
          "train_loss": 0.0003448660718277097,
          "validation_loss": 0.000251723948167637,
          "train_metric": 3.473364267847501e-05,
          "validation_metric": 2.535271596570965e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.6781418489990756
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.00037904694727330934,
          "train_loss": 0.0003402280271984637,
          "validation_loss": 0.0002685183717403561,
          "train_metric": 3.4266522561665624e-05,
          "validation_metric": 2.7044186936109327e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.699137923998933
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.00041750468881218694,
          "train_loss": 0.0003597447066567838,
          "validation_loss": 0.00025031858240254223,
          "train_metric": 3.6232173442840576e-05,
          "validation_metric": 2.5211171305272728e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.720437182999376
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0006041453025318333,
          "train_loss": 0.00032441000803373754,
          "validation_loss": 0.0002659092715475708,
          "train_metric": 3.267338979640044e-05,
          "validation_metric": 2.6781406631926075e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.741084356999636
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0003904484010490705,
          "train_loss": 0.0003045335179194808,
          "validation_loss": 0.0002487014280632138,
          "train_metric": 3.067150100832805e-05,
          "validation_metric": 2.5048298994079232e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7626082329989003
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0003688784308906179,
          "train_loss": 0.0006275029736571014,
          "validation_loss": 0.00044795399298891425,
          "train_metric": 6.319979729596525e-05,
          "validation_metric": 4.5116285036783665e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7833250160001626
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0007111332306521945,
          "train_loss": 0.0021258313208818436,
          "validation_loss": 0.0018453336087986827,
          "train_metric": 0.00021410595218185335,
          "validation_metric": 0.00018585524230729789,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8041362399999343
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0010522935954213608,
          "train_loss": 0.000982508179731667,
          "validation_loss": 0.0008631922537460923,
          "train_metric": 9.89546169876121e-05,
          "validation_metric": 8.693755808053538e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8243015440002637
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.000768384636103292,
          "train_loss": 0.00032966729486361146,
          "validation_loss": 0.0002778651542030275,
          "train_metric": 3.320288669783622e-05,
          "validation_metric": 2.7985559427179396e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.843787986999814
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0003478840171737829,
          "train_loss": 0.0003117503074463457,
          "validation_loss": 0.00024146052601281554,
          "train_metric": 3.139835098409094e-05,
          "validation_metric": 2.431902066746261e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8651128849996894
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.00036173600619804347,
          "train_loss": 0.000820610614027828,
          "validation_loss": 0.0006031462689861655,
          "train_metric": 8.264889765996486e-05,
          "validation_metric": 6.0746693634428084e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8860449160001735
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0004490610517677851,
          "train_loss": 0.00029991596238687634,
          "validation_loss": 0.00021668038971256465,
          "train_metric": 3.020643998752348e-05,
          "validation_metric": 2.182325624744408e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9079377540001587
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.00030620830602856586,
          "train_loss": 0.0006654991302639246,
          "validation_loss": 0.00042452028719708323,
          "train_metric": 6.702663813484833e-05,
          "validation_metric": 4.2756124457810074e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.929137957000421
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0005116830016049789,
          "train_loss": 0.0003119233588222414,
          "validation_loss": 0.00022276115487329662,
          "train_metric": 3.141577690257691e-05,
          "validation_metric": 2.2435688151745126e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.950059238999529
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0003531554921210045,
          "train_loss": 0.00027423532446846366,
          "validation_loss": 0.00020882398530375212,
          "train_metric": 2.7619977117865346e-05,
          "validation_metric": 2.1031988580944017e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.971476351000092
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.00035998752719024196,
          "train_loss": 0.0006240218644961715,
          "validation_loss": 0.0004997733631171286,
          "train_metric": 6.284919800236821e-05,
          "validation_metric": 5.0335343985352665e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.992064329999266
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0003371636921656318,
          "train_loss": 0.0003034695109818131,
          "validation_loss": 0.0002213252300862223,
          "train_metric": 3.056434070458636e-05,
          "validation_metric": 2.229106758022681e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.01236129300014
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0004825937148780213,
          "train_loss": 0.0006174672162160277,
          "validation_loss": 0.0005699259927496314,
          "train_metric": 6.218903581611812e-05,
          "validation_metric": 5.7400859077461064e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.0327421429992683
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.00038741136359021766,
          "train_loss": 0.0002657816803548485,
          "validation_loss": 0.00019705368322320282,
          "train_metric": 2.6768560928758234e-05,
          "validation_metric": 1.9846524082822725e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.05440395199912
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0003964353427363676,
          "train_loss": 0.0008399318321608007,
          "validation_loss": 0.0006169424159452319,
          "train_metric": 8.459485979983583e-05,
          "validation_metric": 6.213619053596631e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.0760804599995026
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0006074074481148273,
          "train_loss": 0.0002821312809828669,
          "validation_loss": 0.0001935452310135588,
          "train_metric": 2.8415230190148577e-05,
          "validation_metric": 1.9493167201289907e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.0972011989997554
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0005031224463891704,
          "train_loss": 0.00042860067333094776,
          "validation_loss": 0.0003237599739804864,
          "train_metric": 4.316708873375319e-05,
          "validation_metric": 3.260792072978802e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.118198262000078
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0003811296173807932,
          "train_loss": 0.00032635184470564127,
          "validation_loss": 0.0002559703425504267,
          "train_metric": 3.2868963899090886e-05,
          "validation_metric": 2.5780393116292544e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.1391564940004173
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0003331890820845729,
          "train_loss": 0.0003935746499337256,
          "validation_loss": 0.00036410827306099236,
          "train_metric": 3.9639402530156076e-05,
          "validation_metric": 3.667164855869487e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.1592621809995762
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0004481005780689884,
          "train_loss": 0.0003914705594070256,
          "validation_loss": 0.00030650716507807374,
          "train_metric": 3.9427486626664177e-05,
          "validation_metric": 3.0870276532368734e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.179927115999817
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0002869001518774894,
          "train_loss": 0.0006783499266020954,
          "validation_loss": 0.0005354427848942578,
          "train_metric": 6.832091457908973e-05,
          "validation_metric": 5.3927829867461696e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.2008988460002
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0006123257917352021,
          "train_loss": 0.0011126358294859529,
          "validation_loss": 0.000758785055950284,
          "train_metric": 0.0001120605957112275,
          "validation_metric": 7.642206765012816e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.236309411999173
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.00039765636302036,
          "train_loss": 0.0003088544763159007,
          "validation_loss": 0.0002605428162496537,
          "train_metric": 3.1106694223126397e-05,
          "validation_metric": 2.624091757752467e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.2891058749992226
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0006069705614208942,
          "train_loss": 0.0005056319059804082,
          "validation_loss": 0.0005796283949166536,
          "train_metric": 5.092539868201129e-05,
          "validation_metric": 5.837804565089755e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.317764031999104
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.00031974642024579225,
          "train_loss": 0.00047997443471103907,
          "validation_loss": 0.00035590617335401475,
          "train_metric": 4.834127321373671e-05,
          "validation_metric": 3.5845572710968554e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.3380724010003178
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0004999093880542205,
          "train_loss": 0.00038301540189422667,
          "validation_loss": 0.00034783780574798584,
          "train_metric": 3.8575912185478956e-05,
          "validation_metric": 3.5032950108870864e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.3577865949991974
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0003517675531838904,
          "train_loss": 0.00031076959567144513,
          "validation_loss": 0.00029436880140565336,
          "train_metric": 3.129957622149959e-05,
          "validation_metric": 2.9647744668181986e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.3776476829989406
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.00028644320809689816,
          "train_loss": 0.0002948589390143752,
          "validation_loss": 0.00026483333203941584,
          "train_metric": 2.969711204059422e-05,
          "validation_metric": 2.6673042157199234e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.398189547000584
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0002584164990366844,
          "train_loss": 0.0002268178650410846,
          "validation_loss": 0.0001910840073833242,
          "train_metric": 2.284426591359079e-05,
          "validation_metric": 1.9245282601332292e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.419280643000093
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0003866503584504244,
          "train_loss": 0.0002743983641266823,
          "validation_loss": 0.00025097225443460047,
          "train_metric": 2.7636397135211155e-05,
          "validation_metric": 2.527700416976586e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.444988467999792
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0005636369814965292,
          "train_loss": 0.0005463949637487531,
          "validation_loss": 0.0004908788832835853,
          "train_metric": 5.503089050762355e-05,
          "validation_metric": 4.943953172187321e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4920881260004535
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0004801316317752935,
          "train_loss": 0.00025242145056836307,
          "validation_loss": 0.000188563484698534,
          "train_metric": 2.5422963517485186e-05,
          "validation_metric": 1.8991420802194625e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.551820417998897
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0003250063446103013,
          "train_loss": 0.00034747799509204924,
          "validation_loss": 0.0002625536872074008,
          "train_metric": 3.499671583995223e-05,
          "validation_metric": 2.6443447495694272e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.576000594999641
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0004047630818604375,
          "train_loss": 0.00024908105842769146,
          "validation_loss": 0.00019663479179143906,
          "train_metric": 2.508653051336296e-05,
          "validation_metric": 1.9804336261586286e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.5963091180001356
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.00029615815856232075,
          "train_loss": 0.00029795951559208333,
          "validation_loss": 0.00020598665287252516,
          "train_metric": 3.000938886543736e-05,
          "validation_metric": 2.0746221707668155e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.616816874999131
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.00041523096206219634,
          "train_loss": 0.0004350168746896088,
          "validation_loss": 0.000323029060382396,
          "train_metric": 4.38133065472357e-05,
          "validation_metric": 3.253430259064771e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.6371981319989573
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.00047772297830306343,
          "train_loss": 0.00021317134087439626,
          "validation_loss": 0.00016684246656950563,
          "train_metric": 2.146983752027154e-05,
          "validation_metric": 1.6803764083306305e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.6586054750005133
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0002261233598801482,
          "train_loss": 0.00020944596326444298,
          "validation_loss": 0.00016011275874916464,
          "train_metric": 2.109463093802333e-05,
          "validation_metric": 1.6125970432767645e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.6802604290005547
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.00026793491997523233,
          "train_loss": 0.0003631651634350419,
          "validation_loss": 0.00031121482606977224,
          "train_metric": 3.657666457002051e-05,
          "validation_metric": 3.1344417948275805e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7006710219993693
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0004104247436771402,
          "train_loss": 0.0005075770895928144,
          "validation_loss": 0.0004380839818622917,
          "train_metric": 5.1121311116730794e-05,
          "validation_metric": 4.412221460370347e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.721351149999464
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0003357457735546632,
          "train_loss": 0.00024625493097119033,
          "validation_loss": 0.00016511519788764417,
          "train_metric": 2.48018932325067e-05,
          "validation_metric": 1.6629797755740583e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7420008829994913
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.00021767637326775002,
          "train_loss": 0.00025487912353128195,
          "validation_loss": 0.00020541901176329702,
          "train_metric": 2.567049341450911e-05,
          "validation_metric": 2.0689049051725306e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.763341899999432
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.00022050567440601299,
          "train_loss": 0.0002039228711510077,
          "validation_loss": 0.00015409319894388318,
          "train_metric": 2.0538367607514374e-05,
          "validation_metric": 1.5519703083555214e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7857381399990118
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0002509558080419083,
          "train_loss": 0.00019998646166641265,
          "validation_loss": 0.00016245729057118297,
          "train_metric": 2.0141902496106923e-05,
          "validation_metric": 1.6362104361178353e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.806609926999954
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0002406021258138935,
          "train_loss": 0.00023250981757882982,
          "validation_loss": 0.00017436749476473778,
          "train_metric": 2.3417536795022897e-05,
          "validation_metric": 1.7561655113240704e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.8274613310004497
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0002468565617164131,
          "train_loss": 0.0002822740643750876,
          "validation_loss": 0.00020007371495012194,
          "train_metric": 2.8429607482394204e-05,
          "validation_metric": 2.015069367189426e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.848238569999012
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.00021601728212772286,
          "train_loss": 0.00025612389435991645,
          "validation_loss": 0.00019108643755316734,
          "train_metric": 2.5795859983190894e-05,
          "validation_metric": 1.924552816490177e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.868351191000329
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.00028545680879688007,
          "train_loss": 0.0001925661927089095,
          "validation_loss": 0.00013712367217522115,
          "train_metric": 1.939456160471309e-05,
          "validation_metric": 1.3810593372909352e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.897657681000055
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.00019955384868808324,
          "train_loss": 0.0001948920835275203,
          "validation_loss": 0.00016359683650080115,
          "train_metric": 1.962881833605934e-05,
          "validation_metric": 1.6476875316584483e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.951913651999348
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0002421858771413099,
          "train_loss": 0.0002277054009027779,
          "validation_loss": 0.00019054840959142894,
          "train_metric": 2.2933656509849243e-05,
          "validation_metric": 1.919134047057014e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.0025694699997985
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0002074286949209636,
          "train_loss": 0.00019043123757001013,
          "validation_loss": 0.00014628912322223186,
          "train_metric": 1.917953704833053e-05,
          "validation_metric": 1.4733704119862523e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.023088803000064
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.00021398349417722784,
          "train_loss": 0.0005510141490958631,
          "validation_loss": 0.00041311574750579894,
          "train_metric": 5.549612615141086e-05,
          "validation_metric": 4.160750540904701e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.042773551998835
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0003620385537033144,
          "train_loss": 0.0008017574436962605,
          "validation_loss": 0.0005906072910875082,
          "train_metric": 8.07500837254338e-05,
          "validation_metric": 5.9483809309313074e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.063078461000259
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.00045080598965796526,
          "train_loss": 0.00029394004377536476,
          "validation_loss": 0.00023609267373103648,
          "train_metric": 2.9604561859741807e-05,
          "validation_metric": 2.377838609390892e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.083558615000584
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0003109308054263238,
          "train_loss": 0.0002258072345284745,
          "validation_loss": 0.00013948159175924957,
          "train_metric": 2.274247890454717e-05,
          "validation_metric": 1.4048075172468089e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.161993530999098
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.00028704806663881755,
          "train_loss": 0.00020912753825541586,
          "validation_loss": 0.00017350616690237075,
          "train_metric": 2.1062558516860008e-05,
          "validation_metric": 1.74749056895962e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.239969244999884
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.00030609657187596895,
          "train_loss": 0.000272127625066787,
          "validation_loss": 0.0002603580942377448,
          "train_metric": 2.740769923548214e-05,
          "validation_metric": 2.6222309315926395e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.3155950300006225
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0006169639491417911,
          "train_loss": 0.0008312284480780363,
          "validation_loss": 0.0005312629509717226,
          "train_metric": 8.371826697839424e-05,
          "validation_metric": 5.3506853873841465e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.418992203000016
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.00040703573904465884,
          "train_loss": 0.000219659399590455,
          "validation_loss": 0.0001492545852670446,
          "train_metric": 2.212328945461195e-05,
          "validation_metric": 1.503237581346184e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.493354206999356
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.00023483772065446828,
          "train_loss": 0.0001831488189054653,
          "validation_loss": 0.00014253355038817972,
          "train_metric": 1.8446076865075156e-05,
          "validation_metric": 1.4355458915815689e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.567699580999033
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0002381908830102475,
          "train_loss": 0.00017332899733446538,
          "validation_loss": 0.000133434179588221,
          "train_metric": 1.7457061403547414e-05,
          "validation_metric": 1.343900294159539e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.644970651999756
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.00019288523208160768,
          "train_loss": 0.0002122207952197641,
          "validation_loss": 0.00017226394265890121,
          "train_metric": 2.1374100469984114e-05,
          "validation_metric": 1.7349791960441507e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.739419161000114
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0001909866359710577,
          "train_loss": 0.00017354890587739646,
          "validation_loss": 0.00013615851639769971,
          "train_metric": 1.7479207599535584e-05,
          "validation_metric": 1.3713387488678563e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.857385428998896
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.00021954392968837055,
          "train_loss": 0.0001755991397658363,
          "validation_loss": 0.00013984553515911102,
          "train_metric": 1.7685702914604917e-05,
          "validation_metric": 1.4084730537433643e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.9653849770002125
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.0002221235563411028,
          "train_loss": 0.00018217200704384595,
          "validation_loss": 0.00013991589366924018,
          "train_metric": 1.834769682318438e-05,
          "validation_metric": 1.4091816410655156e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.979671336999672
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.00022061976324039279,
          "train_loss": 0.00022677393280901015,
          "validation_loss": 0.00016838945157360286,
          "train_metric": 2.2839838493382558e-05,
          "validation_metric": 1.6959569620667025e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.993137655999817
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0002112208667313098,
          "train_loss": 0.0001737186685204506,
          "validation_loss": 0.0001423424982931465,
          "train_metric": 1.749630791891832e-05,
          "validation_metric": 1.4336216736410279e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.007911741000498
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.00019560735563572962,
          "train_loss": 0.00023969030007719994,
          "validation_loss": 0.00018801490659825504,
          "train_metric": 2.4140726964105852e-05,
          "validation_metric": 1.893617445603013e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.0238868100004765
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.00019152779032083345,
          "train_loss": 0.00017086473235394806,
          "validation_loss": 0.00013447242963593453,
          "train_metric": 1.72088693943806e-05,
          "validation_metric": 1.3543572094931733e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.036830246999671
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.00018027242867901805,
          "train_loss": 0.00020412685989867896,
          "validation_loss": 0.0001477296755183488,
          "train_metric": 2.0558909454848617e-05,
          "validation_metric": 1.487879126216285e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.049675969999953
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.00017462019422964659,
          "train_loss": 0.00017260668391827494,
          "validation_loss": 0.00013375090202316642,
          "train_metric": 1.7384312741342e-05,
          "validation_metric": 1.347090073977597e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.062333468998986
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.00017644088075030595,
          "train_loss": 0.00017135590314865112,
          "validation_loss": 0.00013424821372609586,
          "train_metric": 1.7258338630199432e-05,
          "validation_metric": 1.3520989341486711e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.075180792000538
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.00017693703239274328,
          "train_loss": 0.0001789337256923318,
          "validation_loss": 0.00013429073442239314,
          "train_metric": 1.8021552023128606e-05,
          "validation_metric": 1.3525270333047956e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.088011093999739
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0001921771827255725,
          "train_loss": 0.00016558755305595696,
          "validation_loss": 0.00012955306738149375,
          "train_metric": 1.667737160460092e-05,
          "validation_metric": 1.3048110304225702e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.10140795299958
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.00017881759094962035,
          "train_loss": 0.00016593669715803117,
          "validation_loss": 0.00013553421013057232,
          "train_metric": 1.671253630775027e-05,
          "validation_metric": 1.3650509572471492e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.114062889999332
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.00018966156767419307,
          "train_loss": 0.00016906003293115646,
          "validation_loss": 0.00013356014096643776,
          "train_metric": 1.7027108697220683e-05,
          "validation_metric": 1.3451687664201017e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.126505955000539
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0001686354275989288,
          "train_loss": 0.00016444384527858347,
          "validation_loss": 0.00013341540761757642,
          "train_metric": 1.6562182281631976e-05,
          "validation_metric": 1.3437111192615703e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.138831096999638
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.00016910136810110998,
          "train_loss": 0.0001697281695669517,
          "validation_loss": 0.0001283138699363917,
          "train_metric": 1.709439675323665e-05,
          "validation_metric": 1.2923302165290806e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.151788748000399
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.00017533852724227472,
          "train_loss": 0.0001584547571837902,
          "validation_loss": 0.00012432481162250042,
          "train_metric": 1.5958983567543328e-05,
          "validation_metric": 1.252154015674023e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.16521044000001
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.00016808173631943646,
          "train_loss": 0.0001582900295034051,
          "validation_loss": 0.0001278498093597591,
          "train_metric": 1.5942392565193586e-05,
          "validation_metric": 1.28765650515561e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.177951526999095
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0001724461462799809,
          "train_loss": 0.00015899755817372352,
          "validation_loss": 0.00012512155808508396,
          "train_metric": 1.6013651475077495e-05,
          "validation_metric": 1.2601783964782953e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.1902112329989905
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.00018300419424122083,
          "train_loss": 0.0001682780566625297,
          "validation_loss": 0.0001409952819813043,
          "train_metric": 1.6948350094025955e-05,
          "validation_metric": 1.4200530131347477e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.2025339159990835
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.00016299367416650057,
          "train_loss": 0.00016573297034483403,
          "validation_loss": 0.00012830815103370696,
          "train_metric": 1.6692018107278273e-05,
          "validation_metric": 1.2922726455144584e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.214850718999514
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.00019057797499044682,
          "train_loss": 0.00015954651462379843,
          "validation_loss": 0.00012303152470849454,
          "train_metric": 1.6068941476987675e-05,
          "validation_metric": 1.2391284144541714e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.228383603000111
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.00017596558882360114,
          "train_loss": 0.0001748956274241209,
          "validation_loss": 0.00014836477930657566,
          "train_metric": 1.761484418238979e-05,
          "validation_metric": 1.494275602453854e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.241373716000453
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.00016648792143314495,
          "train_loss": 0.0001668957556830719,
          "validation_loss": 0.00012553718988783658,
          "train_metric": 1.6809128283057362e-05,
          "validation_metric": 1.2643646186916158e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.254329452000093
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.00016494239071107586,
          "train_loss": 0.0001553035544930026,
          "validation_loss": 0.00012152733688708395,
          "train_metric": 1.5641604477423243e-05,
          "validation_metric": 1.2239789612067398e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.267961681000088
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.00016111069180624327,
          "train_loss": 0.00016266675083898008,
          "validation_loss": 0.00013522146036848426,
          "train_metric": 1.6383201000280678e-05,
          "validation_metric": 1.3619010132970288e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.2807386059994315
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.00016142015783771058,
          "train_loss": 0.00016027076344471425,
          "validation_loss": 0.00012306190910749137,
          "train_metric": 1.6141882952069864e-05,
          "validation_metric": 1.239434459421318e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.293581865000306
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0001674862496656715,
          "train_loss": 0.00016953803424257785,
          "validation_loss": 0.00013831765681970865,
          "train_metric": 1.7075250070774928e-05,
          "validation_metric": 1.3930848581367172e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.306125932998839
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.00017088445247281925,
          "train_loss": 0.00016218898235820234,
          "validation_loss": 0.00013252333155833185,
          "train_metric": 1.6335081454599276e-05,
          "validation_metric": 1.3347263120522257e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.318679780999446
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.00016113983565446688,
          "train_loss": 0.00016288187180180103,
          "validation_loss": 0.00012246874393895268,
          "train_metric": 1.640486516407691e-05,
          "validation_metric": 1.2334603525232524e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.331338030999177
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.00016957834941422334,
          "train_loss": 0.00016950565623119473,
          "validation_loss": 0.00012820778647437692,
          "train_metric": 1.707198862277437e-05,
          "validation_metric": 1.2912620150018483e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.343772943999284
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.00016269777097477345,
          "train_loss": 0.00015231393626891077,
          "validation_loss": 0.00012143931962782517,
          "train_metric": 1.534050352347549e-05,
          "validation_metric": 1.2230922948219813e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.356959572000051
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.00015858231790844002,
          "train_loss": 0.0002497771929483861,
          "validation_loss": 0.0002021567488554865,
          "train_metric": 2.5156645278912038e-05,
          "validation_metric": 2.0360484995762818e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.369905272000324
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.00021554091608777526,
          "train_loss": 0.00015777067164890468,
          "validation_loss": 0.0001296417322009802,
          "train_metric": 1.589008570590522e-05,
          "validation_metric": 1.3057041542197112e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.3825492129999475
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.00015791271243870142,
          "train_loss": 0.00016162989777512848,
          "validation_loss": 0.0001232550130225718,
          "train_metric": 1.6278770999633707e-05,
          "validation_metric": 1.2413795047905296e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.395138541000051
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.00015707138891229988,
          "train_loss": 0.00016427742957603186,
          "validation_loss": 0.0001316763664362952,
          "train_metric": 1.65454202942783e-05,
          "validation_metric": 1.3261961612442974e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.407752741999502
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.00016491395854245638,
          "train_loss": 0.0001511566515546292,
          "validation_loss": 0.00012219144264236093,
          "train_metric": 1.5223943591990974e-05,
          "validation_metric": 1.2306675671425182e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.420262748999448
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.00016016758536352427,
          "train_loss": 0.00015513176913373172,
          "validation_loss": 0.00012186868116259575,
          "train_metric": 1.562430406920612e-05,
          "validation_metric": 1.2274167602299713e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.432826305999697
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0001593631068317336,
          "train_loss": 0.00015553904813714325,
          "validation_loss": 0.00012870585487689823,
          "train_metric": 1.5665322280256078e-05,
          "validation_metric": 1.2962780601810664e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.4452777879996574
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.00015242234667312005,
          "train_loss": 0.00016039938782341778,
          "validation_loss": 0.00012111449905205518,
          "train_metric": 1.6154837794601917e-05,
          "validation_metric": 1.2198210242786445e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.458607069998834
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.00017113809099100763,
          "train_loss": 0.00014897040091454983,
          "validation_loss": 0.00011947266466449946,
          "train_metric": 1.5003754015197046e-05,
          "validation_metric": 1.2032847735099494e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.472048409999843
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.00016252553768936195,
          "train_loss": 0.0001570174645166844,
          "validation_loss": 0.0001296235859626904,
          "train_metric": 1.5814222933840938e-05,
          "validation_metric": 1.3055213457846548e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.484844227999929
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.00016825611101012328,
          "train_loss": 0.0001507794513599947,
          "validation_loss": 0.00012316122592892498,
          "train_metric": 1.518595490779262e-05,
          "validation_metric": 1.240434812643798e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.497633400000268
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.00015442142239407985,
          "train_loss": 0.00015438236005138606,
          "validation_loss": 0.00012114206765545532,
          "train_metric": 1.5548825103905983e-05,
          "validation_metric": 1.2200986020616256e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.510199492000538
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.00016305746430589352,
          "train_loss": 0.0001537699718028307,
          "validation_loss": 0.00012110365787521005,
          "train_metric": 1.548714681121055e-05,
          "validation_metric": 1.219711521116551e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.522928170999876
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.00015594661635986995,
          "train_loss": 0.00015167059609666467,
          "validation_loss": 0.00012721589882858098,
          "train_metric": 1.527570748294238e-05,
          "validation_metric": 1.281271852349164e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.535618736999822
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.00015730635050204,
          "train_loss": 0.000153502551256679,
          "validation_loss": 0.00012577304732985795,
          "train_metric": 1.546021485410165e-05,
          "validation_metric": 1.2667399460042361e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.5481231230005505
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.00016225741001107963,
          "train_loss": 0.0001548809086671099,
          "validation_loss": 0.00012994848657399416,
          "train_metric": 1.5599036487401463e-05,
          "validation_metric": 1.3087936167721637e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.5607277670005715
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.00015986752578101004,
          "train_loss": 0.00016795439296402037,
          "validation_loss": 0.000126347309560515,
          "train_metric": 1.691575198492501e-05,
          "validation_metric": 1.272523877560161e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.573425487998975
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.00016344196683348855,
          "train_loss": 0.00015008471382316202,
          "validation_loss": 0.00011694519344018772,
          "train_metric": 1.511598293291172e-05,
          "validation_metric": 1.1778291082009673e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.586687290000555
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.00015242089557432337,
          "train_loss": 0.00015679239004384726,
          "validation_loss": 0.0001325382909271866,
          "train_metric": 1.5791556506883353e-05,
          "validation_metric": 1.3348771062737796e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.5998409439998795
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.00014189978128342773,
          "train_loss": 0.000204338546609506,
          "validation_loss": 0.0001438927138224244,
          "train_metric": 2.0580229829647578e-05,
          "validation_metric": 1.4492345144390129e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.6123525540006085
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.00021200962828515912,
          "train_loss": 0.00017812332953326404,
          "validation_loss": 0.0001433926518075168,
          "train_metric": 1.7939930330612697e-05,
          "validation_metric": 1.4441980965784751e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.624770554999486
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.00019249887509431574,
          "train_loss": 0.0001446761016268283,
          "validation_loss": 0.00012072749086655676,
          "train_metric": 1.4571245628758334e-05,
          "validation_metric": 1.215923111885786e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.637083290999726
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.00017839595921032014,
          "train_loss": 0.00015473015082534403,
          "validation_loss": 0.00012707163114100695,
          "train_metric": 1.5583853382850066e-05,
          "validation_metric": 1.2798189345630817e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.64973784599897
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.0001526986850421963,
          "train_loss": 0.00015846433234401047,
          "validation_loss": 0.0001317514688707888,
          "train_metric": 1.5959945812937804e-05,
          "validation_metric": 1.3269523151393514e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.662368620000052
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.00015213426695481758,
          "train_loss": 0.00019425329810474068,
          "validation_loss": 0.0001424271467840299,
          "train_metric": 1.956447886186652e-05,
          "validation_metric": 1.4344740520755295e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 5.675083714999346
        }
      ],
      "best": {
        "epoch": 233,
        "updates": 3728
      },
      "best_validation_metric": 1.1778291082009673e-05,
      "config": {
        "label": "LayerNorm + silu + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.003431951497619055
    },
    "13e54c648cc2_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.19096134742721915,
          "train_loss": 0.07242120057344437,
          "validation_loss": 0.0500011146068573,
          "train_metric": 0.007293997798115015,
          "validation_metric": 0.005035929381847382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02711385399925348
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.05849547719117254,
          "train_loss": 0.038079965859651566,
          "validation_loss": 0.02615676075220108,
          "train_metric": 0.003835274837911129,
          "validation_metric": 0.0026344130747020245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05951293599900964
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.03498553414829075,
          "train_loss": 0.02507447637617588,
          "validation_loss": 0.02035215124487877,
          "train_metric": 0.0025254094507545233,
          "validation_metric": 0.002049794187769294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08335360899945954
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.023305899696424603,
          "train_loss": 0.019549626857042313,
          "validation_loss": 0.014797462150454521,
          "train_metric": 0.0019689667969942093,
          "validation_metric": 0.0014903461560606956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11825258299904817
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.03700241423211992,
          "train_loss": 0.023921405896544456,
          "validation_loss": 0.018056131899356842,
          "train_metric": 0.002409276319667697,
          "validation_metric": 0.0018185474909842014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14472404399930383
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.028315134666627273,
          "train_loss": 0.022200098261237144,
          "validation_loss": 0.014051035977900028,
          "train_metric": 0.0022359120193868876,
          "validation_metric": 0.0014151687500998378,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16596955400018487
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.03941670642234385,
          "train_loss": 0.047134917229413986,
          "validation_loss": 0.04291398823261261,
          "train_metric": 0.004747255705296993,
          "validation_metric": 0.004322139546275139,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18803240200031723
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.03125104267382994,
          "train_loss": 0.025500396266579628,
          "validation_loss": 0.018091389909386635,
          "train_metric": 0.002568306401371956,
          "validation_metric": 0.0018220983911305666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21189567900000839
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.01859823355334811,
          "train_loss": 0.014343221671879292,
          "validation_loss": 0.014042512513697147,
          "train_metric": 0.0014445967972278595,
          "validation_metric": 0.0014143104199320078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23311794899927918
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.01476419367827475,
          "train_loss": 0.011451940052211285,
          "validation_loss": 0.011174211278557777,
          "train_metric": 0.0011533974902704358,
          "validation_metric": 0.0011254255659878254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24923394699908386
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.013102602475555614,
          "train_loss": 0.009634414687752724,
          "validation_loss": 0.009227216243743896,
          "train_metric": 0.000970343011431396,
          "validation_metric": 0.0009293314651586115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26519892800024536
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.01181740511674434,
          "train_loss": 0.027752339839935303,
          "validation_loss": 0.021214960142970085,
          "train_metric": 0.002795114181935787,
          "validation_metric": 0.0021366931032389402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28037657300046703
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.024153543607098982,
          "train_loss": 0.010684475302696228,
          "validation_loss": 0.0072398437187075615,
          "train_metric": 0.001076101209037006,
          "validation_metric": 0.0007291705696843565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.370214919999853
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.020216720033204183,
          "train_loss": 0.012447123415768147,
          "validation_loss": 0.013840309344232082,
          "train_metric": 0.0012536287540569901,
          "validation_metric": 0.0013939451891928911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44573807300002954
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.0160851250984706,
          "train_loss": 0.020313572138547897,
          "validation_loss": 0.014188272878527641,
          "train_metric": 0.002045908709987998,
          "validation_metric": 0.0014289908576756716,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5211689879997721
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.01265521373716183,
          "train_loss": 0.01249550748616457,
          "validation_loss": 0.008259164169430733,
          "train_metric": 0.001258501666598022,
          "validation_metric": 0.000831832701805979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5958391269996355
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.008663570624776185,
          "train_loss": 0.008145946078002453,
          "validation_loss": 0.005313525442034006,
          "train_metric": 0.0008204297628253698,
          "validation_metric": 0.0005351588479243219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6728470460002427
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.010117787925992161,
          "train_loss": 0.006461773533374071,
          "validation_loss": 0.004270083270967007,
          "train_metric": 0.0006508060614578426,
          "validation_metric": 0.0004300671280361712,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.747894782000003
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.011943971047003288,
          "train_loss": 0.006918198429048061,
          "validation_loss": 0.005362193565815687,
          "train_metric": 0.0006967756198719144,
          "validation_metric": 0.0005400604568421841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8240008630000375
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.007399222507956438,
          "train_loss": 0.005954752676188946,
          "validation_loss": 0.00478105153888464,
          "train_metric": 0.000599740887992084,
          "validation_metric": 0.0004815299471374601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8989945270004682
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.007807362839230336,
          "train_loss": 0.005216581281274557,
          "validation_loss": 0.003698101732879877,
          "train_metric": 0.0005253948620520532,
          "validation_metric": 0.00037245929706841707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9749258119991282
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.007189471689343918,
          "train_loss": 0.005692244507372379,
          "validation_loss": 0.006105087231844664,
          "train_metric": 0.0005733020952902734,
          "validation_metric": 0.000614881981164217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0473717820004822
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.01745169445348438,
          "train_loss": 0.02158437669277191,
          "validation_loss": 0.019700421020388603,
          "train_metric": 0.0021738994400948286,
          "validation_metric": 0.0019841541070491076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0877970620003907
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.016935257881414145,
          "train_loss": 0.011732752434909344,
          "validation_loss": 0.015319348312914371,
          "train_metric": 0.0011816798942163587,
          "validation_metric": 0.0015429084887728095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.122354670000277
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.008408344830968417,
          "train_loss": 0.007929813116788864,
          "validation_loss": 0.005407745484262705,
          "train_metric": 0.0007986616692505777,
          "validation_metric": 0.0005446483264677227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.159762752999086
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.007061928350594826,
          "train_loss": 0.006596588063985109,
          "validation_loss": 0.006484700832515955,
          "train_metric": 0.0006643841043114662,
          "validation_metric": 0.0006531153339892626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1879069309998158
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.012771852227160707,
          "train_loss": 0.007400437258183956,
          "validation_loss": 0.008532926440238953,
          "train_metric": 0.0007453449070453644,
          "validation_metric": 0.0008594051469117403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2203806569996232
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.017370946967275813,
          "train_loss": 0.011434217914938927,
          "validation_loss": 0.008804738521575928,
          "train_metric": 0.0011516126105561852,
          "validation_metric": 0.0008867810829542577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2502338659996894
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.02165296656312421,
          "train_loss": 0.010835460387170315,
          "validation_loss": 0.009168559685349464,
          "train_metric": 0.0010913078440353274,
          "validation_metric": 0.0009234237368218601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.284770550999383
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.015899773105047643,
          "train_loss": 0.007458425592631102,
          "validation_loss": 0.005842586047947407,
          "train_metric": 0.0007511853473260999,
          "validation_metric": 0.0005884440615773201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3169347589991958
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008075340345385484,
          "train_loss": 0.011918378062546253,
          "validation_loss": 0.0076078916899859905,
          "train_metric": 0.0012003753799945116,
          "validation_metric": 0.0007662390125915408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3497507499996573
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.014930206874851137,
          "train_loss": 0.01142839714884758,
          "validation_loss": 0.011296026408672333,
          "train_metric": 0.0011510264594107866,
          "validation_metric": 0.0011376944603398442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3813925239992386
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.008419537691224832,
          "train_loss": 0.00673066359013319,
          "validation_loss": 0.005948117934167385,
          "train_metric": 0.0006778877577744424,
          "validation_metric": 0.0005990726058371365,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4138118449991453
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.0060232751711737365,
          "train_loss": 0.006307695992290974,
          "validation_loss": 0.005712182726711035,
          "train_metric": 0.0006352879572659731,
          "validation_metric": 0.0005753101431764662,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.448960915000498
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.004673073803132866,
          "train_loss": 0.004261978901922703,
          "validation_loss": 0.0036122885067015886,
          "train_metric": 0.00042925088200718164,
          "validation_metric": 0.0003638164489530027,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.4834023829989746
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.004866482049692422,
          "train_loss": 0.003506392240524292,
          "validation_loss": 0.0031671205069869757,
          "train_metric": 0.00035315100103616714,
          "validation_metric": 0.00031898077577352524,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5050884119991679
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.004182442084129434,
          "train_loss": 0.0037244250997900963,
          "validation_loss": 0.0042476532980799675,
          "train_metric": 0.0003751104522962123,
          "validation_metric": 0.00042780805961228907,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.521746976999566
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.0033289198036072776,
          "train_loss": 0.0033291378058493137,
          "validation_loss": 0.0033484972082078457,
          "train_metric": 0.0003352985659148544,
          "validation_metric": 0.0003372483770363033,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.53776052899957
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.005780861814855598,
          "train_loss": 0.020466893911361694,
          "validation_loss": 0.011570031754672527,
          "train_metric": 0.0020613507367670536,
          "validation_metric": 0.0011652912944555283,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5532836900001712
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012012778868665919,
          "train_loss": 0.007817122153937817,
          "validation_loss": 0.005699011962860823,
          "train_metric": 0.0007873119320720434,
          "validation_metric": 0.0005739836487919092,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5682538479995856
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.005877621122635901,
          "train_loss": 0.003909142687916756,
          "validation_loss": 0.0036250378470867872,
          "train_metric": 0.0003937145520467311,
          "validation_metric": 0.0003651005099527538,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5840950149995479
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.004109784596948884,
          "train_loss": 0.003253644797950983,
          "validation_loss": 0.0028266971930861473,
          "train_metric": 0.00032769516110420227,
          "validation_metric": 0.00028469457174651325,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.599535516999822
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.002991187764564529,
          "train_loss": 0.0044402494095265865,
          "validation_loss": 0.0031102511566132307,
          "train_metric": 0.0004472056753002107,
          "validation_metric": 0.00031325314193964005,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6152781239998149
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.0037397220730781555,
          "train_loss": 0.00262434221804142,
          "validation_loss": 0.0023731093388050795,
          "train_metric": 0.0002643141779117286,
          "validation_metric": 0.00023901090025901794,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6315518549999979
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.0027903653026442043,
          "train_loss": 0.0037203801330178976,
          "validation_loss": 0.004583532456308603,
          "train_metric": 0.0003747030277736485,
          "validation_metric": 0.0004616366350091994,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.647393381999791
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.0037627266501658596,
          "train_loss": 0.005069808568805456,
          "validation_loss": 0.0037985427770763636,
          "train_metric": 0.0005106126191094518,
          "validation_metric": 0.00038257535197772086,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.663174691999302
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0039194662240333855,
          "train_loss": 0.0027494276873767376,
          "validation_loss": 0.002865391783416271,
          "train_metric": 0.000276912294793874,
          "validation_metric": 0.00028859174926765263,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6783566100002645
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0028392383937898558,
          "train_loss": 0.002893801312893629,
          "validation_loss": 0.0020938196685165167,
          "train_metric": 0.0002914530341513455,
          "validation_metric": 0.00021088187349960208,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6961981509994075
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.0029030974255874753,
          "train_loss": 0.0026001506485044956,
          "validation_loss": 0.0025033329147845507,
          "train_metric": 0.0002618776634335518,
          "validation_metric": 0.0002521264832466841,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7115842780003732
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.003035804635146633,
          "train_loss": 0.002211825456470251,
          "validation_loss": 0.0013373836409300566,
          "train_metric": 0.00022276696108747274,
          "validation_metric": 0.00013469636905938387,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7275795229998039
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.004009369098639581,
          "train_loss": 0.0054374272003769875,
          "validation_loss": 0.006675504148006439,
          "train_metric": 0.0005476378137245774,
          "validation_metric": 0.0006723323022015393,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7413971179994405
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.008843945121043362,
          "train_loss": 0.003147720592096448,
          "validation_loss": 0.0023638326674699783,
          "train_metric": 0.00031702694832347333,
          "validation_metric": 0.00023807653633411974,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7562625680002384
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.004230756803735858,
          "train_loss": 0.0027058853302150965,
          "validation_loss": 0.001981703331694007,
          "train_metric": 0.000272526842309162,
          "validation_metric": 0.00019958992197643965,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7708608969987836
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.002749866478552576,
          "train_loss": 0.0023987486492842436,
          "validation_loss": 0.002011180855333805,
          "train_metric": 0.000241593224927783,
          "validation_metric": 0.00020255876006558537,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7851052339992748
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.0031872097606537864,
          "train_loss": 0.006567193660885096,
          "validation_loss": 0.004181003663688898,
          "train_metric": 0.0006614237208850682,
          "validation_metric": 0.0004210954066365957,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7992846309989545
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0040504113712813705,
          "train_loss": 0.0038834644947201014,
          "validation_loss": 0.0025692638009786606,
          "train_metric": 0.0003911282983608544,
          "validation_metric": 0.00025876687141135335,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8132055579990265
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.0031791906876605935,
          "train_loss": 0.0033707311376929283,
          "validation_loss": 0.002545719500631094,
          "train_metric": 0.00033948765485547483,
          "validation_metric": 0.0002563955495133996,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.827364034999846
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.004306578030082164,
          "train_loss": 0.004744565114378929,
          "validation_loss": 0.004423259757459164,
          "train_metric": 0.00047785521019250154,
          "validation_metric": 0.0004454944864846766,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.841852352999922
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.011845290544442832,
          "train_loss": 0.008682968094944954,
          "validation_loss": 0.008212940767407417,
          "train_metric": 0.0008745167287997901,
          "validation_metric": 0.0008271773112937808,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.856285380999907
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.00589088567357976,
          "train_loss": 0.0032587270252406597,
          "validation_loss": 0.0030251448042690754,
          "train_metric": 0.0003282070392742753,
          "validation_metric": 0.0003046815108973533,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8704493209988868
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.003233864590583835,
          "train_loss": 0.0021209847182035446,
          "validation_loss": 0.0015765625284984708,
          "train_metric": 0.00021361783728934824,
          "validation_metric": 0.00015878560952842236,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8844633249991602
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0021993615227984264,
          "train_loss": 0.0018497803248465061,
          "validation_loss": 0.0014311582781374454,
          "train_metric": 0.00018630307749845088,
          "validation_metric": 0.0001441409986000508,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.8979415929989045
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.0025348320523335133,
          "train_loss": 0.00437528733164072,
          "validation_loss": 0.0029956388752907515,
          "train_metric": 0.0004406629013828933,
          "validation_metric": 0.0003017098060809076,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.912885138999627
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0037865651538595557,
          "train_loss": 0.0023272214457392693,
          "validation_loss": 0.0025570227298885584,
          "train_metric": 0.0002343892410863191,
          "validation_metric": 0.0002575339749455452,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9271244600004138
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.002583433990366757,
          "train_loss": 0.0019464744254946709,
          "validation_loss": 0.001781255821697414,
          "train_metric": 0.00019604177214205265,
          "validation_metric": 0.00017940156976692379,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9414711099998385
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0018684041897358838,
          "train_loss": 0.0015007505426183343,
          "validation_loss": 0.0012852641521021724,
          "train_metric": 0.00015115010319277644,
          "validation_metric": 0.00012944710033480078,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9561448189997463
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0017632504059292842,
          "train_loss": 0.0016148935537785292,
          "validation_loss": 0.0013212879421189427,
          "train_metric": 0.00016264617443084717,
          "validation_metric": 0.0001330752857029438,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9698812289989291
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0021042677835794166,
          "train_loss": 0.0019466588273644447,
          "validation_loss": 0.002111217239871621,
          "train_metric": 0.0001960603694897145,
          "validation_metric": 0.00021263408416416496,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.997897680999813
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.002125705230355379,
          "train_loss": 0.002378548262640834,
          "validation_loss": 0.0018497775308787823,
          "train_metric": 0.00023955866345204413,
          "validation_metric": 0.00018630280101206154,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.130272750999211
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.0020998310974391643,
          "train_loss": 0.002550025936216116,
          "validation_loss": 0.002754666842520237,
          "train_metric": 0.0002568292838986963,
          "validation_metric": 0.00027743997634388506,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1618628020005417
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0013822922901454149,
          "train_loss": 0.0013652424095198512,
          "validation_loss": 0.0009722681134007871,
          "train_metric": 0.0001375022402498871,
          "validation_metric": 9.792328637558967e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1837448329988547
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.001340479828286334,
          "train_loss": 0.0011734109139069915,
          "validation_loss": 0.0013409514212980866,
          "train_metric": 0.00011818166240118444,
          "validation_metric": 0.00013505571405403316,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2039612759999727
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0012831277890654746,
          "train_loss": 0.0011626267805695534,
          "validation_loss": 0.0013831340475007892,
          "train_metric": 0.00011709552200045437,
          "validation_metric": 0.00013930417480878532,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2247039640005823
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0012865321623394266,
          "train_loss": 0.0013982495293021202,
          "validation_loss": 0.000984483165666461,
          "train_metric": 0.00014082655252423137,
          "validation_metric": 9.915353439282626e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.246046309999656
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0015238871856126934,
          "train_loss": 0.0032170426566153765,
          "validation_loss": 0.001395162078551948,
          "train_metric": 0.000324008782627061,
          "validation_metric": 0.0001405156363034621,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2666552169994247
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.00917989496520022,
          "train_loss": 0.0029154890216886997,
          "validation_loss": 0.0022291080094873905,
          "train_metric": 0.0002936373930424452,
          "validation_metric": 0.00022450760297942907,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2872258339994005
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.009327516076155007,
          "train_loss": 0.0037408850621432066,
          "validation_loss": 0.0024357882793992758,
          "train_metric": 0.0003767682646866888,
          "validation_metric": 0.0002453236375004053,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.308931899999152
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.004098504556168336,
          "train_loss": 0.005117382854223251,
          "validation_loss": 0.0056041209027171135,
          "train_metric": 0.0005154040409252048,
          "validation_metric": 0.0005644265911541879,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3307022680000955
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.004834919891436584,
          "train_loss": 0.002767703030258417,
          "validation_loss": 0.00194906210526824,
          "train_metric": 0.00027875290834344923,
          "validation_metric": 0.0001963023969437927,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.351643031999629
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.0024070317449513823,
          "train_loss": 0.0014774608425796032,
          "validation_loss": 0.0012961975298821926,
          "train_metric": 0.00014880446542520076,
          "validation_metric": 0.00013054827286396176,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3720644699988043
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0020394499952089973,
          "train_loss": 0.0024413273204118013,
          "validation_loss": 0.0021273442544043064,
          "train_metric": 0.0002458815579302609,
          "validation_metric": 0.00021425829618237913,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.392977941999561
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0021871748685953207,
          "train_loss": 0.001280031050555408,
          "validation_loss": 0.0009266159031540155,
          "train_metric": 0.00012892004451714456,
          "validation_metric": 9.332536137662828e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4156846739988396
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0014159942547848914,
          "train_loss": 0.0011689498787745833,
          "validation_loss": 0.0012920433655381203,
          "train_metric": 0.00011773234291467816,
          "validation_metric": 0.00013012987619731575,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4372614479998447
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0019502067189023364,
          "train_loss": 0.0020715631544589996,
          "validation_loss": 0.0014526501763612032,
          "train_metric": 0.00020864023827016354,
          "validation_metric": 0.0001463056105421856,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4590493319992675
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0035985013091703877,
          "train_loss": 0.007072011940181255,
          "validation_loss": 0.0056052859872579575,
          "train_metric": 0.0007122671231627464,
          "validation_metric": 0.0005645438213832676,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.480936689000373
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.005503465581568889,
          "train_loss": 0.009836027398705482,
          "validation_loss": 0.007717814762145281,
          "train_metric": 0.0009906487539410591,
          "validation_metric": 0.0007773100514896214,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.501758374999554
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.007217426420538686,
          "train_loss": 0.004359635524451733,
          "validation_loss": 0.005277134012430906,
          "train_metric": 0.0004390865215100348,
          "validation_metric": 0.0005314936861395836,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5227929209995636
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.0028454756320570596,
          "train_loss": 0.0014330454869195819,
          "validation_loss": 0.0012547599617391825,
          "train_metric": 0.00014433107571676373,
          "validation_metric": 0.00012637482723221183,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5590379959994607
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0016214200295507908,
          "train_loss": 0.0014170532813295722,
          "validation_loss": 0.0009488885989412665,
          "train_metric": 0.00014272041153162718,
          "validation_metric": 9.556858276482671e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5922955879996152
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.001475065218983218,
          "train_loss": 0.0011199433356523514,
          "validation_loss": 0.0012644421076402068,
          "train_metric": 0.00011279658792773262,
          "validation_metric": 0.00012734998017549515,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.620362669000315
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.001303806569922017,
          "train_loss": 0.0012208110420033336,
          "validation_loss": 0.001109238713979721,
          "train_metric": 0.00012295562191866338,
          "validation_metric": 0.00011171845835633576,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.648751433000143
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0012409669070621021,
          "train_loss": 0.001781233586370945,
          "validation_loss": 0.0021014062222093344,
          "train_metric": 0.00017939932877197862,
          "validation_metric": 0.00021164595091249794,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.670116519999283
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0018898851158155594,
          "train_loss": 0.0012283617397770286,
          "validation_loss": 0.0013221341650933027,
          "train_metric": 0.00012371607590466738,
          "validation_metric": 0.00013316050171852112,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.6914926770004968
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0010668466511560837,
          "train_loss": 0.002210868988186121,
          "validation_loss": 0.0022857426665723324,
          "train_metric": 0.00022267061285674572,
          "validation_metric": 0.00023021161905489862,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7131247149991395
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0010258891816192772,
          "train_loss": 0.0013494559098035097,
          "validation_loss": 0.0017974043730646372,
          "train_metric": 0.00013591225433629006,
          "validation_metric": 0.0001810279645724222,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.73397359000046
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.00126638446454308,
          "train_loss": 0.0013460364425554872,
          "validation_loss": 0.0008310183766297996,
          "train_metric": 0.00013556786871049553,
          "validation_metric": 8.369712304556742e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7564729019995866
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0009186930237774504,
          "train_loss": 0.0008125420426949859,
          "validation_loss": 0.0008431181195192039,
          "train_metric": 8.183626778190956e-05,
          "validation_metric": 8.491576591040939e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7782569419996435
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0010077269489556784,
          "train_loss": 0.0008370034629479051,
          "validation_loss": 0.0010657833190634847,
          "train_metric": 8.429991430602968e-05,
          "validation_metric": 0.00010734178795246407,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.799224864998905
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0014753078467037994,
          "train_loss": 0.0012691402807831764,
          "validation_loss": 0.0015864407178014517,
          "train_metric": 0.00012782315025106072,
          "validation_metric": 0.0001597805239725858,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8188603870003135
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.001370498641335871,
          "train_loss": 0.0014210465596988797,
          "validation_loss": 0.0009151682606898248,
          "train_metric": 0.00014312259736470878,
          "validation_metric": 9.21723767532967e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8386528179999004
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0016597416033619083,
          "train_loss": 0.0011400147341191769,
          "validation_loss": 0.001797438831999898,
          "train_metric": 0.00011481811088742688,
          "validation_metric": 0.00018103147158399224,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.858450030998938
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0008897268562577665,
          "train_loss": 0.0010071308352053165,
          "validation_loss": 0.0008516609668731689,
          "train_metric": 0.00010143452527699992,
          "validation_metric": 8.577616972615942e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8781906929998513
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0011060504948545713,
          "train_loss": 0.0015836695674806833,
          "validation_loss": 0.0021985177882015705,
          "train_metric": 0.00015950138913467526,
          "validation_metric": 0.00022142667148727924,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.898071882000295
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0008709523463039659,
          "train_loss": 0.0012017281260341406,
          "validation_loss": 0.0008127872133627534,
          "train_metric": 0.0001210336631629616,
          "validation_metric": 8.186095510609448e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9198441690004984
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0009271223225368885,
          "train_loss": 0.0009953181724995375,
          "validation_loss": 0.0012863494921475649,
          "train_metric": 0.00010024481161963195,
          "validation_metric": 0.0001295564288739115,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9403134359999967
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0009875291943899356,
          "train_loss": 0.001880997559055686,
          "validation_loss": 0.0009553995332680643,
          "train_metric": 0.00018944717885460705,
          "validation_metric": 9.622435027267784e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.960509651000393
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0019020123945665546,
          "train_loss": 0.0046744984574615955,
          "validation_loss": 0.004984553437680006,
          "train_metric": 0.0004707983462139964,
          "validation_metric": 0.0005020259995944798,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.981173730999217
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0014769537119718734,
          "train_loss": 0.0011395589681342244,
          "validation_loss": 0.0007334442343562841,
          "train_metric": 0.00011477219231892377,
          "validation_metric": 7.386982906609774e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.009188637999614
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0010206940223724814,
          "train_loss": 0.0006933643599040806,
          "validation_loss": 0.0008058531675487757,
          "train_metric": 6.983311322983354e-05,
          "validation_metric": 8.11625795904547e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.0339884949989937
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.001232464437634917,
          "train_loss": 0.0011716826120391488,
          "validation_loss": 0.0008866343414410949,
          "train_metric": 0.00011800757783930749,
          "validation_metric": 8.929856267059222e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.054158949000339
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.0024299891338159796,
          "train_loss": 0.0035711105447262526,
          "validation_loss": 0.005760783329606056,
          "train_metric": 0.0003596691822167486,
          "validation_metric": 0.0005802050000056624,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.0746927879990835
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.001881768357634428,
          "train_loss": 0.0008101369021460414,
          "validation_loss": 0.0008735927985981107,
          "train_metric": 8.159402932506055e-05,
          "validation_metric": 8.798507042229176e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.100130179998814
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0009292882714362349,
          "train_loss": 0.0008190026856027544,
          "validation_loss": 0.0009469925425946712,
          "train_metric": 8.248696394730359e-05,
          "validation_metric": 9.537761798128486e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.1213203389997943
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0007986494038050296,
          "train_loss": 0.001102045294828713,
          "validation_loss": 0.0007419926696456969,
          "train_metric": 0.00011099397670477629,
          "validation_metric": 7.473080040654168e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.141916153999773
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.001027336043989635,
          "train_loss": 0.001405928866006434,
          "validation_loss": 0.0013679183321073651,
          "train_metric": 0.0001416000013705343,
          "validation_metric": 0.00013777169806417078,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.161765472999832
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0009564147876517382,
          "train_loss": 0.0007144286064431071,
          "validation_loss": 0.00066844851244241,
          "train_metric": 7.195462967501953e-05,
          "validation_metric": 6.732368638040498e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.182641604998935
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0008141277667164104,
          "train_loss": 0.0007697175024077296,
          "validation_loss": 0.0006308048032224178,
          "train_metric": 7.75231237639673e-05,
          "validation_metric": 6.353235221467912e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.2053242450001562
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0007707466702413512,
          "train_loss": 0.000583234301302582,
          "validation_loss": 0.0007818989688530564,
          "train_metric": 5.8741221437230706e-05,
          "validation_metric": 7.875000301282853e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.2260516059996007
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0006709761619276833,
          "train_loss": 0.0008465687860734761,
          "validation_loss": 0.0007517293561249971,
          "train_metric": 8.526330930180848e-05,
          "validation_metric": 7.571142486995086e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.246060072000546
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0008544942811568035,
          "train_loss": 0.0006671838345937431,
          "validation_loss": 0.000866624410264194,
          "train_metric": 6.719630619045347e-05,
          "validation_metric": 8.728323155082762e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.2656008260000817
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0012931839573866455,
          "train_loss": 0.0024842543061822653,
          "validation_loss": 0.0011663241311907768,
          "train_metric": 0.0002502050483599305,
          "validation_metric": 0.00011746789095923305,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.285856254000464
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0017601039235159988,
          "train_loss": 0.0015265346737578511,
          "validation_loss": 0.0023588992189615965,
          "train_metric": 0.00015374697977676988,
          "validation_metric": 0.00023757970484439284,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.3061985799995455
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0012111575124436058,
          "train_loss": 0.0017914241179823875,
          "validation_loss": 0.0009404468000866473,
          "train_metric": 0.0001804256608011201,
          "validation_metric": 9.471836528973654e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.328603359999761
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0008294688359455904,
          "train_loss": 0.0006560354959219694,
          "validation_loss": 0.0006721991230733693,
          "train_metric": 6.607350223930553e-05,
          "validation_metric": 6.770143227186054e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.3551091569988785
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0007435656680172542,
          "train_loss": 0.0007309943903237581,
          "validation_loss": 0.000575352692976594,
          "train_metric": 7.36230649636127e-05,
          "validation_metric": 5.794741082354449e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.3769297340004414
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0007757867406326113,
          "train_loss": 0.0006920704036019742,
          "validation_loss": 0.0007874350412748754,
          "train_metric": 6.970280082896352e-05,
          "validation_metric": 7.93075596448034e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.3968174460005685
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0006069427254260518,
          "train_loss": 0.0005492178024724126,
          "validation_loss": 0.0007043098448775709,
          "train_metric": 5.5315205827355385e-05,
          "validation_metric": 7.093550084391609e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.4163777220001066
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0005738459785789018,
          "train_loss": 0.0004982063546776772,
          "validation_loss": 0.0006790204788558185,
          "train_metric": 5.017752482672222e-05,
          "validation_metric": 6.83884572936222e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.4361919179991673
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0007063046323310118,
          "train_loss": 0.0006497118738479912,
          "validation_loss": 0.0009879922727122903,
          "train_metric": 6.543660128954798e-05,
          "validation_metric": 9.950696403393522e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.4559955280001304
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0006429844124795636,
          "train_loss": 0.0004883087240159512,
          "validation_loss": 0.0006984137580730021,
          "train_metric": 4.9180664063896984e-05,
          "validation_metric": 7.034168083919212e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.47626994500024
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0007240740178531269,
          "train_loss": 0.001904166187159717,
          "validation_loss": 0.002510675461962819,
          "train_metric": 0.00019178066577296704,
          "validation_metric": 0.00025286604068242013,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.496090650000042
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0015301357743737753,
          "train_loss": 0.003679818706586957,
          "validation_loss": 0.0018083740724250674,
          "train_metric": 0.00037061789771541953,
          "validation_metric": 0.00018213280418422073,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.5159358469991275
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.003264478946221061,
          "train_loss": 0.00383895356208086,
          "validation_loss": 0.0045320806093513966,
          "train_metric": 0.0003866453771479428,
          "validation_metric": 0.00045645463978871703,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.5356862179996824
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.002526641696022125,
          "train_loss": 0.000792948470916599,
          "validation_loss": 0.0009679279173724353,
          "train_metric": 7.986286072991788e-05,
          "validation_metric": 9.748615411808714e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.556591373999254
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0021575057471636683,
          "train_loss": 0.001008152379654348,
          "validation_loss": 0.0011618819553405046,
          "train_metric": 0.00010153740004170686,
          "validation_metric": 0.00011702049960149452,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.576505516999532
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0014950085151212988,
          "train_loss": 0.000763461459428072,
          "validation_loss": 0.0008542913710698485,
          "train_metric": 7.689304038649425e-05,
          "validation_metric": 8.604110189480707e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.59605170299983
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0012214869075251045,
          "train_loss": 0.0006441519944928586,
          "validation_loss": 0.0005257397424429655,
          "train_metric": 6.48766290396452e-05,
          "validation_metric": 5.29505850863643e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.617097491000095
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0008626706667200779,
          "train_loss": 0.0005961322458460927,
          "validation_loss": 0.0006439675344154239,
          "train_metric": 6.0040259995730594e-05,
          "validation_metric": 6.485805351985618e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.6367010489993845
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.000646810355647176,
          "train_loss": 0.000625841727014631,
          "validation_loss": 0.0007790563395246863,
          "train_metric": 6.30324866506271e-05,
          "validation_metric": 7.846370863262564e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.65559892499914
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0006690470727335196,
          "train_loss": 0.000613735115621239,
          "validation_loss": 0.0008658351143822074,
          "train_metric": 6.18131598457694e-05,
          "validation_metric": 8.720374171389267e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.6746175850003056
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0007130144531402038,
          "train_loss": 0.0007182543049566448,
          "validation_loss": 0.0005261166952550411,
          "train_metric": 7.233994256239384e-05,
          "validation_metric": 5.29885473952163e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.6940711269999156
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.000600367144215852,
          "train_loss": 0.00051941501442343,
          "validation_loss": 0.0005939400289207697,
          "train_metric": 5.231357499724254e-05,
          "validation_metric": 5.9819463785970584e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7130113279999932
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0005342835211195052,
          "train_loss": 0.0004931337316520512,
          "validation_loss": 0.0006522658513858914,
          "train_metric": 4.9666625272948295e-05,
          "validation_metric": 6.56938282190822e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7313610309993237
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0005169023779671988,
          "train_loss": 0.0004955320036970079,
          "validation_loss": 0.0007390070823021233,
          "train_metric": 4.990816523786634e-05,
          "validation_metric": 7.443008507834747e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.7499604919994454
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0004734117192128906,
          "train_loss": 0.0007573200855404139,
          "validation_loss": 0.00053916044998914,
          "train_metric": 7.627451122971252e-05,
          "validation_metric": 5.430226883618161e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.768907757999841
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0006135230469226371,
          "train_loss": 0.0005529584013856947,
          "validation_loss": 0.0005358633934520185,
          "train_metric": 5.5691940360702574e-05,
          "validation_metric": 5.3970197768649086e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.787959258999763
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0005109049379825592,
          "train_loss": 0.00047132742474786937,
          "validation_loss": 0.0007435277220793068,
          "train_metric": 4.7470370191149414e-05,
          "validation_metric": 7.488538540201262e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.8070443960004923
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.00048239194074994884,
          "train_loss": 0.00047947673010639846,
          "validation_loss": 0.0008232196560129523,
          "train_metric": 4.829114186577499e-05,
          "validation_metric": 8.291167614515871e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.8264783329996135
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.00044641327167482814,
          "train_loss": 0.0004859869659412652,
          "validation_loss": 0.000530469580553472,
          "train_metric": 4.894682933809236e-05,
          "validation_metric": 5.342695294530131e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8456817860005685
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0005176022777959588,
          "train_loss": 0.0005142988520674407,
          "validation_loss": 0.0005274429568089545,
          "train_metric": 5.1798298954963684e-05,
          "validation_metric": 5.312211578711867e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8651069509996887
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0004966176556990831,
          "train_loss": 0.0004703609738498926,
          "validation_loss": 0.0005429123411886394,
          "train_metric": 4.737303970614448e-05,
          "validation_metric": 5.468014569487423e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8842265000002953
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.000752310117604793,
          "train_loss": 0.0006379257538355887,
          "validation_loss": 0.0004939570790156722,
          "train_metric": 6.424954335670918e-05,
          "validation_metric": 4.974955663783476e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9041401199992833
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0005809562844660832,
          "train_loss": 0.00046903890324756503,
          "validation_loss": 0.0005621277377940714,
          "train_metric": 4.7239878767868504e-05,
          "validation_metric": 5.661544855684042e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9234722090004652
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.00045920490265416447,
          "train_loss": 0.00047340610763058066,
          "validation_loss": 0.0008113561198115349,
          "train_metric": 4.7679732233518735e-05,
          "validation_metric": 8.171681110979989e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9428122579993214
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.000452362236501358,
          "train_loss": 0.00042168100480921566,
          "validation_loss": 0.0006282172398641706,
          "train_metric": 4.247016477165744e-05,
          "validation_metric": 6.32717419648543e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9613987150005414
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.00044402642652130453,
          "train_loss": 0.0004162231634836644,
          "validation_loss": 0.0006871765945106745,
          "train_metric": 4.19204770878423e-05,
          "validation_metric": 6.920992018422112e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9797868959994958
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.000433572430665663,
          "train_loss": 0.00041370055987499654,
          "validation_loss": 0.0006505486671812832,
          "train_metric": 4.166640428593382e-05,
          "validation_metric": 6.552088598255068e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.998357100999783
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0004768107792187948,
          "train_loss": 0.0004640541737899184,
          "validation_loss": 0.0007988137658685446,
          "train_metric": 4.673783405451104e-05,
          "validation_metric": 8.045360300457105e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.017260622998947
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.00046351950823009247,
          "train_loss": 0.0004270866047590971,
          "validation_loss": 0.0007572005852125585,
          "train_metric": 4.3014595576096326e-05,
          "validation_metric": 7.626246951986104e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.036572876999344
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0004459673200472025,
          "train_loss": 0.0004063142987433821,
          "validation_loss": 0.0006371114868670702,
          "train_metric": 4.092249582754448e-05,
          "validation_metric": 6.41675287624821e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.055742952999935
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.00044859246463602176,
          "train_loss": 0.00044764019548892975,
          "validation_loss": 0.0005189859075471759,
          "train_metric": 4.508468191488646e-05,
          "validation_metric": 5.227035580901429e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.075233710998873
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0004957019118592143,
          "train_loss": 0.0006668452988378704,
          "validation_loss": 0.0004790916864294559,
          "train_metric": 6.71622110530734e-05,
          "validation_metric": 4.825236101169139e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.095315271999425
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0005060087778474553,
          "train_loss": 0.0004062271327711642,
          "validation_loss": 0.0006157740135677159,
          "train_metric": 4.0913710108725354e-05,
          "validation_metric": 6.201850192155689e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.114070481000454
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0004384267749628634,
          "train_loss": 0.00040205553523264825,
          "validation_loss": 0.0006327427108772099,
          "train_metric": 4.049355993629433e-05,
          "validation_metric": 6.372752977767959e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.132517738000388
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0004279835611669114,
          "train_loss": 0.00041581378900446,
          "validation_loss": 0.0005637601134367287,
          "train_metric": 4.1879240598063916e-05,
          "validation_metric": 5.67798524571117e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.151234868999381
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.00044815714409196516,
          "train_loss": 0.000393998168874532,
          "validation_loss": 0.0007031454588286579,
          "train_metric": 3.968205419369042e-05,
          "validation_metric": 7.08182342350483e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.169968449999942
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.00047776594146853313,
          "train_loss": 0.0004571430617943406,
          "validation_loss": 0.0008958046091720462,
          "train_metric": 4.6041775931371376e-05,
          "validation_metric": 9.022215817822143e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.189093212000444
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.00043478821316966787,
          "train_loss": 0.00040160538628697395,
          "validation_loss": 0.0006386403692886233,
          "train_metric": 4.0448223444400355e-05,
          "validation_metric": 6.432151712942868e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.208241140999235
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0004162772456766106,
          "train_loss": 0.0003855069517157972,
          "validation_loss": 0.0006169043481349945,
          "train_metric": 3.8826845411676913e-05,
          "validation_metric": 6.213233427843079e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.226905346999047
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0003976148182118777,
          "train_loss": 0.0003854965034406632,
          "validation_loss": 0.0005971852806396782,
          "train_metric": 3.882580131175928e-05,
          "validation_metric": 6.014631799189374e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.245675279998977
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.00040854942926671356,
          "train_loss": 0.0003835650277324021,
          "validation_loss": 0.0006442789454013109,
          "train_metric": 3.8631264033028856e-05,
          "validation_metric": 6.488942017313093e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.264376740999069
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.00039182295950013213,
          "train_loss": 0.0003798771067522466,
          "validation_loss": 0.0005945026641711593,
          "train_metric": 3.8259833672782406e-05,
          "validation_metric": 5.987613621982746e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.282957703999273
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.000391454248529044,
          "train_loss": 0.0003750375472009182,
          "validation_loss": 0.0006555180880241096,
          "train_metric": 3.777240635827184e-05,
          "validation_metric": 6.602138455491513e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.301746200999332
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0003851742321785423,
          "train_loss": 0.00039932093932293355,
          "validation_loss": 0.0005469305324368179,
          "train_metric": 4.021813947474584e-05,
          "validation_metric": 5.508484900929034e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.320470912000019
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0004595452783178189,
          "train_loss": 0.00037481391336768866,
          "validation_loss": 0.0005892112385481596,
          "train_metric": 3.774988726945594e-05,
          "validation_metric": 5.934319779044017e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.339567242999692
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0004875089935012511,
          "train_loss": 0.0003749260795302689,
          "validation_loss": 0.000701985671184957,
          "train_metric": 3.776117955567315e-05,
          "validation_metric": 7.07014260115102e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.358559215999776
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.00042576905707392143,
          "train_loss": 0.0003955374413635582,
          "validation_loss": 0.0007156161009334028,
          "train_metric": 3.983708666055463e-05,
          "validation_metric": 7.20742391422391e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.377689441000257
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.000377603267224913,
          "train_loss": 0.00038518020301125944,
          "validation_loss": 0.0005663146148435771,
          "train_metric": 3.879393989336677e-05,
          "validation_metric": 5.703712668037042e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.396055750999949
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0004180894884484587,
          "train_loss": 0.0003861081204377115,
          "validation_loss": 0.0005551600479520857,
          "train_metric": 3.8887399568920955e-05,
          "validation_metric": 5.591368972091004e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.414448649000406
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.00040306566552317236,
          "train_loss": 0.0003873966634273529,
          "validation_loss": 0.0005634764675050974,
          "train_metric": 3.901717354892753e-05,
          "validation_metric": 5.675128704751842e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.433663051999247
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0004076540826645214,
          "train_loss": 0.000380915094865486,
          "validation_loss": 0.0005472126649692655,
          "train_metric": 3.8364374631782994e-05,
          "validation_metric": 5.511325798579492e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.452735364999171
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.0003861696422973182,
          "train_loss": 0.0003668606514111161,
          "validation_loss": 0.0006863857270218432,
          "train_metric": 3.694885890581645e-05,
          "validation_metric": 6.913025572430342e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.470742657998926
        }
      ],
      "best": {
        "epoch": 162,
        "updates": 2592
      },
      "best_validation_metric": 4.825236101169139e-05,
      "config": {
        "label": "LayerNorm + tanh + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.006946391941986242
    },
    "13e54c648cc2_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.17469824547879398,
          "train_loss": 0.08774837106466293,
          "validation_loss": 0.086602583527565,
          "train_metric": 0.008837694302201271,
          "validation_metric": 0.008722295053303242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.015357198999481625
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07360453333240002,
          "train_loss": 0.06181218847632408,
          "validation_loss": 0.06474826484918594,
          "train_metric": 0.0062254974618554115,
          "validation_metric": 0.006521208211779594,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.030770989000302507
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.056057537789456546,
          "train_loss": 0.04291826859116554,
          "validation_loss": 0.046800144016742706,
          "train_metric": 0.004322570748627186,
          "validation_metric": 0.004713539034128189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.045901356999820564
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.038856737315654755,
          "train_loss": 0.03218131884932518,
          "validation_loss": 0.0270075686275959,
          "train_metric": 0.0032411853317171335,
          "validation_metric": 0.0027201033663004637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0612920990006387
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.02863936615176499,
          "train_loss": 0.023211568593978882,
          "validation_loss": 0.02074548229575157,
          "train_metric": 0.0023377842735499144,
          "validation_metric": 0.002089408691972494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07610438600022462
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.020648730802349746,
          "train_loss": 0.021016666665673256,
          "validation_loss": 0.016835160553455353,
          "train_metric": 0.0021167215891182423,
          "validation_metric": 0.0016955757746472955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09143947499978822
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.017860822612419724,
          "train_loss": 0.015319674275815487,
          "validation_loss": 0.016857700422406197,
          "train_metric": 0.0015429414343088865,
          "validation_metric": 0.0016978459898382425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10567971199998283
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.01622412708820775,
          "train_loss": 0.019327949732542038,
          "validation_loss": 0.01822485774755478,
          "train_metric": 0.0019466402009129524,
          "validation_metric": 0.0018355409847572446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12013948099956906
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.015191223297733814,
          "train_loss": 0.00894016120582819,
          "validation_loss": 0.0067612831480801105,
          "train_metric": 0.0009004203020595014,
          "validation_metric": 0.0006809717160649598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13508915999955207
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.01109477988211438,
          "train_loss": 0.008814625442028046,
          "validation_loss": 0.00746913580223918,
          "train_metric": 0.0008877768414095044,
          "validation_metric": 0.0007522639352828264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14908399500018277
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.018847736122552305,
          "train_loss": 0.02247755043208599,
          "validation_loss": 0.015438548289239407,
          "train_metric": 0.0022638565860688686,
          "validation_metric": 0.0015549140516668558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16355996700076503
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.018711894837906584,
          "train_loss": 0.012158442288637161,
          "validation_loss": 0.008228214457631111,
          "train_metric": 0.0012245536781847477,
          "validation_metric": 0.000828715565148741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17768667400014238
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.010454037867020816,
          "train_loss": 0.008276775479316711,
          "validation_loss": 0.005153849255293608,
          "train_metric": 0.000833606522064656,
          "validation_metric": 0.0005190768279135227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19290304199967068
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.008383314532693475,
          "train_loss": 0.009224709123373032,
          "validation_loss": 0.010711843147873878,
          "train_metric": 0.000929078902117908,
          "validation_metric": 0.0010788575746119022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2070370120000007
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.011629403757979162,
          "train_loss": 0.012009760364890099,
          "validation_loss": 0.009356722235679626,
          "train_metric": 0.001209579175338149,
          "validation_metric": 0.0009423746960237622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22058513599949947
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.01043891730660107,
          "train_loss": 0.006014615297317505,
          "validation_loss": 0.0039059207774698734,
          "train_metric": 0.0006057699793018401,
          "validation_metric": 0.0003933900152333081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2349585499996465
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.00892931772978045,
          "train_loss": 0.008275914937257767,
          "validation_loss": 0.0037234548944979906,
          "train_metric": 0.000833519792649895,
          "validation_metric": 0.0003750127216335386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24936869100019976
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.016686872142599896,
          "train_loss": 0.0080908527597785,
          "validation_loss": 0.008409524336457253,
          "train_metric": 0.0008148810593411326,
          "validation_metric": 0.000846976472530514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26317926900082966
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.007357539521763101,
          "train_loss": 0.0050854990258812904,
          "validation_loss": 0.004172328859567642,
          "train_metric": 0.0005121927824802697,
          "validation_metric": 0.0004202217096462846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2766353179995349
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.007062774151563644,
          "train_loss": 0.013032539747655392,
          "validation_loss": 0.010284306481480598,
          "train_metric": 0.0013125897385179996,
          "validation_metric": 0.0010357977589592338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2899900680004066
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.013464375690091401,
          "train_loss": 0.01278369128704071,
          "validation_loss": 0.010428732261061668,
          "train_metric": 0.0012875264510512352,
          "validation_metric": 0.0010503436205908656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3037308090006263
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.007588754087919369,
          "train_loss": 0.005317991133779287,
          "validation_loss": 0.00449811015278101,
          "train_metric": 0.0005356085603125393,
          "validation_metric": 0.00045303325168788433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3171723010000278
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.005772954718850087,
          "train_loss": 0.004529461730271578,
          "validation_loss": 0.00530380941927433,
          "train_metric": 0.0004561908426694572,
          "validation_metric": 0.0005341802607290447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33099209099964355
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.003951733167923521,
          "train_loss": 0.0032853365410119295,
          "validation_loss": 0.0028877228032797575,
          "train_metric": 0.0003308870655018836,
          "validation_metric": 0.0002908408350776881,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34589341200080526
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.005694857194612268,
          "train_loss": 0.012830698862671852,
          "validation_loss": 0.011499912478029728,
          "train_metric": 0.001292260829359293,
          "validation_metric": 0.001158229075372219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35943758100074774
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.011381016200175509,
          "train_loss": 0.016322482377290726,
          "validation_loss": 0.01610724814236164,
          "train_metric": 0.001643940806388855,
          "validation_metric": 0.0016222628764808178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3727275279998139
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.008822661446174607,
          "train_loss": 0.008458982221782207,
          "validation_loss": 0.006830762606114149,
          "train_metric": 0.0008519577095285058,
          "validation_metric": 0.0006879693828523159,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.386458066999694
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.0055612625073990785,
          "train_loss": 0.003302248427644372,
          "validation_loss": 0.003444814821705222,
          "train_metric": 0.00033259036717936397,
          "validation_metric": 0.00034694914938881993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4014946190000046
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.007565966276160907,
          "train_loss": 0.016289830207824707,
          "validation_loss": 0.011221380904316902,
          "train_metric": 0.001640652073547244,
          "validation_metric": 0.0011301763588562608,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4149857949996658
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.013868523805285804,
          "train_loss": 0.01236303336918354,
          "validation_loss": 0.007309946231544018,
          "train_metric": 0.0012451594229787588,
          "validation_metric": 0.0007362310425378382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4291912589997082
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.008568004821427166,
          "train_loss": 0.007436853833496571,
          "validation_loss": 0.006095121614634991,
          "train_metric": 0.0007490128045901656,
          "validation_metric": 0.000613878364674747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4431290970005648
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.011385732956114225,
          "train_loss": 0.006194804795086384,
          "validation_loss": 0.007034803740680218,
          "train_metric": 0.0006239180220291018,
          "validation_metric": 0.0007085197139531374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4566224530008185
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.005738529376685619,
          "train_loss": 0.00688930694013834,
          "validation_loss": 0.007670516613870859,
          "train_metric": 0.0006938657606951892,
          "validation_metric": 0.0007725464529357851,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5511111510004412
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.005401998438173905,
          "train_loss": 0.00349409319460392,
          "validation_loss": 0.002726582810282707,
          "train_metric": 0.0003519122547004372,
          "validation_metric": 0.00027461143326945603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.565957963000983
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.005597987765213475,
          "train_loss": 0.005804476328194141,
          "validation_loss": 0.005776972509920597,
          "train_metric": 0.0005846054991707206,
          "validation_metric": 0.0005818354547955096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5799000040005922
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.006143338585388847,
          "train_loss": 0.01148177683353424,
          "validation_loss": 0.009007774293422699,
          "train_metric": 0.0011564027518033981,
          "validation_metric": 0.0009072301327250898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5935449640001025
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.009111446946917567,
          "train_loss": 0.009930150583386421,
          "validation_loss": 0.010570409707725048,
          "train_metric": 0.0010001285700127482,
          "validation_metric": 0.0010646129958331585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.606934524999815
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.02144408703316003,
          "train_loss": 0.01571727730333805,
          "validation_loss": 0.009488346055150032,
          "train_metric": 0.0015829865587875247,
          "validation_metric": 0.0009556313161738217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6205023599995911
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.01252610661322251,
          "train_loss": 0.00984510499984026,
          "validation_loss": 0.010160131379961967,
          "train_metric": 0.000991562963463366,
          "validation_metric": 0.0010232911445200443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6340218959994672
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.00745236522925552,
          "train_loss": 0.005789437796920538,
          "validation_loss": 0.005320598836988211,
          "train_metric": 0.0005830909358337522,
          "validation_metric": 0.0005358712514862418,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6496321630002058
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.004231854632962495,
          "train_loss": 0.004675413481891155,
          "validation_loss": 0.004513945430517197,
          "train_metric": 0.00047089048894122243,
          "validation_metric": 0.0004546281124930829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6742723810002644
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.007189390700659715,
          "train_loss": 0.004233115818351507,
          "validation_loss": 0.00276603433303535,
          "train_metric": 0.00042634387500584126,
          "validation_metric": 0.0002785848337225616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.69842082499963
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.007435428553435486,
          "train_loss": 0.005871464032679796,
          "validation_loss": 0.003420013003051281,
          "train_metric": 0.0005913523491472006,
          "validation_metric": 0.0003444512258283794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7202783200009435
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.009470527118537575,
          "train_loss": 0.004352807067334652,
          "validation_loss": 0.0026026819832623005,
          "train_metric": 0.00043839873978868127,
          "validation_metric": 0.0002621326129883528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.743408358999659
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.005118723045598017,
          "train_loss": 0.00682386988773942,
          "validation_loss": 0.0061017912812530994,
          "train_metric": 0.000687275140080601,
          "validation_metric": 0.0006145501392893493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7648687889995927
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.005086616001790389,
          "train_loss": 0.005078463815152645,
          "validation_loss": 0.004163194447755814,
          "train_metric": 0.0005114842788316309,
          "validation_metric": 0.00041930173756554723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7871197700005723
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.0032088879524962977,
          "train_loss": 0.00489709060639143,
          "validation_loss": 0.0024873053189367056,
          "train_metric": 0.0004932170268148184,
          "validation_metric": 0.0002505122683942318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8104971930006286
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0029426619221339934,
          "train_loss": 0.0027947891503572464,
          "validation_loss": 0.002015306381508708,
          "train_metric": 0.00028148089768365026,
          "validation_metric": 0.00020297427545301616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8339955600004032
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.004175201407633722,
          "train_loss": 0.020366378128528595,
          "validation_loss": 0.015269857831299305,
          "train_metric": 0.0020512270275503397,
          "validation_metric": 0.0015379241667687893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8559950580001896
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.017829494070610963,
          "train_loss": 0.00337790884077549,
          "validation_loss": 0.0019126050174236298,
          "train_metric": 0.00034021062310785055,
          "validation_metric": 0.00019263055582996458,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8791505730005156
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.007706464035436511,
          "train_loss": 0.004119453486055136,
          "validation_loss": 0.004866297822445631,
          "train_metric": 0.000414896261645481,
          "validation_metric": 0.0004901156644336879,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8998069549998036
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.003865843806124758,
          "train_loss": 0.0030894684605300426,
          "validation_loss": 0.0026276682037860155,
          "train_metric": 0.0003111599071417004,
          "validation_metric": 0.00026464913389645517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9194662330010033
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.003945239663153188,
          "train_loss": 0.0032585756853222847,
          "validation_loss": 0.0023356659803539515,
          "train_metric": 0.0003281918470747769,
          "validation_metric": 0.000235239727771841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.939244339000652
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.005527104272914585,
          "train_loss": 0.005947537254542112,
          "validation_loss": 0.004396939184516668,
          "train_metric": 0.0005990142817609012,
          "validation_metric": 0.00044284359319135547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9605372900005023
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.006053970202628989,
          "train_loss": 0.0036551710218191147,
          "validation_loss": 0.002675779862329364,
          "train_metric": 0.0003681354282889515,
          "validation_metric": 0.00026949477614834905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9811570229994686
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.003581159413442947,
          "train_loss": 0.0024149450473487377,
          "validation_loss": 0.0019068822730332613,
          "train_metric": 0.0002432243782095611,
          "validation_metric": 0.00019205421267542988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0022422279998864
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.008579878223827109,
          "train_loss": 0.007476395927369595,
          "validation_loss": 0.008935678750276566,
          "train_metric": 0.0007529952563345432,
          "validation_metric": 0.0008999688434414566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.023387264000121
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.004014553069282556,
          "train_loss": 0.003338887356221676,
          "validation_loss": 0.0019609611481428146,
          "train_metric": 0.0003362804709468037,
          "validation_metric": 0.00019750080537050962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.043724610000936
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.005568284024775494,
          "train_loss": 0.009784785099327564,
          "validation_loss": 0.009299298748373985,
          "train_metric": 0.0009854877134785056,
          "validation_metric": 0.000936591241043061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0638024050003878
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.008419338599196635,
          "train_loss": 0.006721618585288525,
          "validation_loss": 0.004362899344414473,
          "train_metric": 0.0006769767496734858,
          "validation_metric": 0.00043941522017121315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0842144079997524
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.004889800802629907,
          "train_loss": 0.0033947343472391367,
          "validation_loss": 0.0016815293347463012,
          "train_metric": 0.00034190522274002433,
          "validation_metric": 0.0001693574886303395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.106549358999473
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.005923350501689129,
          "train_loss": 0.012823661789298058,
          "validation_loss": 0.01178542710840702,
          "train_metric": 0.0012915523257106543,
          "validation_metric": 0.001186985056847334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1284040570008074
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.007817275763954967,
          "train_loss": 0.0061249337159097195,
          "validation_loss": 0.006626131013035774,
          "train_metric": 0.0006168808904476464,
          "validation_metric": 0.0006673596799373627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1492247730002418
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0035805948573397473,
          "train_loss": 0.002339717233553529,
          "validation_loss": 0.0020089170429855585,
          "train_metric": 0.00023564770526718348,
          "validation_metric": 0.00020233076065778732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.173681354999644
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.0028934544207004365,
          "train_loss": 0.0026947136502712965,
          "validation_loss": 0.002496195025742054,
          "train_metric": 0.00027140165911987424,
          "validation_metric": 0.0002514075895305723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.193781754000156
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.0026622053919709288,
          "train_loss": 0.002684344770386815,
          "validation_loss": 0.0018297166097909212,
          "train_metric": 0.00027035741368308663,
          "validation_metric": 0.00018428235489409417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2138259660005133
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0030751192825846374,
          "train_loss": 0.0015367658343166113,
          "validation_loss": 0.0007716359687037766,
          "train_metric": 0.00015477740089409053,
          "validation_metric": 7.771634409436956e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.234182898000654
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0028058795105607715,
          "train_loss": 0.0019020121544599533,
          "validation_loss": 0.0013743768213316798,
          "train_metric": 0.0001915636967169121,
          "validation_metric": 0.0001384222268825397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.254278771000827
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.002211427541624289,
          "train_loss": 0.0022421651519834995,
          "validation_loss": 0.0025325354654341936,
          "train_metric": 0.00022582264500670135,
          "validation_metric": 0.0002550677163526416,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3924317920009344
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.004392391329020029,
          "train_loss": 0.011406540870666504,
          "validation_loss": 0.007262186612933874,
          "train_metric": 0.0011488250456750393,
          "validation_metric": 0.0007314208196476102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4160545270005969
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.011072731198510155,
          "train_loss": 0.003727383678779006,
          "validation_loss": 0.002349167363718152,
          "train_metric": 0.0003754084464162588,
          "validation_metric": 0.00023659953149035573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4364510020004673
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.00408680804684991,
          "train_loss": 0.005674190353602171,
          "validation_loss": 0.004940127022564411,
          "train_metric": 0.0005714837461709976,
          "validation_metric": 0.0004975515184924006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4565124189994094
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.004902782951830886,
          "train_loss": 0.005039860028773546,
          "validation_loss": 0.004312317352741957,
          "train_metric": 0.0005075962981209159,
          "validation_metric": 0.00043432082748040557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.476896554000632
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0032235977851087227,
          "train_loss": 0.008413396775722504,
          "validation_loss": 0.005775026511400938,
          "train_metric": 0.0008473665802739561,
          "validation_metric": 0.0005816395278088748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4969832010010578
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.00586096119513968,
          "train_loss": 0.002768790815025568,
          "validation_loss": 0.002736549824476242,
          "train_metric": 0.00027886248426511884,
          "validation_metric": 0.00027561525348573923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5163803140003438
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.003293814654171001,
          "train_loss": 0.004184024874120951,
          "validation_loss": 0.002594922436401248,
          "train_metric": 0.00042139962897635996,
          "validation_metric": 0.0002613510878290981,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.536302505999629
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.009261588580557145,
          "train_loss": 0.009118457324802876,
          "validation_loss": 0.008033856749534607,
          "train_metric": 0.0009183775982819498,
          "validation_metric": 0.0008091406198218465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5562264060008602
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.007043960300507024,
          "train_loss": 0.00428296672180295,
          "validation_loss": 0.003561281831935048,
          "train_metric": 0.00043136466410942376,
          "validation_metric": 0.0003586792736314237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5765107309998712
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.0025962845465983264,
          "train_loss": 0.002497623907402158,
          "validation_loss": 0.001775078708305955,
          "train_metric": 0.0002515515370760113,
          "validation_metric": 0.00017877943173516542,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.5969737340001302
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.002939523648819886,
          "train_loss": 0.002795238047838211,
          "validation_loss": 0.0017248245421797037,
          "train_metric": 0.00028152612503618,
          "validation_metric": 0.0001737179991323501,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.616981520999616
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.004058317510498455,
          "train_loss": 0.004059186205267906,
          "validation_loss": 0.0038835632149130106,
          "train_metric": 0.00040882633766159415,
          "validation_metric": 0.0003911382518708706,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6367170970006555
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.005290762725053355,
          "train_loss": 0.004099313635379076,
          "validation_loss": 0.003008968662470579,
          "train_metric": 0.0004128678410779685,
          "validation_metric": 0.0003030523075722158,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.656576485000187
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0024910525244195014,
          "train_loss": 0.0022627387661486864,
          "validation_loss": 0.001937519060447812,
          "train_metric": 0.00022789476497564465,
          "validation_metric": 0.00019513981533236802,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.6768752640000457
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0017443027754779905,
          "train_loss": 0.0016951101133599877,
          "validation_loss": 0.0014761193888261914,
          "train_metric": 0.00017072529590222985,
          "validation_metric": 0.00014866937999613583,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.69743323900002
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.001857054188803886,
          "train_loss": 0.0011487056035548449,
          "validation_loss": 0.0007523156236857176,
          "train_metric": 0.00011569341586437076,
          "validation_metric": 7.577047654194757e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7197391919999063
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.001402487305313116,
          "train_loss": 0.0011886422289535403,
          "validation_loss": 0.0008236111607402563,
          "train_metric": 0.00011971570347668603,
          "validation_metric": 8.295109728351235e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7449317530008557
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0015786551157361828,
          "train_loss": 0.0021511325612664223,
          "validation_loss": 0.0015478070126846433,
          "train_metric": 0.0002166541962651536,
          "validation_metric": 0.00015588945825584233,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7649129410001478
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.001941722332048812,
          "train_loss": 0.0024398015812039375,
          "validation_loss": 0.0019073360599577427,
          "train_metric": 0.00024572788970544934,
          "validation_metric": 0.00019209990568924695,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.7854277490005188
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.003177905127813574,
          "train_loss": 0.00643165735527873,
          "validation_loss": 0.003862375859171152,
          "train_metric": 0.0006477730348706245,
          "validation_metric": 0.0003890043299179524,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.805944615000044
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0035735715173359495,
          "train_loss": 0.0011286077788099647,
          "validation_loss": 0.0011948838364332914,
          "train_metric": 0.00011366925173206255,
          "validation_metric": 0.00012034432438667864,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8259374080007547
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0020342722564237192,
          "train_loss": 0.001761940191499889,
          "validation_loss": 0.0016197413206100464,
          "train_metric": 0.00017745618242770433,
          "validation_metric": 0.00016313442029058933,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.845975804999398
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0014335668347484898,
          "train_loss": 0.001138987485319376,
          "validation_loss": 0.0010189765598624945,
          "train_metric": 0.00011471466859802604,
          "validation_metric": 0.00010262760770274326,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8663138770007208
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.001560866330692079,
          "train_loss": 0.0043457625433802605,
          "validation_loss": 0.003135166596621275,
          "train_metric": 0.000437689246609807,
          "validation_metric": 0.00031576253240928054,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.8859784420001233
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.004655667151382659,
          "train_loss": 0.0012632639845833182,
          "validation_loss": 0.0010986779816448689,
          "train_metric": 0.00012723132385872304,
          "validation_metric": 0.00011065483704442158,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9062181190001866
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0030838023085379973,
          "train_loss": 0.002098455559462309,
          "validation_loss": 0.0011456431820988655,
          "train_metric": 0.000211348757147789,
          "validation_metric": 0.00011538499529706314,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9269988060004835
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.002094716670399066,
          "train_loss": 0.0014379899948835373,
          "validation_loss": 0.0007574636838398874,
          "train_metric": 0.00014482908591162413,
          "validation_metric": 7.628896128153428e-05,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 1.9474414819997037
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0013513256453734357,
          "train_loss": 0.0010369188385084271,
          "validation_loss": 0.0007908647530712187,
          "train_metric": 0.00010443466453580186,
          "validation_metric": 7.965300028445199e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9682530259997293
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.001095116882424918,
          "train_loss": 0.0007944180397316813,
          "validation_loss": 0.0006238891510292888,
          "train_metric": 8.001088281162083e-05,
          "validation_metric": 6.283583934418857e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 1.9901270890004525
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0011290007259958657,
          "train_loss": 0.0012237364426255226,
          "validation_loss": 0.0010847172234207392,
          "train_metric": 0.00012325025454629213,
          "validation_metric": 0.00010924875095952302,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.010630086000674
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.0008900403536244994,
          "train_loss": 0.0013807319337502122,
          "validation_loss": 0.0008232604013755918,
          "train_metric": 0.00013906226377002895,
          "validation_metric": 8.291576523333788e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0309562550010014
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.001211605502248858,
          "train_loss": 0.001294093206524849,
          "validation_loss": 0.0010900251800194383,
          "train_metric": 0.0001303363242186606,
          "validation_metric": 0.00010978333011735231,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.050768875000358
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0010931178521786933,
          "train_loss": 0.000940984464250505,
          "validation_loss": 0.0006058767903596163,
          "train_metric": 9.477249841438606e-05,
          "validation_metric": 6.1021692090434954e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.071646018001047
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0008931586326070828,
          "train_loss": 0.0013376802671700716,
          "validation_loss": 0.0008782227523624897,
          "train_metric": 0.0001347262441413477,
          "validation_metric": 8.845137926982716e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.0922331780002423
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0016160012310137972,
          "train_loss": 0.0009686824632808566,
          "validation_loss": 0.0008224044577218592,
          "train_metric": 9.75621587713249e-05,
          "validation_metric": 8.282957423944026e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.1129447170005733
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0010367323138780193,
          "train_loss": 0.0006913921097293496,
          "validation_loss": 0.0004474239540286362,
          "train_metric": 6.963448686292395e-05,
          "validation_metric": 4.50629013357684e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.229529358999571
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0007500854408135638,
          "train_loss": 0.000868633680511266,
          "validation_loss": 0.0007868894026614726,
          "train_metric": 8.748561231186613e-05,
          "validation_metric": 7.92526188888587e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2497606720007752
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0008215767720685108,
          "train_loss": 0.0006633929442614317,
          "validation_loss": 0.0006112762494012713,
          "train_metric": 6.681450759060681e-05,
          "validation_metric": 6.156549352454022e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2690856660010468
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0010877830973186065,
          "train_loss": 0.001021894277073443,
          "validation_loss": 0.0006885760230943561,
          "train_metric": 0.00010292144725099206,
          "validation_metric": 6.935084820725024e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.2891610099995887
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.000779628562668222,
          "train_loss": 0.0006250169826671481,
          "validation_loss": 0.0005550393834710121,
          "train_metric": 6.294942431850359e-05,
          "validation_metric": 5.5901538871694356e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3092218620004132
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.0009772868015716085,
          "train_loss": 0.001907695084810257,
          "validation_loss": 0.0013171631144359708,
          "train_metric": 0.00019213606719858944,
          "validation_metric": 0.00013265984307508916,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3291512840005453
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.0011819609135272913,
          "train_loss": 0.0008636261918582022,
          "validation_loss": 0.0010471458081156015,
          "train_metric": 8.698126475792378e-05,
          "validation_metric": 0.00010546469275141135,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.3490030850007315
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.0007469284355465788,
          "train_loss": 0.0005976164247840643,
          "validation_loss": 0.0006085670902393758,
          "train_metric": 6.018972635501996e-05,
          "validation_metric": 6.129264511400834e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.368318883000029
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0008162979738699505,
          "train_loss": 0.0029434242751449347,
          "validation_loss": 0.0019129668362438679,
          "train_metric": 0.00029645091854035854,
          "validation_metric": 0.0001926670374814421,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.388445488999423
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.002174868724978296,
          "train_loss": 0.0008205909980461001,
          "validation_loss": 0.0005909703904762864,
          "train_metric": 8.264692587545142e-05,
          "validation_metric": 5.952036735834554e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.408260001000599
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0007719736349827144,
          "train_loss": 0.0005514742806553841,
          "validation_loss": 0.00038355254218913615,
          "train_metric": 5.55424667254556e-05,
          "validation_metric": 3.863000893034041e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.429230670999459
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0006062279835532536,
          "train_loss": 0.0006022661691531539,
          "validation_loss": 0.00045950140338391066,
          "train_metric": 6.065804336685687e-05,
          "validation_metric": 4.6279299567686394e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.4492435030006163
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.000542231704457663,
          "train_loss": 0.0006209922721609473,
          "validation_loss": 0.00044847186654806137,
          "train_metric": 6.25440661679022e-05,
          "validation_metric": 4.516845365287736e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.468510957000035
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0006059293082216755,
          "train_loss": 0.0005026155849918723,
          "validation_loss": 0.00032328953966498375,
          "train_metric": 5.0621601985767484e-05,
          "validation_metric": 3.256053969380446e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.489663477999784
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0005397186014306499,
          "train_loss": 0.0006309909513220191,
          "validation_loss": 0.0005278136814013124,
          "train_metric": 6.355109508149326e-05,
          "validation_metric": 5.315945963957347e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5098226070003875
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0006726056963088922,
          "train_loss": 0.000642479513771832,
          "validation_loss": 0.00053028593538329,
          "train_metric": 6.470818334491923e-05,
          "validation_metric": 5.340845382306725e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5298832839998795
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0007319213927985402,
          "train_loss": 0.0004635310615412891,
          "validation_loss": 0.00044655572855845094,
          "train_metric": 4.6685152483405545e-05,
          "validation_metric": 4.4975458877161145e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5505458299994643
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0005447097328215023,
          "train_loss": 0.0007038611802272499,
          "validation_loss": 0.0004404880746733397,
          "train_metric": 7.089032442308962e-05,
          "validation_metric": 4.4364343921188265e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5720432159996562
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0007445207793352893,
          "train_loss": 0.0012477472191676497,
          "validation_loss": 0.0007473511504940689,
          "train_metric": 0.00012566853547468781,
          "validation_metric": 7.527045818278566e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.5919189090000145
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0009264774726034375,
          "train_loss": 0.0005726512754336,
          "validation_loss": 0.0005750525742769241,
          "train_metric": 5.767533730249852e-05,
          "validation_metric": 5.7917190133593976e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.612421132000236
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0006934442926649353,
          "train_loss": 0.000571719603613019,
          "validation_loss": 0.0005966495373286307,
          "train_metric": 5.758150655310601e-05,
          "validation_metric": 6.009235221426934e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.633005834000869
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0009706980508781271,
          "train_loss": 0.0006871556397527456,
          "validation_loss": 0.00047371647087857127,
          "train_metric": 6.920780288055539e-05,
          "validation_metric": 4.771098610945046e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.653195398999742
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0011590366084419657,
          "train_loss": 0.0029417730402201414,
          "validation_loss": 0.0018700499786064029,
          "train_metric": 0.0002962845901492983,
          "validation_metric": 0.0001883445802377537,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.6813004790001287
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0010215065340162255,
          "train_loss": 0.0004436893796082586,
          "validation_loss": 0.00047860483755357563,
          "train_metric": 4.468677070690319e-05,
          "validation_metric": 4.820332833332941e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.704705534000823
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0007597426110805827,
          "train_loss": 0.0012930938974022865,
          "validation_loss": 0.00075758364982903,
          "train_metric": 0.00013023569772485644,
          "validation_metric": 7.630104664713144e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 2.7253402620008274
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0014409284849534743,
          "train_loss": 0.0008326545939780772,
          "validation_loss": 0.0005232553230598569,
          "train_metric": 8.386192348552868e-05,
          "validation_metric": 5.270035762805492e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.7452542430000904
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0008410859963987605,
          "train_loss": 0.00044763035839423537,
          "validation_loss": 0.00036889963666908443,
          "train_metric": 4.5083688746672124e-05,
          "validation_metric": 3.715422280947678e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.765605120001055
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0006775264728275943,
          "train_loss": 0.0004199701070319861,
          "validation_loss": 0.0002846120041795075,
          "train_metric": 4.229785190545954e-05,
          "validation_metric": 2.8665075660683215e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.786433159999433
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0004574044878609129,
          "train_loss": 0.00036819768138229847,
          "validation_loss": 0.000357998680556193,
          "train_metric": 3.7083525967318565e-05,
          "validation_metric": 3.605631354730576e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.806342656000197
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0005058065089542652,
          "train_loss": 0.00049494398990646,
          "validation_loss": 0.0003901071904692799,
          "train_metric": 4.98489462188445e-05,
          "validation_metric": 3.92901711165905e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8261822479998955
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0005542812368730665,
          "train_loss": 0.0006166774546727538,
          "validation_loss": 0.0003973704588133842,
          "train_metric": 6.210950232343748e-05,
          "validation_metric": 4.002169953309931e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8458818559993233
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0005785923458461184,
          "train_loss": 0.0010811140527948737,
          "validation_loss": 0.0009453119710087776,
          "train_metric": 0.00010888585529755801,
          "validation_metric": 9.520836465526372e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.865776239999832
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0005606236536550568,
          "train_loss": 0.0003532733826432377,
          "validation_loss": 0.00034398920251987875,
          "train_metric": 3.5580404073698446e-05,
          "validation_metric": 3.464533438091166e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.8852776959993207
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.00041916745976777747,
          "train_loss": 0.0010330474469810724,
          "validation_loss": 0.0005936893285252154,
          "train_metric": 0.00010404476779513061,
          "validation_metric": 5.9794219851028174e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9057572570000048
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0007013363892838242,
          "train_loss": 0.00048466818407177925,
          "validation_loss": 0.00032339696190319955,
          "train_metric": 4.881400673184544e-05,
          "validation_metric": 3.257135904277675e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.926099542999509
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0005970181937300367,
          "train_loss": 0.0006076542777009308,
          "validation_loss": 0.00045423064148053527,
          "train_metric": 6.12006988376379e-05,
          "validation_metric": 4.574844933813438e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 2.9468675259995507
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0004123433386666875,
          "train_loss": 0.00033692942815832794,
          "validation_loss": 0.0003230622678529471,
          "train_metric": 3.39342986990232e-05,
          "validation_metric": 3.253764589317143e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.097644761999618
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0003611987767726532,
          "train_loss": 0.00041353690903633833,
          "validation_loss": 0.00032307161018252373,
          "train_metric": 4.16499242419377e-05,
          "validation_metric": 3.2538588129682466e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.120643709000433
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.00046523534365405794,
          "train_loss": 0.0003870528016705066,
          "validation_loss": 0.0003788710746448487,
          "train_metric": 3.898253999068402e-05,
          "validation_metric": 3.8158512325026095e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.1445611780000036
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0003975089839514112,
          "train_loss": 0.0003905136836692691,
          "validation_loss": 0.000304902350762859,
          "train_metric": 3.93311092921067e-05,
          "validation_metric": 3.0708648409927264e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.1673753960003523
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.00034208051329187583,
          "train_loss": 0.00034549128031358123,
          "validation_loss": 0.0003495221899356693,
          "train_metric": 3.4796616091625765e-05,
          "validation_metric": 3.5202596336603165e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.189345835000495
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.00039940020542417187,
          "train_loss": 0.0003238556382711977,
          "validation_loss": 0.00028284097788855433,
          "train_metric": 3.261755045969039e-05,
          "validation_metric": 2.848670737876091e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.211608352999974
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0003202685666110483,
          "train_loss": 0.0003366815799381584,
          "validation_loss": 0.0003164767404086888,
          "train_metric": 3.3909338526427746e-05,
          "validation_metric": 3.187438051099889e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.232508950999545
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0003516687975206878,
          "train_loss": 0.0003372791688889265,
          "validation_loss": 0.000344881322234869,
          "train_metric": 3.396952524781227e-05,
          "validation_metric": 3.473518154351041e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.2539409620003426
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.00032194946379604517,
          "train_loss": 0.0003341204428579658,
          "validation_loss": 0.00027156894793733954,
          "train_metric": 3.365138763911091e-05,
          "validation_metric": 2.735142879828345e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.276416877999509
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0003372856772330124,
          "train_loss": 0.0003162481007166207,
          "validation_loss": 0.0003185129025951028,
          "train_metric": 3.185135210515e-05,
          "validation_metric": 3.207945337635465e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.297539348999635
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.00041427685300732264,
          "train_loss": 0.00043609377462416887,
          "validation_loss": 0.0003177968319505453,
          "train_metric": 4.392176924739033e-05,
          "validation_metric": 3.200733772246167e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.318815523000012
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.00043857970922545064,
          "train_loss": 0.00035489044967107475,
          "validation_loss": 0.0003153902944177389,
          "train_metric": 3.574326910893433e-05,
          "validation_metric": 3.176495374646038e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.340244700000767
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0003403350347070955,
          "train_loss": 0.0006319247768260539,
          "validation_loss": 0.0004305939073674381,
          "train_metric": 6.364515138557181e-05,
          "validation_metric": 4.336784331826493e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.360922293999465
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0004771759295181255,
          "train_loss": 0.0003767572052311152,
          "validation_loss": 0.0003382318827789277,
          "train_metric": 3.794560325331986e-05,
          "validation_metric": 3.406547330087051e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.382116586999473
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0003897949454767513,
          "train_loss": 0.00029803888173773885,
          "validation_loss": 0.0002824947005137801,
          "train_metric": 3.0017385142855346e-05,
          "validation_metric": 2.8451831894926727e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4049306390006677
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.00031561407377012074,
          "train_loss": 0.00030102761229500175,
          "validation_loss": 0.00029998813988640904,
          "train_metric": 3.0318396966322325e-05,
          "validation_metric": 3.021370866918005e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4247442800005956
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0004077597391187737,
          "train_loss": 0.0002956202661152929,
          "validation_loss": 0.0002916674711741507,
          "train_metric": 2.9773789719911292e-05,
          "validation_metric": 2.9375680242083035e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.444472785000471
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0003290874310550862,
          "train_loss": 0.0003096971195191145,
          "validation_loss": 0.00033477143733762205,
          "train_metric": 3.119156463071704e-05,
          "validation_metric": 3.3716951293172315e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4641098040010547
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.00034112261164409574,
          "train_loss": 0.00034735805820673704,
          "validation_loss": 0.0003510713577270508,
          "train_metric": 3.498463411233388e-05,
          "validation_metric": 3.535861833370291e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.4833738729994366
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0003436386487010168,
          "train_loss": 0.00029547844314947724,
          "validation_loss": 0.00027260236674919724,
          "train_metric": 2.9759503377135843e-05,
          "validation_metric": 2.7455513190943748e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 3.502988345000631
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.00034816664265235886,
          "train_loss": 0.00038528471486642957,
          "validation_loss": 0.00040015392005443573,
          "train_metric": 3.8804464566055685e-05,
          "validation_metric": 4.030203854199499e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.5231195779997506
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.00035195943110011285,
          "train_loss": 0.0002866552968043834,
          "validation_loss": 0.0002773437008727342,
          "train_metric": 2.8870870664832182e-05,
          "validation_metric": 2.7933043384109624e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.5428178229994955
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.0003086779438490339,
          "train_loss": 0.00034617824712768197,
          "validation_loss": 0.0002821616071742028,
          "train_metric": 3.4865803172579035e-05,
          "validation_metric": 2.841827881638892e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.5627277540006617
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.00030913383034203434,
          "train_loss": 0.0003310449537821114,
          "validation_loss": 0.00034287231392227113,
          "train_metric": 3.33416392095387e-05,
          "validation_metric": 3.453284807619639e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.582304859000942
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0003324223425806849,
          "train_loss": 0.00028501771157607436,
          "validation_loss": 0.00028542164363898337,
          "train_metric": 2.870593743864447e-05,
          "validation_metric": 2.8746624593622983e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.601708176000102
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.00032067055553852697,
          "train_loss": 0.00030026669264771044,
          "validation_loss": 0.0002694029244594276,
          "train_metric": 3.0241764761740342e-05,
          "validation_metric": 2.7133275580126792e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.6243750619996717
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.0002808113022183534,
          "train_loss": 0.00028073464636690915,
          "validation_loss": 0.0002894120698329061,
          "train_metric": 2.827456410159357e-05,
          "validation_metric": 2.914852484536823e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.6456267989997286
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0003172246752001229,
          "train_loss": 0.00039763428503647447,
          "validation_loss": 0.00029577952227555215,
          "train_metric": 4.0048271330306306e-05,
          "validation_metric": 2.9789831387461163e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.6645828489999985
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0003460499010543572,
          "train_loss": 0.00028934169677086174,
          "validation_loss": 0.00032511065364815295,
          "train_metric": 2.9141438062652014e-05,
          "validation_metric": 3.27439556713216e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.683041485999638
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0002822172064043116,
          "train_loss": 0.0002772154111880809,
          "validation_loss": 0.0002627221401780844,
          "train_metric": 2.7920121283386834e-05,
          "validation_metric": 2.6460413209861144e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.704108043000815
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.00028258864585950505,
          "train_loss": 0.00027454286464489996,
          "validation_loss": 0.00028097903123125434,
          "train_metric": 2.765095268841833e-05,
          "validation_metric": 2.829918230418116e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.7236282440007926
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.0002774251624941826,
          "train_loss": 0.0002770032733678818,
          "validation_loss": 0.0002794186875689775,
          "train_metric": 2.7898755433852784e-05,
          "validation_metric": 2.81420234387042e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.7423151019993384
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.000281498214462772,
          "train_loss": 0.00027336610946804285,
          "validation_loss": 0.0002670305548235774,
          "train_metric": 2.7532434614840895e-05,
          "validation_metric": 2.689434404601343e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.760904983000728
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0003047121990675805,
          "train_loss": 0.00028182900859974325,
          "validation_loss": 0.0002679592289496213,
          "train_metric": 2.8384787583490834e-05,
          "validation_metric": 2.6987871024175547e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.779958274000819
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0002949428144347621,
          "train_loss": 0.00037207064451649785,
          "validation_loss": 0.0003863939782604575,
          "train_metric": 3.747359733097255e-05,
          "validation_metric": 3.891617961926386e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.7990331940000033
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.0003296904374110454,
          "train_loss": 0.00032887861016206443,
          "validation_loss": 0.00027232340653426945,
          "train_metric": 3.312344415462576e-05,
          "validation_metric": 2.7427418899605982e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.8180608410002606
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.0002932949237219873,
          "train_loss": 0.0002864811394829303,
          "validation_loss": 0.00032269765506498516,
          "train_metric": 2.8853335606982e-05,
          "validation_metric": 3.2500924135092646e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9339910840008088
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.00031233494519256055,
          "train_loss": 0.0003354971413500607,
          "validation_loss": 0.00027295181644149125,
          "train_metric": 3.37900455633644e-05,
          "validation_metric": 2.7490706997923553e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.9545838159992854
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.000305868627037853,
          "train_loss": 0.0002735816815402359,
          "validation_loss": 0.0002744573284871876,
          "train_metric": 2.7554144253372215e-05,
          "validation_metric": 2.7642336135613732e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.973455780000222
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.00028166409447294427,
          "train_loss": 0.0002830291632562876,
          "validation_loss": 0.00030536408303305507,
          "train_metric": 2.8505659429356456e-05,
          "validation_metric": 3.075515269301832e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 3.991812943999321
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.0002765267099675839,
          "train_loss": 0.0002776173932943493,
          "validation_loss": 0.0002520274429116398,
          "train_metric": 2.7960606530541554e-05,
          "validation_metric": 2.538328044465743e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.011373231000107
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.00028787116661987966,
          "train_loss": 0.0002983140875585377,
          "validation_loss": 0.00032546688453294337,
          "train_metric": 3.0045106541365385e-05,
          "validation_metric": 3.277983341831714e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.030977956999777
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.00027445032174000517,
          "train_loss": 0.00030929030617699027,
          "validation_loss": 0.00026881572557613254,
          "train_metric": 3.115058643743396e-05,
          "validation_metric": 2.7074138415628113e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.050346881000223
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0002880951542465482,
          "train_loss": 0.0002639041340444237,
          "validation_loss": 0.00028102315263822675,
          "train_metric": 2.6579456971376203e-05,
          "validation_metric": 2.8303622457315214e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.0691152830004285
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0003018040683855361,
          "train_loss": 0.00026237673591822386,
          "validation_loss": 0.00026851051370613277,
          "train_metric": 2.6425623218528926e-05,
          "validation_metric": 2.704339931369759e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.088554441999804
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.00027693030551745323,
          "train_loss": 0.0002774022286757827,
          "validation_loss": 0.00028516838210634887,
          "train_metric": 2.7938933271798305e-05,
          "validation_metric": 2.8721115086227655e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.107955670000592
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.000292743418867758,
          "train_loss": 0.00030179021996445954,
          "validation_loss": 0.0002531386271584779,
          "train_metric": 3.0395207431865856e-05,
          "validation_metric": 2.5495195586699992e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.126563509000334
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.0002841609748429619,
          "train_loss": 0.0002864125417545438,
          "validation_loss": 0.00034233726910315454,
          "train_metric": 2.8846421628259122e-05,
          "validation_metric": 3.4478947782190517e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.1451297650000924
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.00028262842215553974,
          "train_loss": 0.00029193400405347347,
          "validation_loss": 0.0002361818915233016,
          "train_metric": 2.940252306871116e-05,
          "validation_metric": 2.3787375539541245e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.164749855999617
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.00027356098416930763,
          "train_loss": 0.00026882460224442184,
          "validation_loss": 0.00031252397457137704,
          "train_metric": 2.7075027901446447e-05,
          "validation_metric": 3.147626921418123e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.18384940299984
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.00027205079641134944,
          "train_loss": 0.0002974964736495167,
          "validation_loss": 0.0002925922744907439,
          "train_metric": 2.996275361510925e-05,
          "validation_metric": 2.94688215944916e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.2027780980006355
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.0002762031163001666,
          "train_loss": 0.0003075434360653162,
          "validation_loss": 0.0002585119509603828,
          "train_metric": 3.097464650636539e-05,
          "validation_metric": 2.603637767606415e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.222051398999611
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.00027356992723071016,
          "train_loss": 0.0002791658043861389,
          "validation_loss": 0.000306255416944623,
          "train_metric": 2.811655940604396e-05,
          "validation_metric": 3.084492345806211e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.2407643149999785
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.00029273270774865523,
          "train_loss": 0.0003021188313141465,
          "validation_loss": 0.00029132605413906276,
          "train_metric": 3.0428303944063373e-05,
          "validation_metric": 2.9341294066398405e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.259558198000377
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0003225418145120784,
          "train_loss": 0.00027510785730555654,
          "validation_loss": 0.000306733330944553,
          "train_metric": 2.7707856133929454e-05,
          "validation_metric": 3.089305755565874e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.278187223000714
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.0002822446399477485,
          "train_loss": 0.0002588585193734616,
          "validation_loss": 0.0002432977344142273,
          "train_metric": 2.607128226372879e-05,
          "validation_metric": 2.4504057364538312e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.296848906000378
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.00031507431503996486,
          "train_loss": 0.00026554428040981293,
          "validation_loss": 0.0002507561875972897,
          "train_metric": 2.674464849405922e-05,
          "validation_metric": 2.5255241780541837e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.315587741000854
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0002837461324816104,
          "train_loss": 0.0002929348556790501,
          "validation_loss": 0.00025727623142302036,
          "train_metric": 2.9503327823476866e-05,
          "validation_metric": 2.591192242107354e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.334437969000646
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0002699688475331641,
          "train_loss": 0.00025828974321484566,
          "validation_loss": 0.00028611949528567493,
          "train_metric": 2.6013993192464113e-05,
          "validation_metric": 2.881690852518659e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.353252655999313
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0002958355607916019,
          "train_loss": 0.0002502142160665244,
          "validation_loss": 0.0002510654157958925,
          "train_metric": 2.5200664822477847e-05,
          "validation_metric": 2.528638651710935e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 4.3732075879997865
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.0002572743542259559,
          "train_loss": 0.00024769813171587884,
          "validation_loss": 0.0002454916248098016,
          "train_metric": 2.4947246856754646e-05,
          "validation_metric": 2.472502092132345e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.392699743999401
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0002592805003587273,
          "train_loss": 0.0002469761238899082,
          "validation_loss": 0.0002592137607280165,
          "train_metric": 2.4874530936358497e-05,
          "validation_metric": 2.6107059966307133e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.41130972999963
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.00026691292214309215,
          "train_loss": 0.0002450831525493413,
          "validation_loss": 0.0002507681492716074,
          "train_metric": 2.468387538101524e-05,
          "validation_metric": 2.525644958950579e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.429640189000565
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0002524776509744697,
          "train_loss": 0.0002470760955475271,
          "validation_loss": 0.000256396277109161,
          "train_metric": 2.4884595404728316e-05,
          "validation_metric": 2.5823293981375173e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.4490267599994695
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0002755154564511031,
          "train_loss": 0.00024444516748189926,
          "validation_loss": 0.00025709415785968304,
          "train_metric": 2.4619621399324387e-05,
          "validation_metric": 2.5893581550917588e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.468019165000442
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.00025699631669340306,
          "train_loss": 0.00024352163018193096,
          "validation_loss": 0.00024917160044424236,
          "train_metric": 2.4526603738195263e-05,
          "validation_metric": 2.5095649107242934e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.486119896000673
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0002510589843041089,
          "train_loss": 0.00024340096570085734,
          "validation_loss": 0.000254352722549811,
          "train_metric": 2.451445470796898e-05,
          "validation_metric": 2.5617477149353363e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.504639982000299
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.00024904265092118294,
          "train_loss": 0.00024239730555564165,
          "validation_loss": 0.0002537426189519465,
          "train_metric": 2.4413368009845726e-05,
          "validation_metric": 2.5556028049322776e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.523147371999585
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.0002453253355270135,
          "train_loss": 0.00024363023112528026,
          "validation_loss": 0.00024607841623947024,
          "train_metric": 2.4537543140468188e-05,
          "validation_metric": 2.4784119887044653e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 4.541590148000978
        }
      ],
      "best": {
        "epoch": 189,
        "updates": 3024
      },
      "best_validation_metric": 2.3787375539541245e-05,
      "config": {
        "label": "LayerNorm + tanh + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.004877230314383487
    },
    "d5c218b8a06d_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 1.2026782371103764,
          "train_loss": 0.7293098568916321,
          "validation_loss": 0.31126153469085693,
          "train_metric": 0.07345342636108398,
          "validation_metric": 0.031349122524261475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06936033000056341
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.6360595803707838,
          "train_loss": 0.5844699740409851,
          "validation_loss": 0.10958640277385712,
          "train_metric": 0.05886567384004593,
          "validation_metric": 0.01103714108467102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13854980699943553
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.5737623050808907,
          "train_loss": 0.5421966910362244,
          "validation_loss": 0.13522636890411377,
          "train_metric": 0.0546080656349659,
          "validation_metric": 0.013619504868984222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20357936800064635
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.5317818811163306,
          "train_loss": 0.5187839865684509,
          "validation_loss": 0.08521861582994461,
          "train_metric": 0.05225002393126488,
          "validation_metric": 0.008582905866205692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2729798309992475
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.5140035655349493,
          "train_loss": 0.5039367079734802,
          "validation_loss": 0.07849752902984619,
          "train_metric": 0.05075465887784958,
          "validation_metric": 0.007905984297394753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34077430899924366
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.5079674245789647,
          "train_loss": 0.5003489255905151,
          "validation_loss": 0.09958729892969131,
          "train_metric": 0.05039330944418907,
          "validation_metric": 0.010030066594481468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5964558480009146
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.5020176793914288,
          "train_loss": 0.4900491237640381,
          "validation_loss": 0.0717223733663559,
          "train_metric": 0.04935595765709877,
          "validation_metric": 0.007223614025861025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6683210619994497
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.4932086248882115,
          "train_loss": 0.4820261001586914,
          "validation_loss": 0.06344542652368546,
          "train_metric": 0.048547904938459396,
          "validation_metric": 0.006389991380274296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7419364050001604
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.4890028042718768,
          "train_loss": 0.48184463381767273,
          "validation_loss": 0.07463790476322174,
          "train_metric": 0.048529624938964844,
          "validation_metric": 0.0075172558426856995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8067441279999912
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.49038014374673367,
          "train_loss": 0.47763216495513916,
          "validation_loss": 0.0683656558394432,
          "train_metric": 0.04810535907745361,
          "validation_metric": 0.006885537877678871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.872795421999399
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.4851382430642843,
          "train_loss": 0.4773828685283661,
          "validation_loss": 0.05736415088176727,
          "train_metric": 0.04808025807142258,
          "validation_metric": 0.005777507554739714,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9505100389997096
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.47870074585080147,
          "train_loss": 0.48097729682922363,
          "validation_loss": 0.043865133076906204,
          "train_metric": 0.04844226688146591,
          "validation_metric": 0.004417935851961374,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0678028830006951
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.4835465317592025,
          "train_loss": 0.47345614433288574,
          "validation_loss": 0.07218296080827713,
          "train_metric": 0.047684770077466965,
          "validation_metric": 0.007270003203302622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1439109570001165
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.47679297626018524,
          "train_loss": 0.4731188118457794,
          "validation_loss": 0.060062505304813385,
          "train_metric": 0.047650791704654694,
          "validation_metric": 0.0060492753982543945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2299347080006555
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.47279727179557085,
          "train_loss": 0.4682210385799408,
          "validation_loss": 0.05355948209762573,
          "train_metric": 0.04715751111507416,
          "validation_metric": 0.005394314415752888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.275515033999909
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.4770183628425002,
          "train_loss": 0.46543341875076294,
          "validation_loss": 0.051297713071107864,
          "train_metric": 0.04687674716114998,
          "validation_metric": 0.005166517570614815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.286530182000206
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.4727508861105889,
          "train_loss": 0.47011470794677734,
          "validation_loss": 0.07857362926006317,
          "train_metric": 0.047348231077194214,
          "validation_metric": 0.007913649082183838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2974892940001155
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.47039608377963305,
          "train_loss": 0.47392737865448,
          "validation_loss": 0.07112348824739456,
          "train_metric": 0.04773223400115967,
          "validation_metric": 0.007163296919316053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.308543056000417
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.4693222362548113,
          "train_loss": 0.46241268515586853,
          "validation_loss": 0.06525902450084686,
          "train_metric": 0.04657251387834549,
          "validation_metric": 0.006572650745511055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.321996250000666
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.46601996943354607,
          "train_loss": 0.4704171419143677,
          "validation_loss": 0.035987354815006256,
          "train_metric": 0.047378696501255035,
          "validation_metric": 0.0036245149094611406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3365676590001385
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.47726463340222836,
          "train_loss": 0.4878505766391754,
          "validation_loss": 0.03944571316242218,
          "train_metric": 0.049134526401758194,
          "validation_metric": 0.003972827922552824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3478856820001965
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.4859011508524418,
          "train_loss": 0.4679509699344635,
          "validation_loss": 0.06192736700177193,
          "train_metric": 0.04713031277060509,
          "validation_metric": 0.006237098015844822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3593222159997822
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.4832295193336904,
          "train_loss": 0.4720529317855835,
          "validation_loss": 0.0576460100710392,
          "train_metric": 0.04754344001412392,
          "validation_metric": 0.005805895663797855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3709308260004036
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.4674167283810675,
          "train_loss": 0.4680629372596741,
          "validation_loss": 0.05802706256508827,
          "train_metric": 0.047141581773757935,
          "validation_metric": 0.005844273138791323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3820330859998649
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.46451911237090826,
          "train_loss": 0.45659714937210083,
          "validation_loss": 0.03545249253511429,
          "train_metric": 0.04598679021000862,
          "validation_metric": 0.003570645349100232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3936030170007143
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.4759107774589211,
          "train_loss": 0.47369635105133057,
          "validation_loss": 0.025964854285120964,
          "train_metric": 0.04770895838737488,
          "validation_metric": 0.002615084871649742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4055753510001523
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.46788898948580027,
          "train_loss": 0.47021612524986267,
          "validation_loss": 0.09218467026948929,
          "train_metric": 0.047358449548482895,
          "validation_metric": 0.00928450282663107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4167262600003596
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.4692828869447112,
          "train_loss": 0.4645330011844635,
          "validation_loss": 0.05885409191250801,
          "train_metric": 0.04678605869412422,
          "validation_metric": 0.005927568301558495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.427828842000963
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.4639119626954198,
          "train_loss": 0.45296669006347656,
          "validation_loss": 0.030596088618040085,
          "train_metric": 0.045621149241924286,
          "validation_metric": 0.003081526141613722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.438786393000555
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.4570441283285618,
          "train_loss": 0.4752497971057892,
          "validation_loss": 0.08237441629171371,
          "train_metric": 0.04786541685461998,
          "validation_metric": 0.008296448737382889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4495154110009025
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.4680082472041249,
          "train_loss": 0.45231494307518005,
          "validation_loss": 0.04164813086390495,
          "train_metric": 0.04555550962686539,
          "validation_metric": 0.004194647539407015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4604252280005312
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.45889217779040337,
          "train_loss": 0.4506063461303711,
          "validation_loss": 0.04289703071117401,
          "train_metric": 0.04538342356681824,
          "validation_metric": 0.004320431966334581,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4710347799991723
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.4541114345192909,
          "train_loss": 0.451497346162796,
          "validation_loss": 0.033135365694761276,
          "train_metric": 0.04547315835952759,
          "validation_metric": 0.0033372726757079363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4817425040000671
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.45904514007270336,
          "train_loss": 0.4502199590206146,
          "validation_loss": 0.04907520115375519,
          "train_metric": 0.045344505459070206,
          "validation_metric": 0.004942674655467272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.492700442999194
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.4556976556777954,
          "train_loss": 0.44840285181999207,
          "validation_loss": 0.030781537294387817,
          "train_metric": 0.04516149312257767,
          "validation_metric": 0.0031002035830169916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5034847629995056
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.4542266046628356,
          "train_loss": 0.45433199405670166,
          "validation_loss": 0.03317631036043167,
          "train_metric": 0.04575866088271141,
          "validation_metric": 0.00334139633923769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5143450970008416
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.45609058532863855,
          "train_loss": 0.4489269554615021,
          "validation_loss": 0.03823782876133919,
          "train_metric": 0.045214276760816574,
          "validation_metric": 0.003851174144074321,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5249140279993298
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.45510272309184074,
          "train_loss": 0.46448954939842224,
          "validation_loss": 0.027873510494828224,
          "train_metric": 0.04678168147802353,
          "validation_metric": 0.0028073180001229048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.535990849000882
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.45666081458330154,
          "train_loss": 0.45510581135749817,
          "validation_loss": 0.0277418065816164,
          "train_metric": 0.045836593955755234,
          "validation_metric": 0.0027940531726926565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5466881480006123
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.45143533078953624,
          "train_loss": 0.4498971700668335,
          "validation_loss": 0.03823717683553696,
          "train_metric": 0.04531199485063553,
          "validation_metric": 0.0038511077873408794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5580106760007766
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.45197259448468685,
          "train_loss": 0.44987285137176514,
          "validation_loss": 0.04342999681830406,
          "train_metric": 0.04530954733490944,
          "validation_metric": 0.004374110139906406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5696914910004125
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.4515529852360487,
          "train_loss": 0.45170193910598755,
          "validation_loss": 0.03805416449904442,
          "train_metric": 0.04549377039074898,
          "validation_metric": 0.003832676215097308,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.580810900999495
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.45632106624543667,
          "train_loss": 0.45039570331573486,
          "validation_loss": 0.022302696481347084,
          "train_metric": 0.0453622043132782,
          "validation_metric": 0.002246245974674821,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5924830029998702
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.47361381351947784,
          "train_loss": 0.4476272761821747,
          "validation_loss": 0.04577041044831276,
          "train_metric": 0.045083384960889816,
          "validation_metric": 0.004609828349202871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.603662154999256
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.45513481041416526,
          "train_loss": 0.452519953250885,
          "validation_loss": 0.029892129823565483,
          "train_metric": 0.045576147735118866,
          "validation_metric": 0.003010626183822751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6146442549998028
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.45137087698094547,
          "train_loss": 0.4457034170627594,
          "validation_loss": 0.027299363166093826,
          "train_metric": 0.04488961771130562,
          "validation_metric": 0.002749491948634386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6256938749993424
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.46105102077126503,
          "train_loss": 0.4555639922618866,
          "validation_loss": 0.021578123793005943,
          "train_metric": 0.04588273912668228,
          "validation_metric": 0.0021732694003731012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.637521556000138
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.45624903310090303,
          "train_loss": 0.4472806751728058,
          "validation_loss": 0.026648633182048798,
          "train_metric": 0.04504847154021263,
          "validation_metric": 0.0026839529164135456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.648582900999827
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.4507075088331476,
          "train_loss": 0.44850027561187744,
          "validation_loss": 0.02816079743206501,
          "train_metric": 0.04517130181193352,
          "validation_metric": 0.002836252097040415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6594030329997622
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.4506010664626956,
          "train_loss": 0.45032942295074463,
          "validation_loss": 0.027645353227853775,
          "train_metric": 0.045355528593063354,
          "validation_metric": 0.0027843392454087734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6702385419994243
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.4563529994338751,
          "train_loss": 0.4458363652229309,
          "validation_loss": 0.03072291426360607,
          "train_metric": 0.04490300640463829,
          "validation_metric": 0.003094299230724573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6809762480006611
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.46997447591274977,
          "train_loss": 0.4472334384918213,
          "validation_loss": 0.03811461105942726,
          "train_metric": 0.04504372179508209,
          "validation_metric": 0.003838764037936926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6919306590007182
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.45847533037886024,
          "train_loss": 0.4592421352863312,
          "validation_loss": 0.06249275431036949,
          "train_metric": 0.04625318944454193,
          "validation_metric": 0.006294041872024536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7028521289994387
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.45278312684968114,
          "train_loss": 0.45418581366539,
          "validation_loss": 0.02248762920498848,
          "train_metric": 0.045743927359580994,
          "validation_metric": 0.0022648717276751995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7138242040000478
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.45401423517614603,
          "train_loss": 0.44971805810928345,
          "validation_loss": 0.02279682643711567,
          "train_metric": 0.045293960720300674,
          "validation_metric": 0.0022960130590945482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7246809210009815
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.4566430156119168,
          "train_loss": 0.446811705827713,
          "validation_loss": 0.02537696249783039,
          "train_metric": 0.045001234859228134,
          "validation_metric": 0.002555874641984701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7356072359998507
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.4504860118031502,
          "train_loss": 0.4445638656616211,
          "validation_loss": 0.03525340184569359,
          "train_metric": 0.04477484151721001,
          "validation_metric": 0.0035505937412381172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.746591047000038
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.451172191882506,
          "train_loss": 0.44400396943092346,
          "validation_loss": 0.031488459557294846,
          "train_metric": 0.04471845179796219,
          "validation_metric": 0.003171402495354414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7573112350000883
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.449641983024776,
          "train_loss": 0.4646117687225342,
          "validation_loss": 0.06722179055213928,
          "train_metric": 0.046793997287750244,
          "validation_metric": 0.0067703332751989365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7683655510008975
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.4670361829921603,
          "train_loss": 0.44745850563049316,
          "validation_loss": 0.036180607974529266,
          "train_metric": 0.04506637901067734,
          "validation_metric": 0.003643978387117386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7792321619999711
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.44862031657248735,
          "train_loss": 0.44555041193962097,
          "validation_loss": 0.025208184495568275,
          "train_metric": 0.04487420618534088,
          "validation_metric": 0.0025388761423528194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7900874200004182
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.45436049439013004,
          "train_loss": 0.4450947046279907,
          "validation_loss": 0.03192766755819321,
          "train_metric": 0.04482830688357353,
          "validation_metric": 0.0032156379893422127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8010364349993324
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.4595406362786889,
          "train_loss": 0.46701765060424805,
          "validation_loss": 0.06381319463253021,
          "train_metric": 0.0470363087952137,
          "validation_metric": 0.006427031476050615,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.811932994000017
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.4600594323128462,
          "train_loss": 0.4450894296169281,
          "validation_loss": 0.04165557771921158,
          "train_metric": 0.044827770441770554,
          "validation_metric": 0.004195397719740868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8229736299999786
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.4536797939799726,
          "train_loss": 0.4493331015110016,
          "validation_loss": 0.01679016649723053,
          "train_metric": 0.04525517672300339,
          "validation_metric": 0.0016910441918298602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8342560920009419
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.4518917230889201,
          "train_loss": 0.4492551386356354,
          "validation_loss": 0.047664087265729904,
          "train_metric": 0.04524733126163483,
          "validation_metric": 0.004800552502274513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8454145590003463
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.45824658032506704,
          "train_loss": 0.4534675180912018,
          "validation_loss": 0.0512089729309082,
          "train_metric": 0.04567158967256546,
          "validation_metric": 0.00515757966786623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.856375945999389
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.44995660870335996,
          "train_loss": 0.4522004723548889,
          "validation_loss": 0.05343877151608467,
          "train_metric": 0.04554397612810135,
          "validation_metric": 0.005382157396525145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8673807270006364
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.45292061753571033,
          "train_loss": 0.4434307813644409,
          "validation_loss": 0.0347926989197731,
          "train_metric": 0.04466072842478752,
          "validation_metric": 0.003504193155094981,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8781740720005473
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.4487909721210599,
          "train_loss": 0.4618000388145447,
          "validation_loss": 0.06015397608280182,
          "train_metric": 0.046510808169841766,
          "validation_metric": 0.006058488041162491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.889219917000446
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.46307213022373617,
          "train_loss": 0.44324782490730286,
          "validation_loss": 0.025695152580738068,
          "train_metric": 0.04464229941368103,
          "validation_metric": 0.0025879216846078634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9001475790009863
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.45103039126843214,
          "train_loss": 0.44736942648887634,
          "validation_loss": 0.047724463045597076,
          "train_metric": 0.04505740478634834,
          "validation_metric": 0.00480663264170289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9110274459999346
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.4479784432332963,
          "train_loss": 0.44249966740608215,
          "validation_loss": 0.02925943024456501,
          "train_metric": 0.04456694424152374,
          "validation_metric": 0.00294690253213048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9218342820004182
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.44847920909523964,
          "train_loss": 0.4420109987258911,
          "validation_loss": 0.031020458787679672,
          "train_metric": 0.04451772943139076,
          "validation_metric": 0.0031242670956999063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.932983332000731
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.4514791809488088,
          "train_loss": 0.4467306137084961,
          "validation_loss": 0.04010139778256416,
          "train_metric": 0.04499306157231331,
          "validation_metric": 0.004038865678012371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9437514250003005
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.4533541315468028,
          "train_loss": 0.44415995478630066,
          "validation_loss": 0.031070232391357422,
          "train_metric": 0.0447341650724411,
          "validation_metric": 0.003129279939457774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.025690862999909
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.44916508067399263,
          "train_loss": 0.44273900985717773,
          "validation_loss": 0.02347850427031517,
          "train_metric": 0.04459105804562569,
          "validation_metric": 0.0023646687623113394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.037009143999967
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.44885523873381317,
          "train_loss": 0.4428853392601013,
          "validation_loss": 0.0281156525015831,
          "train_metric": 0.044605791568756104,
          "validation_metric": 0.0028317056130617857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.047829252000156
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.44781280029565096,
          "train_loss": 0.4481898248195648,
          "validation_loss": 0.020853813737630844,
          "train_metric": 0.045140042901039124,
          "validation_metric": 0.0021003198344260454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.058294685999499
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.4485369147732854,
          "train_loss": 0.44777026772499084,
          "validation_loss": 0.018803652375936508,
          "train_metric": 0.04509778320789337,
          "validation_metric": 0.0018938350258395076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.069089229000383
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.4530335636809468,
          "train_loss": 0.4413500726222992,
          "validation_loss": 0.03243809565901756,
          "train_metric": 0.044451162219047546,
          "validation_metric": 0.003267045831307769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.07983713899921
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.44376709032803774,
          "train_loss": 0.44102048873901367,
          "validation_loss": 0.02625206485390663,
          "train_metric": 0.04441796988248825,
          "validation_metric": 0.0026440119836479425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0904621509998833
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.4460496064275503,
          "train_loss": 0.4425477981567383,
          "validation_loss": 0.038809701800346375,
          "train_metric": 0.04457179456949234,
          "validation_metric": 0.003908771090209484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.101033409999218
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.4465923700481653,
          "train_loss": 0.4489843547344208,
          "validation_loss": 0.02658543549478054,
          "train_metric": 0.04522005841135979,
          "validation_metric": 0.0026775880251079798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1115171590008686
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.4539604517631233,
          "train_loss": 0.45346060395240784,
          "validation_loss": 0.019252993166446686,
          "train_metric": 0.04567088931798935,
          "validation_metric": 0.001939091132953763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1216566919993056
        }
      ],
      "best": {
        "epoch": 65,
        "updates": 1040
      },
      "best_validation_metric": 0.0016910441918298602,
      "config": {
        "label": "MSE + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": true,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.04112230771527615
    },
    "fc71c16e3acd_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.7378480453044176,
          "train_loss": 0.5219846367835999,
          "validation_loss": 0.32044413685798645,
          "train_metric": 0.0759110301733017,
          "validation_metric": 0.016800832003355026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02342659299938532
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.4270789474248886,
          "train_loss": 0.3663177490234375,
          "validation_loss": 0.21789206564426422,
          "train_metric": 0.06053704023361206,
          "validation_metric": 0.015553402714431286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.037858765999772004
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.3559560040012002,
          "train_loss": 0.3323574364185333,
          "validation_loss": 0.1885303258895874,
          "train_metric": 0.056650854647159576,
          "validation_metric": 0.01306948997080326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05174165599964908
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.33214769791811705,
          "train_loss": 0.3062162697315216,
          "validation_loss": 0.1631731241941452,
          "train_metric": 0.055497925728559494,
          "validation_metric": 0.011663412675261497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06485805299962522
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.30752558540552855,
          "train_loss": 0.3015317916870117,
          "validation_loss": 0.14575810730457306,
          "train_metric": 0.05449260398745537,
          "validation_metric": 0.009739075787365437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07784222900045279
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.29474763851612806,
          "train_loss": 0.2878584563732147,
          "validation_loss": 0.14004731178283691,
          "train_metric": 0.05353228375315666,
          "validation_metric": 0.008617738261818886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0915679209992959
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.28705803491175175,
          "train_loss": 0.2833433747291565,
          "validation_loss": 0.13528946042060852,
          "train_metric": 0.052503593266010284,
          "validation_metric": 0.007670003920793533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1046097100006591
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.28457442205399275,
          "train_loss": 0.28120744228363037,
          "validation_loss": 0.1345420628786087,
          "train_metric": 0.05232308804988861,
          "validation_metric": 0.008156122639775276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11699717199917359
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.2805907856673002,
          "train_loss": 0.27728036046028137,
          "validation_loss": 0.13238880038261414,
          "train_metric": 0.05202078819274902,
          "validation_metric": 0.00741867208853364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12989866899988556
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.2781245904043317,
          "train_loss": 0.2744764983654022,
          "validation_loss": 0.13053566217422485,
          "train_metric": 0.05148182436823845,
          "validation_metric": 0.006963340565562248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14261304799947538
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.2757938811555505,
          "train_loss": 0.2693432867527008,
          "validation_loss": 0.12101969867944717,
          "train_metric": 0.051602501422166824,
          "validation_metric": 0.0069107976742088795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15581670199935616
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.2699753427878022,
          "train_loss": 0.2636840045452118,
          "validation_loss": 0.1161474958062172,
          "train_metric": 0.05114779621362686,
          "validation_metric": 0.006240795832127333,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16895006199956697
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.2640720484778285,
          "train_loss": 0.259927362203598,
          "validation_loss": 0.11603640764951706,
          "train_metric": 0.05126132443547249,
          "validation_metric": 0.006451419088989496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18123024399938004
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.2560720443725586,
          "train_loss": 0.2490968555212021,
          "validation_loss": 0.10259772837162018,
          "train_metric": 0.04999663680791855,
          "validation_metric": 0.005132453050464392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1935902730001544
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.2507364843040705,
          "train_loss": 0.2573651075363159,
          "validation_loss": 0.10571520030498505,
          "train_metric": 0.05059119686484337,
          "validation_metric": 0.0042732530273497105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20615182999972603
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.24709030333906412,
          "train_loss": 0.23337560892105103,
          "validation_loss": 0.08474212139844894,
          "train_metric": 0.04899875074625015,
          "validation_metric": 0.004179536364972591,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21867100000054052
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.23559070006012917,
          "train_loss": 0.23049812018871307,
          "validation_loss": 0.08401677757501602,
          "train_metric": 0.04813678190112114,
          "validation_metric": 0.0038302212487906218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2313637209990702
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.22861655335873365,
          "train_loss": 0.23293167352676392,
          "validation_loss": 0.0857844427227974,
          "train_metric": 0.04804686829447746,
          "validation_metric": 0.0039003717247396708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24629293699945265
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.23008936108089983,
          "train_loss": 0.2316148579120636,
          "validation_loss": 0.07851789146661758,
          "train_metric": 0.04837990179657936,
          "validation_metric": 0.002919863909482956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26027492200046254
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.22289870120584965,
          "train_loss": 0.21963101625442505,
          "validation_loss": 0.07076691836118698,
          "train_metric": 0.047774508595466614,
          "validation_metric": 0.0029481721576303244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28080196200062346
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.22306269919499755,
          "train_loss": 0.22003327310085297,
          "validation_loss": 0.06756801158189774,
          "train_metric": 0.04777051508426666,
          "validation_metric": 0.0023419184144586325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3035627999997814
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.21871617948636413,
          "train_loss": 0.21808356046676636,
          "validation_loss": 0.06935285031795502,
          "train_metric": 0.04789868742227554,
          "validation_metric": 0.0032228499185293913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35111192499971366
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.2225348656065762,
          "train_loss": 0.22196954488754272,
          "validation_loss": 0.07096520066261292,
          "train_metric": 0.0471591055393219,
          "validation_metric": 0.002390899695456028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37231559800056857
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.21932524256408215,
          "train_loss": 0.21190591156482697,
          "validation_loss": 0.06457315385341644,
          "train_metric": 0.04709157720208168,
          "validation_metric": 0.0022327308543026447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3936829399990529
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.21232172870077193,
          "train_loss": 0.21242228150367737,
          "validation_loss": 0.05858708173036575,
          "train_metric": 0.04762386903166771,
          "validation_metric": 0.0018813060596585274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40568751899991184
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.2109002866782248,
          "train_loss": 0.2056390345096588,
          "validation_loss": 0.05571288987994194,
          "train_metric": 0.047196634113788605,
          "validation_metric": 0.0018976535648107529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.435712546999639
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.20712090656161308,
          "train_loss": 0.20241566002368927,
          "validation_loss": 0.05179627984762192,
          "train_metric": 0.047143395990133286,
          "validation_metric": 0.001771810231730342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4627309150000656
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.20461133029311895,
          "train_loss": 0.20386889576911926,
          "validation_loss": 0.050304561853408813,
          "train_metric": 0.04737682640552521,
          "validation_metric": 0.0015919810393825173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4947037599995383
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.20458990475162864,
          "train_loss": 0.2030000388622284,
          "validation_loss": 0.053936779499053955,
          "train_metric": 0.046604882925748825,
          "validation_metric": 0.0017297726590186357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5219948809990456
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.20176606136374176,
          "train_loss": 0.20603056252002716,
          "validation_loss": 0.05332007259130478,
          "train_metric": 0.0474129319190979,
          "validation_metric": 0.0012597155291587114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.535035445000176
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.20356288691982627,
          "train_loss": 0.1990450620651245,
          "validation_loss": 0.050926852971315384,
          "train_metric": 0.046640682965517044,
          "validation_metric": 0.0015342202968895435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5474669880004512
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.1992226275615394,
          "train_loss": 0.20071223378181458,
          "validation_loss": 0.04731631651520729,
          "train_metric": 0.047321174293756485,
          "validation_metric": 0.001184183289296925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5598489080002764
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.2012207992374897,
          "train_loss": 0.2045070230960846,
          "validation_loss": 0.054142434149980545,
          "train_metric": 0.04671063646674156,
          "validation_metric": 0.001203671214170754,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5715077559998463
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.2008315809071064,
          "train_loss": 0.20253632962703705,
          "validation_loss": 0.05288918316364288,
          "train_metric": 0.046575985848903656,
          "validation_metric": 0.0013064570957794785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5835016809996887
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.1992915398441255,
          "train_loss": 0.20326384902000427,
          "validation_loss": 0.05008623003959656,
          "train_metric": 0.04696865379810333,
          "validation_metric": 0.0015102721517905593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5952369110000291
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.20120514603331685,
          "train_loss": 0.19778300821781158,
          "validation_loss": 0.04480849951505661,
          "train_metric": 0.046824514865875244,
          "validation_metric": 0.0012172497808933258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6071524559993122
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.1963407319271937,
          "train_loss": 0.19182604551315308,
          "validation_loss": 0.03881119191646576,
          "train_metric": 0.04682214558124542,
          "validation_metric": 0.0010089542483910918,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6199866739989375
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.19387923181056976,
          "train_loss": 0.19316747784614563,
          "validation_loss": 0.04260895401239395,
          "train_metric": 0.0466848686337471,
          "validation_metric": 0.0011918529635295272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6319056209995324
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.19338587019592524,
          "train_loss": 0.19074279069900513,
          "validation_loss": 0.03986440971493721,
          "train_metric": 0.04647240787744522,
          "validation_metric": 0.0009937755530700088,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7187705659998755
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.19381978875026107,
          "train_loss": 0.1942221224308014,
          "validation_loss": 0.04312332347035408,
          "train_metric": 0.046624768525362015,
          "validation_metric": 0.0008352624136023223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.731503277000229
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.19221028313040733,
          "train_loss": 0.1918589472770691,
          "validation_loss": 0.039524804800748825,
          "train_metric": 0.04655587673187256,
          "validation_metric": 0.0008329221163876355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.744513747999008
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.19185500172898173,
          "train_loss": 0.1934974640607834,
          "validation_loss": 0.04155174642801285,
          "train_metric": 0.04676981642842293,
          "validation_metric": 0.0009099962189793587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7568725979999726
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.19348505651578307,
          "train_loss": 0.19978435337543488,
          "validation_loss": 0.04891902208328247,
          "train_metric": 0.04634689912199974,
          "validation_metric": 0.0008437424548901618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7686912379995192
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.19724908145144582,
          "train_loss": 0.20439444482326508,
          "validation_loss": 0.055349912494421005,
          "train_metric": 0.04623003676533699,
          "validation_metric": 0.0013596231583505869,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7805630339989875
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.19616154697723687,
          "train_loss": 0.19498592615127563,
          "validation_loss": 0.040691256523132324,
          "train_metric": 0.047148022800683975,
          "validation_metric": 0.0007138139917515218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7928424029996677
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.1907249039504677,
          "train_loss": 0.18981221318244934,
          "validation_loss": 0.03761214762926102,
          "train_metric": 0.04635815694928169,
          "validation_metric": 0.0009694320033304393,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8054096859996207
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.1906367428600788,
          "train_loss": 0.18905727565288544,
          "validation_loss": 0.0367722362279892,
          "train_metric": 0.04641292616724968,
          "validation_metric": 0.0009647932602092624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8177875290002703
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.19210545904934406,
          "train_loss": 0.19237880408763885,
          "validation_loss": 0.0383060984313488,
          "train_metric": 0.04646523296833038,
          "validation_metric": 0.0010633146157488227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.830052903998876
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.19020165875554085,
          "train_loss": 0.1881064772605896,
          "validation_loss": 0.03750837966799736,
          "train_metric": 0.046315331012010574,
          "validation_metric": 0.0007549276924692094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8417304220001824
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.18660650681704283,
          "train_loss": 0.1845930963754654,
          "validation_loss": 0.031471043825149536,
          "train_metric": 0.04625384509563446,
          "validation_metric": 0.0007063179509714246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8546266590001323
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.1876776171848178,
          "train_loss": 0.18807823956012726,
          "validation_loss": 0.034693073481321335,
          "train_metric": 0.04658403992652893,
          "validation_metric": 0.000577325641643256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8670039070002531
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.18969487911090255,
          "train_loss": 0.1881391406059265,
          "validation_loss": 0.03810793533921242,
          "train_metric": 0.04637372866272926,
          "validation_metric": 0.0005902369157411158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8794106599998486
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.18504922045394778,
          "train_loss": 0.18464523553848267,
          "validation_loss": 0.030329540371894836,
          "train_metric": 0.04655207693576813,
          "validation_metric": 0.0005470567848533392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8917890720003925
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.18515092367306352,
          "train_loss": 0.18663109838962555,
          "validation_loss": 0.03267085552215576,
          "train_metric": 0.046674661338329315,
          "validation_metric": 0.0005753807490691543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9042650369992771
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.18818894040305167,
          "train_loss": 0.18511919677257538,
          "validation_loss": 0.03320704773068428,
          "train_metric": 0.04657508060336113,
          "validation_metric": 0.0005447855801321566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9167061360003572
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.19044632650911808,
          "train_loss": 0.18807834386825562,
          "validation_loss": 0.034210871905088425,
          "train_metric": 0.04671395942568779,
          "validation_metric": 0.0006140246987342834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9288570429998799
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.1865712534636259,
          "train_loss": 0.18518252670764923,
          "validation_loss": 0.03229567036032677,
          "train_metric": 0.04621244966983795,
          "validation_metric": 0.000552903744392097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9411465110006247
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.1842410797253251,
          "train_loss": 0.1826099157333374,
          "validation_loss": 0.03383704274892807,
          "train_metric": 0.0462622195482254,
          "validation_metric": 0.0006062203319743276,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9530668900006276
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.18243812629953027,
          "train_loss": 0.1795494109392166,
          "validation_loss": 0.027272284030914307,
          "train_metric": 0.04626334458589554,
          "validation_metric": 0.0005242244224064052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9656295129989303
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.18115591653622687,
          "train_loss": 0.18370576202869415,
          "validation_loss": 0.03367806226015091,
          "train_metric": 0.04612208530306816,
          "validation_metric": 0.0007677469402551651,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9777868589990248
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.18658740352839231,
          "train_loss": 0.18584129214286804,
          "validation_loss": 0.03248004987835884,
          "train_metric": 0.04608244076371193,
          "validation_metric": 0.0006212270236574113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9898821520000638
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.18627081508748233,
          "train_loss": 0.1857871413230896,
          "validation_loss": 0.03411262482404709,
          "train_metric": 0.04663093015551567,
          "validation_metric": 0.0004943498643115163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0026401270006318
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.18489579716697335,
          "train_loss": 0.1885463446378708,
          "validation_loss": 0.03451923653483391,
          "train_metric": 0.04596176743507385,
          "validation_metric": 0.0005801896913908422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0146864670005016
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.18592460127547383,
          "train_loss": 0.18126371502876282,
          "validation_loss": 0.02964635007083416,
          "train_metric": 0.04594191163778305,
          "validation_metric": 0.0005085372249595821,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0265093839989277
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.18263505725190043,
          "train_loss": 0.18493926525115967,
          "validation_loss": 0.03234650939702988,
          "train_metric": 0.04660212621092796,
          "validation_metric": 0.0005332184955477715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.03784596100013
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.1916776760481298,
          "train_loss": 0.1833726316690445,
          "validation_loss": 0.02665817365050316,
          "train_metric": 0.04598241671919823,
          "validation_metric": 0.0004999336088076234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0493281750004826
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.19244241528213024,
          "train_loss": 0.181229367852211,
          "validation_loss": 0.026495283469557762,
          "train_metric": 0.04629097506403923,
          "validation_metric": 0.00044390757102519274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.061398611000186
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.18485109508037567,
          "train_loss": 0.18305490911006927,
          "validation_loss": 0.03265732154250145,
          "train_metric": 0.04610509052872658,
          "validation_metric": 0.0006942471372894943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0732844100002694
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.18314010673202574,
          "train_loss": 0.1797143816947937,
          "validation_loss": 0.028956076130270958,
          "train_metric": 0.04622969031333923,
          "validation_metric": 0.000452417996712029,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0859269760003372
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.18031700188294053,
          "train_loss": 0.17958149313926697,
          "validation_loss": 0.028716275468468666,
          "train_metric": 0.045992009341716766,
          "validation_metric": 0.0004242336144670844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0982583869990776
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.18042015470564365,
          "train_loss": 0.17735880613327026,
          "validation_loss": 0.025593653321266174,
          "train_metric": 0.04623640701174736,
          "validation_metric": 0.0004030067357234657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.110081327999069
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.18165927473455667,
          "train_loss": 0.1846410632133484,
          "validation_loss": 0.03167611360549927,
          "train_metric": 0.04596371203660965,
          "validation_metric": 0.0006307739531621337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1212787280001066
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.18340669269673526,
          "train_loss": 0.1802656352519989,
          "validation_loss": 0.026911333203315735,
          "train_metric": 0.046478211879730225,
          "validation_metric": 0.0003508703375700861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1331143280003744
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.18355254223570228,
          "train_loss": 0.18113458156585693,
          "validation_loss": 0.027740273624658585,
          "train_metric": 0.04630180075764656,
          "validation_metric": 0.00040691925096325576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1447122850004234
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.18159782839938998,
          "train_loss": 0.18589071929454803,
          "validation_loss": 0.035979513078927994,
          "train_metric": 0.04621158912777901,
          "validation_metric": 0.000462769326986745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.155886141999872
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.18576559307985008,
          "train_loss": 0.18224859237670898,
          "validation_loss": 0.027977824211120605,
          "train_metric": 0.04642326012253761,
          "validation_metric": 0.0003643388336058706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1670935250003822
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.18685094779357314,
          "train_loss": 0.1861763447523117,
          "validation_loss": 0.03375057131052017,
          "train_metric": 0.0467674545943737,
          "validation_metric": 0.0003681384841911495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1784888610000053
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.18300181557424366,
          "train_loss": 0.18700598180294037,
          "validation_loss": 0.034784577786922455,
          "train_metric": 0.04637022688984871,
          "validation_metric": 0.0005120155983604491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1896310079991963
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.183273688191548,
          "train_loss": 0.18019966781139374,
          "validation_loss": 0.028633354231715202,
          "train_metric": 0.046155091375112534,
          "validation_metric": 0.00036178133450448513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2010283809995599
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.17901617952156812,
          "train_loss": 0.17992445826530457,
          "validation_loss": 0.025024209171533585,
          "train_metric": 0.0460631437599659,
          "validation_metric": 0.0005361948278732598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.212441771000158
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.1779003869742155,
          "train_loss": 0.1826767772436142,
          "validation_loss": 0.029707346111536026,
          "train_metric": 0.046411145478487015,
          "validation_metric": 0.0004675239324569702,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.223529848999533
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.18172882369253784,
          "train_loss": 0.1844734251499176,
          "validation_loss": 0.0302409827709198,
          "train_metric": 0.04570641368627548,
          "validation_metric": 0.0006308951415121555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.234746592999727
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.18187939235940576,
          "train_loss": 0.19076992571353912,
          "validation_loss": 0.035349562764167786,
          "train_metric": 0.04730835556983948,
          "validation_metric": 0.0006753060151822865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2460358789994643
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.18453336972743273,
          "train_loss": 0.17952215671539307,
          "validation_loss": 0.030538946390151978,
          "train_metric": 0.045923467725515366,
          "validation_metric": 0.00042026268783956766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2572709559990471
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.17897962220013142,
          "train_loss": 0.18406681716442108,
          "validation_loss": 0.031209155917167664,
          "train_metric": 0.04580949991941452,
          "validation_metric": 0.0003558663884177804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2684147130003112
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.18517122115008533,
          "train_loss": 0.17827649414539337,
          "validation_loss": 0.02720264345407486,
          "train_metric": 0.04603227600455284,
          "validation_metric": 0.000319474347634241,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.280092077999143
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.1800336861051619,
          "train_loss": 0.18364755809307098,
          "validation_loss": 0.02817002311348915,
          "train_metric": 0.04605654999613762,
          "validation_metric": 0.0006279092049226165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.291420597999604
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.17933833692222834,
          "train_loss": 0.17856694757938385,
          "validation_loss": 0.026139846071600914,
          "train_metric": 0.046372611075639725,
          "validation_metric": 0.0003215206670574844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3023357379988738
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.17871197825297713,
          "train_loss": 0.17959603667259216,
          "validation_loss": 0.029578616842627525,
          "train_metric": 0.04584839567542076,
          "validation_metric": 0.0004656993260141462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.31322085300053
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.17980400985106826,
          "train_loss": 0.17821303009986877,
          "validation_loss": 0.027551371604204178,
          "train_metric": 0.04619933292269707,
          "validation_metric": 0.0003189174458384514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3247567620001064
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.17784311820287257,
          "train_loss": 0.18072590231895447,
          "validation_loss": 0.02526767924427986,
          "train_metric": 0.045767586678266525,
          "validation_metric": 0.00041625607991591096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3359212270006537
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.17814900050871074,
          "train_loss": 0.17355048656463623,
          "validation_loss": 0.021502889692783356,
          "train_metric": 0.046045634895563126,
          "validation_metric": 0.00025980875943787396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3474662699991313
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.1766621433198452,
          "train_loss": 0.1786571890115738,
          "validation_loss": 0.02595473825931549,
          "train_metric": 0.04604605585336685,
          "validation_metric": 0.000413834466598928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3582778649997636
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.17686200561001897,
          "train_loss": 0.17257821559906006,
          "validation_loss": 0.018422206863760948,
          "train_metric": 0.0460229367017746,
          "validation_metric": 0.00027551595121622086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3695082350004668
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.1748776687309146,
          "train_loss": 0.1753491759300232,
          "validation_loss": 0.02236531861126423,
          "train_metric": 0.045956455171108246,
          "validation_metric": 0.00034599934588186443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3805655910000496
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.1756633969489485,
          "train_loss": 0.17693021893501282,
          "validation_loss": 0.024023093283176422,
          "train_metric": 0.045776598155498505,
          "validation_metric": 0.000443215569248423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3918367229998694
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.17587867309339345,
          "train_loss": 0.17514191567897797,
          "validation_loss": 0.02272060513496399,
          "train_metric": 0.04599231109023094,
          "validation_metric": 0.0002684393257368356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.403307099999438
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.17785494308918715,
          "train_loss": 0.1740894615650177,
          "validation_loss": 0.022190727293491364,
          "train_metric": 0.04577968269586563,
          "validation_metric": 0.000322028121445328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4146020589996624
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.17731111915782094,
          "train_loss": 0.17932912707328796,
          "validation_loss": 0.02510167844593525,
          "train_metric": 0.04648635536432266,
          "validation_metric": 0.00025626804563216865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4269598129994847
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.17787783755920827,
          "train_loss": 0.1771600842475891,
          "validation_loss": 0.02338813804090023,
          "train_metric": 0.04617192596197128,
          "validation_metric": 0.0003743017732631415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.438242688000173
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.17711662163492292,
          "train_loss": 0.17478083074092865,
          "validation_loss": 0.022321170195937157,
          "train_metric": 0.04576057195663452,
          "validation_metric": 0.0002643316111061722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4495657000006759
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.17761714989319444,
          "train_loss": 0.17469049990177155,
          "validation_loss": 0.022921696305274963,
          "train_metric": 0.04574398323893547,
          "validation_metric": 0.000293548044282943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4609766939993278
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.17502721399068832,
          "train_loss": 0.17315886914730072,
          "validation_loss": 0.020734498277306557,
          "train_metric": 0.046044010668992996,
          "validation_metric": 0.0002925422159023583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4720332180004334
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.17373080644756556,
          "train_loss": 0.17357762157917023,
          "validation_loss": 0.02067199908196926,
          "train_metric": 0.04589242488145828,
          "validation_metric": 0.0002664825296960771,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.483151796999664
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.17391886096447706,
          "train_loss": 0.17870838940143585,
          "validation_loss": 0.02715938724577427,
          "train_metric": 0.04614507406949997,
          "validation_metric": 0.0004556428757496178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.494327192000128
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.17702864622697234,
          "train_loss": 0.1785779595375061,
          "validation_loss": 0.025108326226472855,
          "train_metric": 0.04577726125717163,
          "validation_metric": 0.0002580287691671401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5052707260001625
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.17601903853937984,
          "train_loss": 0.178035169839859,
          "validation_loss": 0.02549021877348423,
          "train_metric": 0.04639187827706337,
          "validation_metric": 0.0002383173705311492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5173180469992076
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.17755870008841157,
          "train_loss": 0.1789054423570633,
          "validation_loss": 0.023532480001449585,
          "train_metric": 0.04595733433961868,
          "validation_metric": 0.00021295438637025654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5293670760001987
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.1766331137623638,
          "train_loss": 0.1767653375864029,
          "validation_loss": 0.022254709154367447,
          "train_metric": 0.045769695192575455,
          "validation_metric": 0.00033237607567571104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5404165099989768
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.17825521202757955,
          "train_loss": 0.17693763971328735,
          "validation_loss": 0.022924620658159256,
          "train_metric": 0.046279992908239365,
          "validation_metric": 0.0002741432108450681,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5514214310005627
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.17821543850004673,
          "train_loss": 0.17806348204612732,
          "validation_loss": 0.02621757797896862,
          "train_metric": 0.045864757150411606,
          "validation_metric": 0.0004079557547811419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.562679021999429
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.17881905753165483,
          "train_loss": 0.17714647948741913,
          "validation_loss": 0.024467941373586655,
          "train_metric": 0.046353988349437714,
          "validation_metric": 0.00022596298367716372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.573758505999649
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.1793964677490294,
          "train_loss": 0.17210985720157623,
          "validation_loss": 0.016622792929410934,
          "train_metric": 0.04600471258163452,
          "validation_metric": 0.000195853688637726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5853190859997994
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.17535824328660965,
          "train_loss": 0.17368397116661072,
          "validation_loss": 0.019755620509386063,
          "train_metric": 0.04602806270122528,
          "validation_metric": 0.00021406296582426876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5965565589995094
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.17555687064304948,
          "train_loss": 0.17489582300186157,
          "validation_loss": 0.023512227460741997,
          "train_metric": 0.046340081840753555,
          "validation_metric": 0.00020633878011722118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6074454419995163
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.1740437764674425,
          "train_loss": 0.17691878974437714,
          "validation_loss": 0.024682603776454926,
          "train_metric": 0.04650333896279335,
          "validation_metric": 0.00022151942539494485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6186755879989505
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.17464360455051064,
          "train_loss": 0.1747978776693344,
          "validation_loss": 0.022711055353283882,
          "train_metric": 0.04574066027998924,
          "validation_metric": 0.0002268025273224339,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6296815099995001
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.17236397811211646,
          "train_loss": 0.17072683572769165,
          "validation_loss": 0.01625427044928074,
          "train_metric": 0.0461541973054409,
          "validation_metric": 0.00018466086476109922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6412943550003547
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.17223563231527805,
          "train_loss": 0.17323197424411774,
          "validation_loss": 0.020052654668688774,
          "train_metric": 0.04593968763947487,
          "validation_metric": 0.00019092447473667562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6528081859996746
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.17355623666662723,
          "train_loss": 0.1757117062807083,
          "validation_loss": 0.0234364066272974,
          "train_metric": 0.04569056257605553,
          "validation_metric": 0.00024669544654898345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.66398355399906
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.17839423706755042,
          "train_loss": 0.17182281613349915,
          "validation_loss": 0.018507421016693115,
          "train_metric": 0.045992378145456314,
          "validation_metric": 0.00018683119560591877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6751102859998355
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.1755782295949757,
          "train_loss": 0.17139241099357605,
          "validation_loss": 0.018213633447885513,
          "train_metric": 0.046077486127614975,
          "validation_metric": 0.00019007663649972528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6864579539997067
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.1763691403903067,
          "train_loss": 0.17054946720600128,
          "validation_loss": 0.01762811467051506,
          "train_metric": 0.04603373259305954,
          "validation_metric": 0.00017865284462459385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.698347104998902
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.1743346550501883,
          "train_loss": 0.17551957070827484,
          "validation_loss": 0.022097671404480934,
          "train_metric": 0.04632260650396347,
          "validation_metric": 0.0002087239845423028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.709429051999905
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.17822576547041535,
          "train_loss": 0.17366713285446167,
          "validation_loss": 0.02261810004711151,
          "train_metric": 0.04610065743327141,
          "validation_metric": 0.0002504323492757976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7205809520000912
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.1765668501611799,
          "train_loss": 0.1784449964761734,
          "validation_loss": 0.0277756005525589,
          "train_metric": 0.04674215987324715,
          "validation_metric": 0.0002713730791583657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7318578550002712
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.17642429133411497,
          "train_loss": 0.17264841496944427,
          "validation_loss": 0.02087586186826229,
          "train_metric": 0.045797646045684814,
          "validation_metric": 0.00019994146714452654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7428392059991893
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.1721462057903409,
          "train_loss": 0.1696324497461319,
          "validation_loss": 0.018080420792102814,
          "train_metric": 0.045915793627500534,
          "validation_metric": 0.00023302601766772568,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7541674320000311
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.17317938106134534,
          "train_loss": 0.17191952466964722,
          "validation_loss": 0.018025105819106102,
          "train_metric": 0.04613355174660683,
          "validation_metric": 0.00018782843835651875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7657578319995082
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.1716738543473184,
          "train_loss": 0.17474836111068726,
          "validation_loss": 0.02272423729300499,
          "train_metric": 0.04559396207332611,
          "validation_metric": 0.0002474552020430565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.776656980999178
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.17347601940855384,
          "train_loss": 0.1723286509513855,
          "validation_loss": 0.01974973827600479,
          "train_metric": 0.04614565521478653,
          "validation_metric": 0.00019998614152427763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.787648707999324
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.17739698151126504,
          "train_loss": 0.1837998628616333,
          "validation_loss": 0.031065262854099274,
          "train_metric": 0.04560789093375206,
          "validation_metric": 0.00023984888684935868,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7986705690000235
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.18009601859375834,
          "train_loss": 0.1805107742547989,
          "validation_loss": 0.02875438891351223,
          "train_metric": 0.0456937737762928,
          "validation_metric": 0.00039861921686679125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8097097650006617
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.1827611131593585,
          "train_loss": 0.1898418813943863,
          "validation_loss": 0.033136237412691116,
          "train_metric": 0.04744177684187889,
          "validation_metric": 0.0005935622612014413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8208856290002586
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.18098042346537113,
          "train_loss": 0.18578067421913147,
          "validation_loss": 0.03692229092121124,
          "train_metric": 0.046064555644989014,
          "validation_metric": 0.000495371816214174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8320056310003565
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.1785223549231887,
          "train_loss": 0.17692282795906067,
          "validation_loss": 0.022990822792053223,
          "train_metric": 0.04614555835723877,
          "validation_metric": 0.0001974760089069605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.842877139999473
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.1774408002384007,
          "train_loss": 0.1772485226392746,
          "validation_loss": 0.02455056644976139,
          "train_metric": 0.0460335947573185,
          "validation_metric": 0.00020031738677062094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8537736189991847
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.17579340434167534,
          "train_loss": 0.1720133125782013,
          "validation_loss": 0.019098052754998207,
          "train_metric": 0.04594555124640465,
          "validation_metric": 0.00016202540427912027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.865203751000081
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.1750105395913124,
          "train_loss": 0.1702871471643448,
          "validation_loss": 0.017835041508078575,
          "train_metric": 0.046113189309835434,
          "validation_metric": 0.0001651980564929545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8762384149995341
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.17315029411111027,
          "train_loss": 0.1775672286748886,
          "validation_loss": 0.025447160005569458,
          "train_metric": 0.04590645432472229,
          "validation_metric": 0.0002813817118294537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8875332799998432
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.173821386648342,
          "train_loss": 0.16988909244537354,
          "validation_loss": 0.018417315557599068,
          "train_metric": 0.04606218263506889,
          "validation_metric": 0.00017680325254332274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8989177490002476
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.17514622397720814,
          "train_loss": 0.1719079166650772,
          "validation_loss": 0.018968163058161736,
          "train_metric": 0.04638604447245598,
          "validation_metric": 0.0002110595378326252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9098906489998626
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.17181104514747858,
          "train_loss": 0.17324641346931458,
          "validation_loss": 0.023597512394189835,
          "train_metric": 0.04591693729162216,
          "validation_metric": 0.00020408839918673038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9206595470004686
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.1723056110786274,
          "train_loss": 0.17315807938575745,
          "validation_loss": 0.021366490051150322,
          "train_metric": 0.04636723920702934,
          "validation_metric": 0.0002340788341825828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9313674700006231
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.17347309365868568,
          "train_loss": 0.17130354046821594,
          "validation_loss": 0.01811137981712818,
          "train_metric": 0.04606943577528,
          "validation_metric": 0.0002201858878834173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9415071459989122
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.17030672170221806,
          "train_loss": 0.17055734992027283,
          "validation_loss": 0.018043378368020058,
          "train_metric": 0.0457521416246891,
          "validation_metric": 0.00026861822698265314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.952030062000631
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.1717013157904148,
          "train_loss": 0.16807912290096283,
          "validation_loss": 0.015536642633378506,
          "train_metric": 0.04592816159129143,
          "validation_metric": 0.00017572053184267133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.962329869998939
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.1744023123756051,
          "train_loss": 0.17863836884498596,
          "validation_loss": 0.02427435666322708,
          "train_metric": 0.045941513031721115,
          "validation_metric": 0.00017482567636761814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9723330050001096
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.1769816225860268,
          "train_loss": 0.17930923402309418,
          "validation_loss": 0.024027464911341667,
          "train_metric": 0.04576132819056511,
          "validation_metric": 0.00019094177696388215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.98274900800061
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.17584528867155313,
          "train_loss": 0.17227384448051453,
          "validation_loss": 0.01912374049425125,
          "train_metric": 0.04594222456216812,
          "validation_metric": 0.00015770464960951358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9938570499998605
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.17114606616087258,
          "train_loss": 0.17156754434108734,
          "validation_loss": 0.016537705436348915,
          "train_metric": 0.04636998102068901,
          "validation_metric": 0.00015369110042229295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0050153199990746
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.17086674587335438,
          "train_loss": 0.1710701584815979,
          "validation_loss": 0.019240429624915123,
          "train_metric": 0.045936375856399536,
          "validation_metric": 0.0001586080325068906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0157550389994867
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.1705409251153469,
          "train_loss": 0.1687498539686203,
          "validation_loss": 0.015790734440088272,
          "train_metric": 0.04608285799622536,
          "validation_metric": 0.00015164451906457543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.026546717999736
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.1697879123967141,
          "train_loss": 0.16936029493808746,
          "validation_loss": 0.016922468319535255,
          "train_metric": 0.04579656571149826,
          "validation_metric": 0.0001869300613179803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.036890092998874
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.17240739660337567,
          "train_loss": 0.17316791415214539,
          "validation_loss": 0.023544292896986008,
          "train_metric": 0.046232860535383224,
          "validation_metric": 0.00020663176837842911,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0472279410005285
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.1766290864907205,
          "train_loss": 0.1830231100320816,
          "validation_loss": 0.03188375383615494,
          "train_metric": 0.04539351537823677,
          "validation_metric": 0.00020508519082795829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0574075109998375
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.17878695158287883,
          "train_loss": 0.17623531818389893,
          "validation_loss": 0.02205084078013897,
          "train_metric": 0.045733824372291565,
          "validation_metric": 0.00016106045222841203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0678837849991396
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.17440671287477016,
          "train_loss": 0.1729792356491089,
          "validation_loss": 0.018108472228050232,
          "train_metric": 0.046060968190431595,
          "validation_metric": 0.00014771305723115802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.078788953000185
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.172686209785752,
          "train_loss": 0.17380371689796448,
          "validation_loss": 0.021356074139475822,
          "train_metric": 0.045993201434612274,
          "validation_metric": 0.00019779021386057138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0890940760000376
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.1728067232761532,
          "train_loss": 0.17099353671073914,
          "validation_loss": 0.01947329007089138,
          "train_metric": 0.045782420784235,
          "validation_metric": 0.00018959496810566634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0999766160002764
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.17193971417145804,
          "train_loss": 0.17552350461483002,
          "validation_loss": 0.019902877509593964,
          "train_metric": 0.046358853578567505,
          "validation_metric": 0.00022975758474785835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.110508750000008
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.1728199371136725,
          "train_loss": 0.17482894659042358,
          "validation_loss": 0.025569526478648186,
          "train_metric": 0.046322014182806015,
          "validation_metric": 0.00019479628826957196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1209355520004465
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.1710537092294544,
          "train_loss": 0.1730574667453766,
          "validation_loss": 0.01909671723842621,
          "train_metric": 0.046201810240745544,
          "validation_metric": 0.00014185057079885155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1322459500006516
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.17362511123064905,
          "train_loss": 0.16918392479419708,
          "validation_loss": 0.015400843694806099,
          "train_metric": 0.04581780731678009,
          "validation_metric": 0.00017410406144335866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.143037989999357
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.17077337787486613,
          "train_loss": 0.1710163652896881,
          "validation_loss": 0.0188590120524168,
          "train_metric": 0.04580321162939072,
          "validation_metric": 0.00018525360792409629,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1536438509992877
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.17158479732461274,
          "train_loss": 0.17601069808006287,
          "validation_loss": 0.0223116222769022,
          "train_metric": 0.04615240916609764,
          "validation_metric": 0.00015052627713885158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1641694429999916
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.17210989934392273,
          "train_loss": 0.1672418713569641,
          "validation_loss": 0.012930705212056637,
          "train_metric": 0.0459674708545208,
          "validation_metric": 0.00012244853132870048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1748163219999697
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.17062658723443747,
          "train_loss": 0.17166852951049805,
          "validation_loss": 0.019067006185650826,
          "train_metric": 0.04585281014442444,
          "validation_metric": 0.00017846078844740987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.185519361999468
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.1720340484753251,
          "train_loss": 0.1768781840801239,
          "validation_loss": 0.02255561575293541,
          "train_metric": 0.045810725539922714,
          "validation_metric": 0.00020207752822898328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.196090945999458
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.1720561790280044,
          "train_loss": 0.17768648266792297,
          "validation_loss": 0.025089958682656288,
          "train_metric": 0.04610249400138855,
          "validation_metric": 0.0003556045703589916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2064936909991957
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.17272160714492202,
          "train_loss": 0.1795913726091385,
          "validation_loss": 0.025477394461631775,
          "train_metric": 0.046468336135149,
          "validation_metric": 0.00020551812485791743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2171137539990013
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.17718125088140368,
          "train_loss": 0.17192178964614868,
          "validation_loss": 0.01848621480166912,
          "train_metric": 0.04597330838441849,
          "validation_metric": 0.00017988351464737207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2277472379992105
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.17449924582615495,
          "train_loss": 0.17788822948932648,
          "validation_loss": 0.023743705824017525,
          "train_metric": 0.04630975052714348,
          "validation_metric": 0.00015337171498686075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2380659649988957
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.17470539605710655,
          "train_loss": 0.179411381483078,
          "validation_loss": 0.027014099061489105,
          "train_metric": 0.04594315588474274,
          "validation_metric": 0.00025752800866030157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2488367360001575
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.17835619067773223,
          "train_loss": 0.1760888695716858,
          "validation_loss": 0.023533876985311508,
          "train_metric": 0.04642128944396973,
          "validation_metric": 0.00019601202802732587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.259341520999442
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.1757063965778798,
          "train_loss": 0.17630277574062347,
          "validation_loss": 0.022450974211096764,
          "train_metric": 0.04649054631590843,
          "validation_metric": 0.0001634287618799135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.269739274999665
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.17424672888591886,
          "train_loss": 0.17249442636966705,
          "validation_loss": 0.01937869004905224,
          "train_metric": 0.04587230458855629,
          "validation_metric": 0.00013768297503702343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.280010618000233
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.1718266293173656,
          "train_loss": 0.17256763577461243,
          "validation_loss": 0.018297988921403885,
          "train_metric": 0.04578786343336105,
          "validation_metric": 0.00013535632751882076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2898109319994546
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.17360808607190847,
          "train_loss": 0.1695900708436966,
          "validation_loss": 0.019087307155132294,
          "train_metric": 0.04609617963433266,
          "validation_metric": 0.0001924852840602398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3002557849995355
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.17252339702099562,
          "train_loss": 0.17136603593826294,
          "validation_loss": 0.017387637868523598,
          "train_metric": 0.046216510236263275,
          "validation_metric": 0.0001247378677362576,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3106114029997116
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.1727691856212914,
          "train_loss": 0.17542128264904022,
          "validation_loss": 0.023178160190582275,
          "train_metric": 0.0464683473110199,
          "validation_metric": 0.00022136069310363382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.320880617999137
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.16956017457414418,
          "train_loss": 0.1670360416173935,
          "validation_loss": 0.01462554931640625,
          "train_metric": 0.04585685953497887,
          "validation_metric": 0.00013602209219243377,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3311070349991496
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.16950787778478116,
          "train_loss": 0.1687038391828537,
          "validation_loss": 0.01606609858572483,
          "train_metric": 0.04623900353908539,
          "validation_metric": 0.00012015723041258752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.341454375999092
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.17041675583459437,
          "train_loss": 0.16753506660461426,
          "validation_loss": 0.014905039221048355,
          "train_metric": 0.04593249782919884,
          "validation_metric": 0.000144702717079781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3520044750002853
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.16889883030671626,
          "train_loss": 0.1700553447008133,
          "validation_loss": 0.01668946072459221,
          "train_metric": 0.04598049446940422,
          "validation_metric": 0.00019764019816648215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3625569209998503
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.16891204979037866,
          "train_loss": 0.16751505434513092,
          "validation_loss": 0.015329777263104916,
          "train_metric": 0.0460372231900692,
          "validation_metric": 0.00011854251351905987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.373043695999513
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.17097096843644977,
          "train_loss": 0.16859008371829987,
          "validation_loss": 0.013869480229914188,
          "train_metric": 0.046183645725250244,
          "validation_metric": 0.00011624045873759314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3840067860001
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.1711366195231676,
          "train_loss": 0.17238250374794006,
          "validation_loss": 0.021400101482868195,
          "train_metric": 0.04567611590027809,
          "validation_metric": 0.00020516883523669094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.394359108999197
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.17275395774049684,
          "train_loss": 0.17143207788467407,
          "validation_loss": 0.0189460888504982,
          "train_metric": 0.04604627937078476,
          "validation_metric": 0.00014183799794409424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4045860059995903
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.17731924401596189,
          "train_loss": 0.17375969886779785,
          "validation_loss": 0.021766874939203262,
          "train_metric": 0.045776885002851486,
          "validation_metric": 0.0002492281491868198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.414978618000532
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.1746066955383867,
          "train_loss": 0.16651614010334015,
          "validation_loss": 0.014009094797074795,
          "train_metric": 0.04597930237650871,
          "validation_metric": 0.00010937363549601287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4254669250003644
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.17013752972707152,
          "train_loss": 0.1711183786392212,
          "validation_loss": 0.01611095666885376,
          "train_metric": 0.04633181169629097,
          "validation_metric": 0.00012012016668450087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.43601941399902
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.17185566248372197,
          "train_loss": 0.17314840853214264,
          "validation_loss": 0.02214883640408516,
          "train_metric": 0.04630623012781143,
          "validation_metric": 0.00016505680105183274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4465212659997633
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.1757516919169575,
          "train_loss": 0.17887799441814423,
          "validation_loss": 0.0254904143512249,
          "train_metric": 0.04675819352269173,
          "validation_metric": 0.0002707828825805336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4565730710000935
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.1742791081778705,
          "train_loss": 0.1684924066066742,
          "validation_loss": 0.015315888449549675,
          "train_metric": 0.04606809467077255,
          "validation_metric": 0.0001215418815263547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4668199490006373
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.17447020951658487,
          "train_loss": 0.17337153851985931,
          "validation_loss": 0.021259188652038574,
          "train_metric": 0.045898791402578354,
          "validation_metric": 0.00017356911848764867,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.476762186999622
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.1719466911163181,
          "train_loss": 0.17401979863643646,
          "validation_loss": 0.021513041108846664,
          "train_metric": 0.046609871089458466,
          "validation_metric": 0.00019034072465728968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4867759690005187
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.17480330995749682,
          "train_loss": 0.17420241236686707,
          "validation_loss": 0.023118799552321434,
          "train_metric": 0.04606617987155914,
          "validation_metric": 0.0002739069750532508,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.497069234999799
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.17134911136236042,
          "train_loss": 0.16861079633235931,
          "validation_loss": 0.0163121335208416,
          "train_metric": 0.04598284885287285,
          "validation_metric": 0.00012943662295583636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5069713659995614
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.16955792030785233,
          "train_loss": 0.17072871327400208,
          "validation_loss": 0.019117698073387146,
          "train_metric": 0.04635016247630119,
          "validation_metric": 0.00011914469359908253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5174260219991993
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.17015331611037254,
          "train_loss": 0.16926415264606476,
          "validation_loss": 0.018157729879021645,
          "train_metric": 0.046077463775873184,
          "validation_metric": 0.00017409416614100337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.527649792000375
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.17277943587396294,
          "train_loss": 0.1799537092447281,
          "validation_loss": 0.028273236006498337,
          "train_metric": 0.045835867524147034,
          "validation_metric": 0.00025137682678177953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.537734145998911
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.1750162225216627,
          "train_loss": 0.18146799504756927,
          "validation_loss": 0.024610621854662895,
          "train_metric": 0.04609542340040207,
          "validation_metric": 0.00014760287012904882,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.547979373999624
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.1735907574184239,
          "train_loss": 0.16620121896266937,
          "validation_loss": 0.014303967356681824,
          "train_metric": 0.04603159800171852,
          "validation_metric": 0.00010114820906892419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.558519210999293
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.17023870046250522,
          "train_loss": 0.1729024201631546,
          "validation_loss": 0.018907636404037476,
          "train_metric": 0.04656310752034187,
          "validation_metric": 0.00017514260252937675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5691807179991883
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.17052870010957122,
          "train_loss": 0.1782306730747223,
          "validation_loss": 0.026829630136489868,
          "train_metric": 0.045591067522764206,
          "validation_metric": 0.0003282261313870549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5792859110006248
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.1764840332325548,
          "train_loss": 0.1837332397699356,
          "validation_loss": 0.030420435592532158,
          "train_metric": 0.04703916609287262,
          "validation_metric": 0.0003949553647544235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.589221701000497
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.17291891248896718,
          "train_loss": 0.1713765412569046,
          "validation_loss": 0.01818002015352249,
          "train_metric": 0.045987002551555634,
          "validation_metric": 0.00013221918197814375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.599554669999634
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.17049438296817243,
          "train_loss": 0.17120955884456635,
          "validation_loss": 0.019430022686719894,
          "train_metric": 0.04644930735230446,
          "validation_metric": 0.0001242986327270046,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.609467455999038
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.1696013220353052,
          "train_loss": 0.1742880642414093,
          "validation_loss": 0.02511124685406685,
          "train_metric": 0.04632382467389107,
          "validation_metric": 0.00016776815755292773,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6197293879995414
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.17082175344694406,
          "train_loss": 0.16945037245750427,
          "validation_loss": 0.014790738932788372,
          "train_metric": 0.0461762361228466,
          "validation_metric": 0.0001300629519391805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6299285290006083
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.17079384066164494,
          "train_loss": 0.1737791746854782,
          "validation_loss": 0.018806753680109978,
          "train_metric": 0.046576205641031265,
          "validation_metric": 0.00019080215133726597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.639848659000563
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.1699180081486702,
          "train_loss": 0.1671326458454132,
          "validation_loss": 0.012468130327761173,
          "train_metric": 0.04594101011753082,
          "validation_metric": 0.00010863152419915423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.649943684999016
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.17221017787232995,
          "train_loss": 0.169136181473732,
          "validation_loss": 0.014916656538844109,
          "train_metric": 0.0457993783056736,
          "validation_metric": 0.00015724639524705708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.660019723000005
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.16982907080091536,
          "train_loss": 0.16757096350193024,
          "validation_loss": 0.013391591608524323,
          "train_metric": 0.04586384445428848,
          "validation_metric": 0.0001073753519449383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.670168078999268
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.16912426060298458,
          "train_loss": 0.1689092516899109,
          "validation_loss": 0.01447087712585926,
          "train_metric": 0.046214960515499115,
          "validation_metric": 0.00010420975740998983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.680832209000073
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.17012868530582637,
          "train_loss": 0.17266952991485596,
          "validation_loss": 0.022626712918281555,
          "train_metric": 0.045965977013111115,
          "validation_metric": 0.00015024928143247962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6907838829993125
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.16918828088091686,
          "train_loss": 0.1669912040233612,
          "validation_loss": 0.013510506600141525,
          "train_metric": 0.046146173030138016,
          "validation_metric": 9.546894580125809e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.702157232999525
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.16775815817527473,
          "train_loss": 0.16873909533023834,
          "validation_loss": 0.014574809931218624,
          "train_metric": 0.04574589431285858,
          "validation_metric": 0.00012601270282175392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7126025630004733
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.1680963517865166,
          "train_loss": 0.17030154168605804,
          "validation_loss": 0.014385655522346497,
          "train_metric": 0.046144597232341766,
          "validation_metric": 0.00011444716801634058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.722809306000272
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.17079644557088614,
          "train_loss": 0.17089559137821198,
          "validation_loss": 0.015603764913976192,
          "train_metric": 0.04612616449594498,
          "validation_metric": 0.0001581053074914962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7333117779999156
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.1710938038304448,
          "train_loss": 0.17161446809768677,
          "validation_loss": 0.017661890015006065,
          "train_metric": 0.04629315808415413,
          "validation_metric": 0.0001128949734265916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7435954270004004
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.17086272267624736,
          "train_loss": 0.1701539009809494,
          "validation_loss": 0.017019089311361313,
          "train_metric": 0.04587104544043541,
          "validation_metric": 0.00010379092418588698,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7537089540001034
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.17285438370890915,
          "train_loss": 0.17472383379936218,
          "validation_loss": 0.02121630497276783,
          "train_metric": 0.046417735517024994,
          "validation_metric": 0.0001231234346050769,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7639811529988947
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.17017694772221148,
          "train_loss": 0.17106467485427856,
          "validation_loss": 0.02097029983997345,
          "train_metric": 0.04582098871469498,
          "validation_metric": 0.0002556410036049783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.773994949000553
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.17055958311539143,
          "train_loss": 0.17090603709220886,
          "validation_loss": 0.017548400908708572,
          "train_metric": 0.046280018985271454,
          "validation_metric": 0.00011445321433711797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.784809733999282
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.17398936580866575,
          "train_loss": 0.17237187922000885,
          "validation_loss": 0.02041216753423214,
          "train_metric": 0.0462271124124527,
          "validation_metric": 0.0001662171125644818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.795216279999295
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.17333926528226584,
          "train_loss": 0.17392882704734802,
          "validation_loss": 0.01997162215411663,
          "train_metric": 0.04607423394918442,
          "validation_metric": 0.00012509214866440743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8051759659992967
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.17150004021823406,
          "train_loss": 0.1717405766248703,
          "validation_loss": 0.019696827977895737,
          "train_metric": 0.04604492709040642,
          "validation_metric": 0.00016432511620223522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.815457568000056
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.1713776420801878,
          "train_loss": 0.17012596130371094,
          "validation_loss": 0.017529137432575226,
          "train_metric": 0.04578666016459465,
          "validation_metric": 0.0001754980767145753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8256440080003813
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.16959022404626012,
          "train_loss": 0.16940729320049286,
          "validation_loss": 0.013406787067651749,
          "train_metric": 0.046283695846796036,
          "validation_metric": 0.00013941621000412852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.835778978000235
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.1683334065019153,
          "train_loss": 0.1711612194776535,
          "validation_loss": 0.016898276284337044,
          "train_metric": 0.04599159210920334,
          "validation_metric": 0.00022443731722887605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8458903949995147
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.17401843797415495,
          "train_loss": 0.18017281591892242,
          "validation_loss": 0.025281637907028198,
          "train_metric": 0.04682229831814766,
          "validation_metric": 0.00016646384028717875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.855683592999412
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.1751992084318772,
          "train_loss": 0.17133478820323944,
          "validation_loss": 0.017399931326508522,
          "train_metric": 0.045882850885391235,
          "validation_metric": 0.00010749766079243273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8660433559998637
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.17111357406247407,
          "train_loss": 0.16522058844566345,
          "validation_loss": 0.014476202428340912,
          "train_metric": 0.04587815701961517,
          "validation_metric": 0.00011142249422846362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8761170559992024
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.16968342452310026,
          "train_loss": 0.17329557240009308,
          "validation_loss": 0.019852520897984505,
          "train_metric": 0.04550950974225998,
          "validation_metric": 0.0001910839055199176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8864180599994143
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.17413217015564442,
          "train_loss": 0.18820971250534058,
          "validation_loss": 0.02970125339925289,
          "train_metric": 0.04707365855574608,
          "validation_metric": 0.00047902000369504094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.896678652999981
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.17600318789482117,
          "train_loss": 0.17929500341415405,
          "validation_loss": 0.028559789061546326,
          "train_metric": 0.04562216252088547,
          "validation_metric": 0.0005245792563073337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9066295030006586
        }
      ],
      "best": {
        "epoch": 218,
        "updates": 3488
      },
      "best_validation_metric": 9.546894580125809e-05,
      "config": {
        "label": "MAE + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mae",
        "outliers": true,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009770821142629624
    },
    "772fcca85495_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.1509843599051237,
          "train_loss": 0.1091870442032814,
          "validation_loss": 0.07135053724050522,
          "train_metric": 0.08967095613479614,
          "validation_metric": 0.031792886555194855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.025248463000025367
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.07435780041851103,
          "train_loss": 0.058548104017972946,
          "validation_loss": 0.029746737331151962,
          "train_metric": 0.05592314153909683,
          "validation_metric": 0.01202777773141861,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05047561600076733
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.05426082748454064,
          "train_loss": 0.051037609577178955,
          "validation_loss": 0.023354142904281616,
          "train_metric": 0.05529759079217911,
          "validation_metric": 0.012401859275996685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07836227699954179
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.047890143003314734,
          "train_loss": 0.04583452641963959,
          "validation_loss": 0.01670289970934391,
          "train_metric": 0.052653588354587555,
          "validation_metric": 0.007944162003695965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09462584399989282
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.045362746459431946,
          "train_loss": 0.04462675005197525,
          "validation_loss": 0.01551012322306633,
          "train_metric": 0.05162627995014191,
          "validation_metric": 0.007071496453136206,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10711306600023818
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.04428895364981145,
          "train_loss": 0.044281065464019775,
          "validation_loss": 0.015361291356384754,
          "train_metric": 0.05055435746908188,
          "validation_metric": 0.0062365964986383915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11939887100015767
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.044347371091134846,
          "train_loss": 0.044827863574028015,
          "validation_loss": 0.015588768757879734,
          "train_metric": 0.0506867840886116,
          "validation_metric": 0.005771577823907137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13179513800059794
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.044222898373845965,
          "train_loss": 0.043255239725112915,
          "validation_loss": 0.014420890249311924,
          "train_metric": 0.05014745146036148,
          "validation_metric": 0.005874066613614559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14326679199984937
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04310482554137707,
          "train_loss": 0.04248005896806717,
          "validation_loss": 0.013293217867612839,
          "train_metric": 0.05052916333079338,
          "validation_metric": 0.005558382719755173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15538897199985513
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.042594604135956615,
          "train_loss": 0.042979612946510315,
          "validation_loss": 0.013229373842477798,
          "train_metric": 0.051001276820898056,
          "validation_metric": 0.004696185700595379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16769238300003053
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04237279936205596,
          "train_loss": 0.04257906228303909,
          "validation_loss": 0.012461750768125057,
          "train_metric": 0.05078520253300667,
          "validation_metric": 0.00438260892406106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18029448199922626
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04172906617168337,
          "train_loss": 0.04200659692287445,
          "validation_loss": 0.012859873473644257,
          "train_metric": 0.0495140515267849,
          "validation_metric": 0.0051588332280516624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1919571079997695
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.041652265237644315,
          "train_loss": 0.04078488424420357,
          "validation_loss": 0.011081580072641373,
          "train_metric": 0.049646805971860886,
          "validation_metric": 0.003743993118405342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20422083400080737
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.04040493146749213,
          "train_loss": 0.03960121050477028,
          "validation_loss": 0.010274384170770645,
          "train_metric": 0.04893498867750168,
          "validation_metric": 0.004008551128208637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21593020900036208
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03952821058919653,
          "train_loss": 0.04131082445383072,
          "validation_loss": 0.010992339812219143,
          "train_metric": 0.04984813183546066,
          "validation_metric": 0.0032291742973029613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2280957090006268
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.03950085083488375,
          "train_loss": 0.038575392216444016,
          "validation_loss": 0.00907911453396082,
          "train_metric": 0.04836330562829971,
          "validation_metric": 0.003116786014288664,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2405001290007931
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03902744210790843,
          "train_loss": 0.038495540618896484,
          "validation_loss": 0.00868136901408434,
          "train_metric": 0.04864988476037979,
          "validation_metric": 0.0027106129564344883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25284318599915423
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.03922776260878891,
          "train_loss": 0.0410037524998188,
          "validation_loss": 0.012060548178851604,
          "train_metric": 0.047845665365457535,
          "validation_metric": 0.0045993695966899395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26466896499914583
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.03801533149089664,
          "train_loss": 0.03736613690853119,
          "validation_loss": 0.00750691769644618,
          "train_metric": 0.04833994433283806,
          "validation_metric": 0.002601856365799904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2765416780002852
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.03731139324372634,
          "train_loss": 0.036896418780088425,
          "validation_loss": 0.007259875535964966,
          "train_metric": 0.04809170588850975,
          "validation_metric": 0.002600583480671048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28837650300010864
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.03793109791877214,
          "train_loss": 0.036666080355644226,
          "validation_loss": 0.007598461117595434,
          "train_metric": 0.04729407653212547,
          "validation_metric": 0.0025110128335654736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3002267470001243
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.036871487041935325,
          "train_loss": 0.036815542727708817,
          "validation_loss": 0.007272700779139996,
          "train_metric": 0.04747869074344635,
          "validation_metric": 0.0023322647903114557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31217963000017335
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.03768885612953454,
          "train_loss": 0.036277007311582565,
          "validation_loss": 0.0066861240193247795,
          "train_metric": 0.0477985255420208,
          "validation_metric": 0.002034620149061084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3245317300006718
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.03658458369318396,
          "train_loss": 0.03612127527594566,
          "validation_loss": 0.006423489656299353,
          "train_metric": 0.047851189970970154,
          "validation_metric": 0.001934063620865345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3369063139998616
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.036320292041637,
          "train_loss": 0.03586135804653168,
          "validation_loss": 0.006463924422860146,
          "train_metric": 0.04730145260691643,
          "validation_metric": 0.0020784097723662853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3485600780004461
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.036321464052889496,
          "train_loss": 0.035576097667217255,
          "validation_loss": 0.006424356251955032,
          "train_metric": 0.0469735786318779,
          "validation_metric": 0.001956039573997259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3593410329995095
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.03603135235607624,
          "train_loss": 0.03524567931890488,
          "validation_loss": 0.005876734387129545,
          "train_metric": 0.04714551940560341,
          "validation_metric": 0.001824917970225215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37082618400017964
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.03583337552845478,
          "train_loss": 0.035437487065792084,
          "validation_loss": 0.006384360603988171,
          "train_metric": 0.04672074317932129,
          "validation_metric": 0.0020095340441912413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.38209271999949124
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.03631594526814297,
          "train_loss": 0.035865627229213715,
          "validation_loss": 0.007029848173260689,
          "train_metric": 0.04655224829912186,
          "validation_metric": 0.0022344328463077545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3932066910001595
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.03595336733269505,
          "train_loss": 0.038917671889066696,
          "validation_loss": 0.00861446838825941,
          "train_metric": 0.049685776233673096,
          "validation_metric": 0.0020864943508058786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.40425899000001664
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.03854939673328772,
          "train_loss": 0.03519188612699509,
          "validation_loss": 0.005918712355196476,
          "train_metric": 0.04667738452553749,
          "validation_metric": 0.0016559022478759289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4160062730006757
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.035925466916523874,
          "train_loss": 0.035862330347299576,
          "validation_loss": 0.006602270994335413,
          "train_metric": 0.04664704576134682,
          "validation_metric": 0.0022526767570525408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42721774200072105
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.03507105940661859,
          "train_loss": 0.03497739136219025,
          "validation_loss": 0.005506044253706932,
          "train_metric": 0.04681597277522087,
          "validation_metric": 0.0017409137217327952,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43829114899926935
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.03481864515924826,
          "train_loss": 0.034807730466127396,
          "validation_loss": 0.005100584123283625,
          "train_metric": 0.04731817916035652,
          "validation_metric": 0.0016200003447011113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45012102500004403
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.03492335689952597,
          "train_loss": 0.034633733332157135,
          "validation_loss": 0.004784820601344109,
          "train_metric": 0.047566480934619904,
          "validation_metric": 0.001299576717428863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4617440580004768
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.03573898220201954,
          "train_loss": 0.03511504456400871,
          "validation_loss": 0.005954718682914972,
          "train_metric": 0.0461701974272728,
          "validation_metric": 0.0018570307875052094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4731048930007091
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.03431727894349024,
          "train_loss": 0.034109991043806076,
          "validation_loss": 0.004777577240020037,
          "train_metric": 0.04677891731262207,
          "validation_metric": 0.0012463110033422709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4849012009999569
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.034564262721687555,
          "train_loss": 0.03419986739754677,
          "validation_loss": 0.005085825454443693,
          "train_metric": 0.04634220525622368,
          "validation_metric": 0.001496786717325449,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4960619850007788
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.03406300663482398,
          "train_loss": 0.03379553556442261,
          "validation_loss": 0.004455117974430323,
          "train_metric": 0.04647020623087883,
          "validation_metric": 0.0013013632269576192,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.50716390100024
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.03397383711126167,
          "train_loss": 0.03377070277929306,
          "validation_loss": 0.004159391857683659,
          "train_metric": 0.04685716703534126,
          "validation_metric": 0.001180805847980082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6019951920006861
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.0342369248974137,
          "train_loss": 0.03426253795623779,
          "validation_loss": 0.00482722045853734,
          "train_metric": 0.046389929950237274,
          "validation_metric": 0.0015260869404301047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6131833910003479
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.033966681687161326,
          "train_loss": 0.033882662653923035,
          "validation_loss": 0.004272957798093557,
          "train_metric": 0.04700136184692383,
          "validation_metric": 0.001028174185194075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6246446980003384
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.034100591379683465,
          "train_loss": 0.03356897830963135,
          "validation_loss": 0.0041090622544288635,
          "train_metric": 0.04623580724000931,
          "validation_metric": 0.0011518677929416299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6359012110006006
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.034645955107407644,
          "train_loss": 0.03394952416419983,
          "validation_loss": 0.003939626272767782,
          "train_metric": 0.04739244282245636,
          "validation_metric": 0.0010323842288926244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6472142200000235
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.03468171041458845,
          "train_loss": 0.03347356244921684,
          "validation_loss": 0.004195353947579861,
          "train_metric": 0.046028267592191696,
          "validation_metric": 0.0011511124903336167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6583004639996943
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.03364610148128122,
          "train_loss": 0.03344693407416344,
          "validation_loss": 0.0037084855139255524,
          "train_metric": 0.04699964076280594,
          "validation_metric": 0.001012514578178525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6699356289991556
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.03346014674752951,
          "train_loss": 0.03338612988591194,
          "validation_loss": 0.0038247944321483374,
          "train_metric": 0.04698490351438522,
          "validation_metric": 0.0008845629636198282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6816278909991524
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.033378493040800095,
          "train_loss": 0.03355398774147034,
          "validation_loss": 0.004102626349776983,
          "train_metric": 0.046856775879859924,
          "validation_metric": 0.0009146247757598758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6927817510004388
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.03347741375910118,
          "train_loss": 0.03306224197149277,
          "validation_loss": 0.0036350192967802286,
          "train_metric": 0.04670651629567146,
          "validation_metric": 0.0009024594910442829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7037989210002706
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.033298159192781895,
          "train_loss": 0.03312031179666519,
          "validation_loss": 0.003581533906981349,
          "train_metric": 0.046937063336372375,
          "validation_metric": 0.0008179531432688236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7154365719998168
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.033228220127057284,
          "train_loss": 0.03299107775092125,
          "validation_loss": 0.003650466911494732,
          "train_metric": 0.04641888663172722,
          "validation_metric": 0.0008225632482208312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7264611449991207
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.03340948396362364,
          "train_loss": 0.03503003716468811,
          "validation_loss": 0.005554096307605505,
          "train_metric": 0.045957230031490326,
          "validation_metric": 0.0017127125756815076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7374590399995213
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.03387671837117523,
          "train_loss": 0.03379443287849426,
          "validation_loss": 0.00442791823297739,
          "train_metric": 0.04632730409502983,
          "validation_metric": 0.0012957837898284197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7484156129994517
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.03307683678576723,
          "train_loss": 0.033420152962207794,
          "validation_loss": 0.003965201787650585,
          "train_metric": 0.04683457314968109,
          "validation_metric": 0.0008671280811540782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.759263877000194
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.03380429663229734,
          "train_loss": 0.03352911025285721,
          "validation_loss": 0.004677293356508017,
          "train_metric": 0.04601014778017998,
          "validation_metric": 0.0012887173797935247,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7702293549991737
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.03363144869217649,
          "train_loss": 0.03267711028456688,
          "validation_loss": 0.0031799261923879385,
          "train_metric": 0.04625355452299118,
          "validation_metric": 0.0008600332657806575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7812269639998703
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.03325092571321875,
          "train_loss": 0.03267538920044899,
          "validation_loss": 0.0030539254657924175,
          "train_metric": 0.04678041487932205,
          "validation_metric": 0.0006872487720102072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7926694649995625
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.03283405862748623,
          "train_loss": 0.033668484538793564,
          "validation_loss": 0.004107839893549681,
          "train_metric": 0.04718489199876785,
          "validation_metric": 0.0009025156614370644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8038410560002376
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.03328423673519865,
          "train_loss": 0.032682668417692184,
          "validation_loss": 0.0033551014494150877,
          "train_metric": 0.04605455324053764,
          "validation_metric": 0.0008778201881796122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8149427299995295
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.03284895428805612,
          "train_loss": 0.032859329134225845,
          "validation_loss": 0.003508152673020959,
          "train_metric": 0.0463438481092453,
          "validation_metric": 0.0007537096389569342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8260020819998317
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.03314586130727548,
          "train_loss": 0.03311585634946823,
          "validation_loss": 0.004075467586517334,
          "train_metric": 0.04515931382775307,
          "validation_metric": 0.0009943445911630988,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8367610620007326
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0331990912090987,
          "train_loss": 0.03313937410712242,
          "validation_loss": 0.003201608546078205,
          "train_metric": 0.047309085726737976,
          "validation_metric": 0.0006861303118057549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8481169900005625
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.03256918321130797,
          "train_loss": 0.03243736922740936,
          "validation_loss": 0.002851073630154133,
          "train_metric": 0.04593722149729729,
          "validation_metric": 0.0007157844374887645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8590756009998586
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.032648883236106485,
          "train_loss": 0.03297876939177513,
          "validation_loss": 0.003523380495607853,
          "train_metric": 0.046711500734090805,
          "validation_metric": 0.0007497130427509546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8699995349998062
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.03266937847365625,
          "train_loss": 0.032062698155641556,
          "validation_loss": 0.0028686553705483675,
          "train_metric": 0.04591573029756546,
          "validation_metric": 0.0006357519305311143,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8814318019994971
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.03225621220190078,
          "train_loss": 0.03212161362171173,
          "validation_loss": 0.0028134137392044067,
          "train_metric": 0.04599925875663757,
          "validation_metric": 0.0006951967952772975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8925413409997418
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0327321671647951,
          "train_loss": 0.03244645893573761,
          "validation_loss": 0.0027798800729215145,
          "train_metric": 0.04650432616472244,
          "validation_metric": 0.000683172547724098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9035465910001221
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.03215754410484806,
          "train_loss": 0.03202082961797714,
          "validation_loss": 0.0025815078988671303,
          "train_metric": 0.04593805596232414,
          "validation_metric": 0.0006228298298083246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9150997529995948
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.032575797085883096,
          "train_loss": 0.033008377999067307,
          "validation_loss": 0.0038761503528803587,
          "train_metric": 0.04576218128204346,
          "validation_metric": 0.0009398602996952832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9262279230006243
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.03272392356302589,
          "train_loss": 0.03220364451408386,
          "validation_loss": 0.0030985502526164055,
          "train_metric": 0.04563572630286217,
          "validation_metric": 0.0006824516458436847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9373555309994117
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.03303118026815355,
          "train_loss": 0.03216330334544182,
          "validation_loss": 0.0028182656969875097,
          "train_metric": 0.046569325029850006,
          "validation_metric": 0.0006270828889682889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9484712520006724
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.03218733664834872,
          "train_loss": 0.03176378831267357,
          "validation_loss": 0.0025515847373753786,
          "train_metric": 0.04580128565430641,
          "validation_metric": 0.0005849149310961366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9598988060006377
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.032138411537744105,
          "train_loss": 0.03232291340827942,
          "validation_loss": 0.002888751681894064,
          "train_metric": 0.04639798030257225,
          "validation_metric": 0.0007063561351969838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9710058519995073
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.03221017075702548,
          "train_loss": 0.031794339418411255,
          "validation_loss": 0.0025071208365261555,
          "train_metric": 0.04594752565026283,
          "validation_metric": 0.0005908288294449449,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9821616139997786
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.03207350324373692,
          "train_loss": 0.031950633972883224,
          "validation_loss": 0.002361737657338381,
          "train_metric": 0.046444110572338104,
          "validation_metric": 0.0005022534751333296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9935793010008638
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.03213402579422109,
          "train_loss": 0.032224688678979874,
          "validation_loss": 0.002900139195844531,
          "train_metric": 0.04525457322597504,
          "validation_metric": 0.0006785811274312437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0050140170005761
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.032481275498867035,
          "train_loss": 0.03174924850463867,
          "validation_loss": 0.0026115153450518847,
          "train_metric": 0.04576832428574562,
          "validation_metric": 0.0006064522895030677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0161272650002502
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.031980058760382235,
          "train_loss": 0.03297026455402374,
          "validation_loss": 0.003459882689639926,
          "train_metric": 0.04717138037085533,
          "validation_metric": 0.0007785334018990397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0270039660008479
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.03199474459688645,
          "train_loss": 0.03324267640709877,
          "validation_loss": 0.0037001732271164656,
          "train_metric": 0.04558781161904335,
          "validation_metric": 0.000991502427496016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0377718899999309
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.03197544062277302,
          "train_loss": 0.03207559511065483,
          "validation_loss": 0.0026510474272072315,
          "train_metric": 0.046182338148355484,
          "validation_metric": 0.0006288719596341252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0488021059991297
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0320418579649413,
          "train_loss": 0.03143401816487312,
          "validation_loss": 0.00216789566911757,
          "train_metric": 0.045653343200683594,
          "validation_metric": 0.0004747849015984684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0604422750002414
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.03214771729835775,
          "train_loss": 0.032525837421417236,
          "validation_loss": 0.0028664120472967625,
          "train_metric": 0.0457276850938797,
          "validation_metric": 0.0007008754182606936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0716071759998158
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.03218969303998165,
          "train_loss": 0.03227336332201958,
          "validation_loss": 0.002639578189700842,
          "train_metric": 0.046995386481285095,
          "validation_metric": 0.0005535742966458201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0826250610007264
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.032483152754139155,
          "train_loss": 0.03361009433865547,
          "validation_loss": 0.003859969787299633,
          "train_metric": 0.04603390023112297,
          "validation_metric": 0.0011041054967790842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.093645724999078
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.032495282881427556,
          "train_loss": 0.032355114817619324,
          "validation_loss": 0.0029848963022232056,
          "train_metric": 0.04677821695804596,
          "validation_metric": 0.0006498242728412151,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.104104947999076
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.03229164981166832,
          "train_loss": 0.03201928362250328,
          "validation_loss": 0.002759544877335429,
          "train_metric": 0.04553428664803505,
          "validation_metric": 0.0006332711782306433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1149204840003222
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.03176192409591749,
          "train_loss": 0.03137291595339775,
          "validation_loss": 0.0020944802090525627,
          "train_metric": 0.04577523469924927,
          "validation_metric": 0.00046736799413338304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1261699400001817
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.031495983741478994,
          "train_loss": 0.031154140830039978,
          "validation_loss": 0.0017251332756131887,
          "train_metric": 0.046046849340200424,
          "validation_metric": 0.00036898316466249526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.137437516999853
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.0318266709218733,
          "train_loss": 0.031181663274765015,
          "validation_loss": 0.0018663086229935288,
          "train_metric": 0.04584789276123047,
          "validation_metric": 0.0003929081140086055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.148380239999824
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.031613054638000904,
          "train_loss": 0.03138907998800278,
          "validation_loss": 0.001893143984489143,
          "train_metric": 0.04649810865521431,
          "validation_metric": 0.00039713102160021663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1591297689992643
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.03148384076484945,
          "train_loss": 0.03105495125055313,
          "validation_loss": 0.0017351533751934767,
          "train_metric": 0.04599401354789734,
          "validation_metric": 0.00036378903314471245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1700082959996507
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.03130340081406757,
          "train_loss": 0.031116550788283348,
          "validation_loss": 0.0020538298413157463,
          "train_metric": 0.04566515237092972,
          "validation_metric": 0.0004243200528435409,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1807740280000871
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.03128451401425991,
          "train_loss": 0.031199190765619278,
          "validation_loss": 0.0017577421385794878,
          "train_metric": 0.045619212090969086,
          "validation_metric": 0.0003885402111336589,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1911049129994353
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.03127936110831797,
          "train_loss": 0.03178548440337181,
          "validation_loss": 0.0022187517024576664,
          "train_metric": 0.046996068209409714,
          "validation_metric": 0.0004627658927347511,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2011849369991978
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.03198501974111423,
          "train_loss": 0.03189479559659958,
          "validation_loss": 0.0026905506383627653,
          "train_metric": 0.045634083449840546,
          "validation_metric": 0.0005455354694277048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2113189720002993
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.03165557194733992,
          "train_loss": 0.03162919729948044,
          "validation_loss": 0.001694255624897778,
          "train_metric": 0.046857163310050964,
          "validation_metric": 0.00035216263495385647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2222256740005832
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.03280051192268729,
          "train_loss": 0.032331813126802444,
          "validation_loss": 0.0021835332736372948,
          "train_metric": 0.046484582126140594,
          "validation_metric": 0.0005703879287466407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2325247419994412
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.0328493692795746,
          "train_loss": 0.03108593262732029,
          "validation_loss": 0.0020001018419861794,
          "train_metric": 0.04571889713406563,
          "validation_metric": 0.0004070104332640767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.24290080100036
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.031663056230172515,
          "train_loss": 0.03103019669651985,
          "validation_loss": 0.001766912522725761,
          "train_metric": 0.04622724652290344,
          "validation_metric": 0.0003622799413278699,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2531823179997446
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.03152046073228121,
          "train_loss": 0.03319389000535011,
          "validation_loss": 0.0037468967493623495,
          "train_metric": 0.04530291631817818,
          "validation_metric": 0.0009213483426719904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2635432200004288
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.031893907056655735,
          "train_loss": 0.03146890178322792,
          "validation_loss": 0.002071547321975231,
          "train_metric": 0.04655640572309494,
          "validation_metric": 0.00042551622027531266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2736539649995393
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.03147098187037045,
          "train_loss": 0.031913697719573975,
          "validation_loss": 0.0026433493476361036,
          "train_metric": 0.04622222110629082,
          "validation_metric": 0.0005628349608741701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.283662772999378
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.031724976463010535,
          "train_loss": 0.03167259693145752,
          "validation_loss": 0.0015973804984241724,
          "train_metric": 0.04674998298287392,
          "validation_metric": 0.00033928820630535483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2941363240006467
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.032010350449127145,
          "train_loss": 0.03247925266623497,
          "validation_loss": 0.003340586554259062,
          "train_metric": 0.045107800513505936,
          "validation_metric": 0.000795999076217413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3045231180003611
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.03185751607088605,
          "train_loss": 0.031143445521593094,
          "validation_loss": 0.0017844102112576365,
          "train_metric": 0.046110980212688446,
          "validation_metric": 0.0003625224926508963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3150620150008763
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.03116639766813023,
          "train_loss": 0.030948368832468987,
          "validation_loss": 0.0016320232534781098,
          "train_metric": 0.04588504135608673,
          "validation_metric": 0.0003450466028880328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3252702850004425
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.031241108852555044,
          "train_loss": 0.030890243127942085,
          "validation_loss": 0.0015656271716579795,
          "train_metric": 0.04599776864051819,
          "validation_metric": 0.00031724778818897903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3358225150004728
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.031211071625875775,
          "train_loss": 0.031262606382369995,
          "validation_loss": 0.001935317530296743,
          "train_metric": 0.04625808075070381,
          "validation_metric": 0.0003980931651312858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.346213684999384
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.03160747920628637,
          "train_loss": 0.030918339267373085,
          "validation_loss": 0.0015154011780396104,
          "train_metric": 0.04582350328564644,
          "validation_metric": 0.0003212960727978498,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3564257170000928
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.031642476678825915,
          "train_loss": 0.03185945004224777,
          "validation_loss": 0.002085421234369278,
          "train_metric": 0.04712406545877457,
          "validation_metric": 0.00043444690527394414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.434610480999254
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.03278637764742598,
          "train_loss": 0.031546466052532196,
          "validation_loss": 0.002036271383985877,
          "train_metric": 0.04530302807688713,
          "validation_metric": 0.0004403579223435372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4453363470001932
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.03149598790332675,
          "train_loss": 0.03175274282693863,
          "validation_loss": 0.0020755177829414606,
          "train_metric": 0.047059789299964905,
          "validation_metric": 0.00042957052937708795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4558834939998633
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.0319970563868992,
          "train_loss": 0.03259674832224846,
          "validation_loss": 0.003253184724599123,
          "train_metric": 0.04519197344779968,
          "validation_metric": 0.0007582263788208365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4662975859992002
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.03176709753461182,
          "train_loss": 0.03142641857266426,
          "validation_loss": 0.0021634348668158054,
          "train_metric": 0.04669339209794998,
          "validation_metric": 0.0004559505032375455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4766432069991424
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.03131931950338185,
          "train_loss": 0.031173253431916237,
          "validation_loss": 0.0017029455630108714,
          "train_metric": 0.0466630645096302,
          "validation_metric": 0.00034645767300389707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4867836360008369
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.03167983624734916,
          "train_loss": 0.0318998284637928,
          "validation_loss": 0.002516713924705982,
          "train_metric": 0.04501175135374069,
          "validation_metric": 0.0005543336737900972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4972500870007934
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.03122173564042896,
          "train_loss": 0.031007705256342888,
          "validation_loss": 0.0014857144560664892,
          "train_metric": 0.04611446335911751,
          "validation_metric": 0.00030435575172305107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.507996923999599
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0314287614019122,
          "train_loss": 0.03236522525548935,
          "validation_loss": 0.00304472167044878,
          "train_metric": 0.045270152390003204,
          "validation_metric": 0.0007187154260464013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5180498980007542
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.03235473789391108,
          "train_loss": 0.03147723525762558,
          "validation_loss": 0.002158024813979864,
          "train_metric": 0.046550504863262177,
          "validation_metric": 0.0004539359943009913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.528190936000101
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.03186429908964783,
          "train_loss": 0.03125578910112381,
          "validation_loss": 0.0017651221714913845,
          "train_metric": 0.04679836705327034,
          "validation_metric": 0.00036108054337091744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5381022639994626
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.03179669956443831,
          "train_loss": 0.03178936615586281,
          "validation_loss": 0.002533617429435253,
          "train_metric": 0.044982731342315674,
          "validation_metric": 0.0005550337955355644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5482380499997817
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.031974526995327324,
          "train_loss": 0.03097722865641117,
          "validation_loss": 0.0015235802857205272,
          "train_metric": 0.04554328694939613,
          "validation_metric": 0.00033603853080421686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5583034589999443
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.03104581224033609,
          "train_loss": 0.03092431277036667,
          "validation_loss": 0.0016595731722190976,
          "train_metric": 0.04614729806780815,
          "validation_metric": 0.00033721933141350746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.568314158999783
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.030975760833825916,
          "train_loss": 0.030899012461304665,
          "validation_loss": 0.0014834948815405369,
          "train_metric": 0.045449066907167435,
          "validation_metric": 0.0003089038946200162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5782993930006342
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.030668878396681976,
          "train_loss": 0.030947498977184296,
          "validation_loss": 0.00165486556943506,
          "train_metric": 0.04619577154517174,
          "validation_metric": 0.0003358443209435791,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5880354099990655
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.031225795799400657,
          "train_loss": 0.03131001070141792,
          "validation_loss": 0.0019398984732106328,
          "train_metric": 0.04668104648590088,
          "validation_metric": 0.00039414557977579534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.597943837999992
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.03158190147951245,
          "train_loss": 0.031526342034339905,
          "validation_loss": 0.0019327248446643353,
          "train_metric": 0.04501716047525406,
          "validation_metric": 0.0003958392480853945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6079417979999562
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.031607240438461304,
          "train_loss": 0.03239527717232704,
          "validation_loss": 0.0027867325115948915,
          "train_metric": 0.04762367904186249,
          "validation_metric": 0.0006154202856123447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.617866131999108
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.031829974555876106,
          "train_loss": 0.032550740987062454,
          "validation_loss": 0.0032167641911655664,
          "train_metric": 0.045659977942705154,
          "validation_metric": 0.0008009374141693115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6279187489999458
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.031601097842212766,
          "train_loss": 0.03162560984492302,
          "validation_loss": 0.0023115770891308784,
          "train_metric": 0.04666046053171158,
          "validation_metric": 0.000500815047416836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6377405929997622
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.03220389847410843,
          "train_loss": 0.03249916806817055,
          "validation_loss": 0.0030305718537420034,
          "train_metric": 0.04578380659222603,
          "validation_metric": 0.0007557066855952144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.647831940999822
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.03168538606041693,
          "train_loss": 0.03141395375132561,
          "validation_loss": 0.0017751469276845455,
          "train_metric": 0.046745188534259796,
          "validation_metric": 0.00035791605478152633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6578491469990695
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.031898793240543455,
          "train_loss": 0.03157275170087814,
          "validation_loss": 0.002300266409292817,
          "train_metric": 0.04489709436893463,
          "validation_metric": 0.0004832737613469362,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6676414140001725
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.03097268642159179,
          "train_loss": 0.030855044722557068,
          "validation_loss": 0.0013846615329384804,
          "train_metric": 0.04643123224377632,
          "validation_metric": 0.00027922698063775897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.678292333999707
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.03146003623260185,
          "train_loss": 0.030932417139410973,
          "validation_loss": 0.0011890721507370472,
          "train_metric": 0.04645967483520508,
          "validation_metric": 0.00024476993712596595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6889948410007491
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.031706200607004575,
          "train_loss": 0.03111664205789566,
          "validation_loss": 0.0017035421915352345,
          "train_metric": 0.045153964310884476,
          "validation_metric": 0.000371233734767884,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6992091009997239
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.03175903856754303,
          "train_loss": 0.031097134575247765,
          "validation_loss": 0.0017128719482570887,
          "train_metric": 0.046772561967372894,
          "validation_metric": 0.00035045668482780457,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7092040169991378
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.032721840718295425,
          "train_loss": 0.031701698899269104,
          "validation_loss": 0.002298751613125205,
          "train_metric": 0.04508689045906067,
          "validation_metric": 0.00046339043183252215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7189230780004436
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.03177489718655124,
          "train_loss": 0.030973827466368675,
          "validation_loss": 0.0014034849591553211,
          "train_metric": 0.04669712856411934,
          "validation_metric": 0.00028329662745818496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7289396180003678
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.031431551920832135,
          "train_loss": 0.031127218157052994,
          "validation_loss": 0.001393609563820064,
          "train_metric": 0.04571262374520302,
          "validation_metric": 0.0003025399928446859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.738847235999856
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.03137415071250871,
          "train_loss": 0.030749276280403137,
          "validation_loss": 0.0013557663187384605,
          "train_metric": 0.04540911689400673,
          "validation_metric": 0.00028253349591977894,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7488393029998406
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.030885572487022728,
          "train_loss": 0.030664680525660515,
          "validation_loss": 0.0011493946658447385,
          "train_metric": 0.04634057730436325,
          "validation_metric": 0.00023167523613665253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7593622580006922
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.030795877843047492,
          "train_loss": 0.03137531876564026,
          "validation_loss": 0.0019955073948949575,
          "train_metric": 0.046370234340429306,
          "validation_metric": 0.00042205030331388116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7692993820000993
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.030926437757443637,
          "train_loss": 0.03062150627374649,
          "validation_loss": 0.0012468998320400715,
          "train_metric": 0.04577082023024559,
          "validation_metric": 0.00025442519108764827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7794771360004233
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.030682213848194806,
          "train_loss": 0.030497735366225243,
          "validation_loss": 0.0011008080327883363,
          "train_metric": 0.045608650892972946,
          "validation_metric": 0.0002232875267509371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7897953610008699
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0311648698116187,
          "train_loss": 0.03357233852148056,
          "validation_loss": 0.004464158788323402,
          "train_metric": 0.04575459659099579,
          "validation_metric": 0.0011600054567679763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7997910419999243
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.032512931153178215,
          "train_loss": 0.03089950978755951,
          "validation_loss": 0.0011883648112416267,
          "train_metric": 0.04654490575194359,
          "validation_metric": 0.00024094538821373135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8099929300005897
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.031109248055145144,
          "train_loss": 0.03113509714603424,
          "validation_loss": 0.0016956484178081155,
          "train_metric": 0.04603574797511101,
          "validation_metric": 0.00034217475331388414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8198694639995665
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.030843256401567487,
          "train_loss": 0.030532656237483025,
          "validation_loss": 0.0011793964076787233,
          "train_metric": 0.04573262855410576,
          "validation_metric": 0.00024377221416216344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.829775337000683
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.030964826932176948,
          "train_loss": 0.03322489932179451,
          "validation_loss": 0.004214589949697256,
          "train_metric": 0.04530920833349228,
          "validation_metric": 0.0010181553661823273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8398957459994563
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.03284747194265947,
          "train_loss": 0.0317360945045948,
          "validation_loss": 0.002089897869154811,
          "train_metric": 0.046963829547166824,
          "validation_metric": 0.0004227591853123158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8496412349995808
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0313202696852386,
          "train_loss": 0.03102453425526619,
          "validation_loss": 0.0016438636230304837,
          "train_metric": 0.04546113684773445,
          "validation_metric": 0.00033163317129947245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8596262909995858
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.03097709803842008,
          "train_loss": 0.03112965263426304,
          "validation_loss": 0.001635542488656938,
          "train_metric": 0.04555869847536087,
          "validation_metric": 0.0003415630490053445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8693382250003197
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.030894357340002898,
          "train_loss": 0.031544577330350876,
          "validation_loss": 0.0022434366401284933,
          "train_metric": 0.04589325189590454,
          "validation_metric": 0.0004982675309292972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8793652330004988
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.030980709590949118,
          "train_loss": 0.030605336651206017,
          "validation_loss": 0.001302491407841444,
          "train_metric": 0.04571329057216644,
          "validation_metric": 0.00026721550966612995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8892274789996009
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.031526375154498965,
          "train_loss": 0.03110930696129799,
          "validation_loss": 0.001874716253951192,
          "train_metric": 0.04675282910466194,
          "validation_metric": 0.0003778696700464934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8991228630002297
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.030950131767895073,
          "train_loss": 0.031084977090358734,
          "validation_loss": 0.001732941367663443,
          "train_metric": 0.04566137120127678,
          "validation_metric": 0.00035906260018236935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9091306110003643
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.03089696360984817,
          "train_loss": 0.030576322227716446,
          "validation_loss": 0.0012785530416294932,
          "train_metric": 0.04557299241423607,
          "validation_metric": 0.00026851476286537945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9189531609990809
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.03061990403512027,
          "train_loss": 0.030536958947777748,
          "validation_loss": 0.0010864271316677332,
          "train_metric": 0.04621027782559395,
          "validation_metric": 0.00021887564798817039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.929505706000782
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.03099640090658795,
          "train_loss": 0.03205288574099541,
          "validation_loss": 0.00265776040032506,
          "train_metric": 0.0450524240732193,
          "validation_metric": 0.0006113431882113218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9396684959992854
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.031485513667576015,
          "train_loss": 0.030839113518595695,
          "validation_loss": 0.0014912495389580727,
          "train_metric": 0.04658661037683487,
          "validation_metric": 0.0003043048200197518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9496529059997556
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.030923937971238047,
          "train_loss": 0.03060811012983322,
          "validation_loss": 0.0008955082739703357,
          "train_metric": 0.04614536464214325,
          "validation_metric": 0.00018411809287499636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.960499542999969
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.030748339428100735,
          "train_loss": 0.030359920114278793,
          "validation_loss": 0.000931872520595789,
          "train_metric": 0.04583311080932617,
          "validation_metric": 0.00018771036411635578,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9707274270003836
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.03059462708188221,
          "train_loss": 0.030519690364599228,
          "validation_loss": 0.0011473132763057947,
          "train_metric": 0.045624230057001114,
          "validation_metric": 0.00023110638721846044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9810504199995194
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.030773170001339167,
          "train_loss": 0.030619941651821136,
          "validation_loss": 0.0010150152957066894,
          "train_metric": 0.04553576558828354,
          "validation_metric": 0.00021209596889093518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.991251759000079
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.03071868629194796,
          "train_loss": 0.03178957477211952,
          "validation_loss": 0.0025112221483141184,
          "train_metric": 0.04654295742511749,
          "validation_metric": 0.0005564586608670652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0012096940008632
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.031372925150208175,
          "train_loss": 0.031153274700045586,
          "validation_loss": 0.0015538083389401436,
          "train_metric": 0.04549422860145569,
          "validation_metric": 0.0003469907969702035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0111939910002548
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.03104865673230961,
          "train_loss": 0.03106491081416607,
          "validation_loss": 0.001696386025287211,
          "train_metric": 0.045254092663526535,
          "validation_metric": 0.00035818657488562167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.021239403000436
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.03264034882886335,
          "train_loss": 0.032148249447345734,
          "validation_loss": 0.001911256811581552,
          "train_metric": 0.04656457155942917,
          "validation_metric": 0.00048313100705854595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0317698220005695
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.03167098965786863,
          "train_loss": 0.030910398811101913,
          "validation_loss": 0.0017314699944108725,
          "train_metric": 0.045087531208992004,
          "validation_metric": 0.00034881712053902447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0418311009998433
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.030727704215678386,
          "train_loss": 0.030926303938031197,
          "validation_loss": 0.001693017897196114,
          "train_metric": 0.04649187624454498,
          "validation_metric": 0.00034353669616393745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.051726335999774
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.03052046784432605,
          "train_loss": 0.03037029318511486,
          "validation_loss": 0.0009756509680300951,
          "train_metric": 0.04604138433933258,
          "validation_metric": 0.00019652796618174762,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0617598949993408
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.030534652760252357,
          "train_loss": 0.030531585216522217,
          "validation_loss": 0.0010616894578561187,
          "train_metric": 0.04550008103251457,
          "validation_metric": 0.0002235229912912473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0718562429992744
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.030707989368238486,
          "train_loss": 0.030529240146279335,
          "validation_loss": 0.0012149368412792683,
          "train_metric": 0.04631833732128143,
          "validation_metric": 0.00024473198573105037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.081736775999161
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.03059153605136089,
          "train_loss": 0.03062969446182251,
          "validation_loss": 0.001377733307890594,
          "train_metric": 0.04595872014760971,
          "validation_metric": 0.00027973760734312236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0916239239995775
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.030834097749902867,
          "train_loss": 0.03144065663218498,
          "validation_loss": 0.0015675274189561605,
          "train_metric": 0.046931568533182144,
          "validation_metric": 0.0003208894922863692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1013701779993426
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.03094428659096593,
          "train_loss": 0.03206798806786537,
          "validation_loss": 0.002658720128238201,
          "train_metric": 0.045584600418806076,
          "validation_metric": 0.0006328733870759606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1114026230006857
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.030732276616618037,
          "train_loss": 0.03066093847155571,
          "validation_loss": 0.001243649167008698,
          "train_metric": 0.045868292450904846,
          "validation_metric": 0.0002573156089056283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1212906269993255
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.030667772516608238,
          "train_loss": 0.030540384352207184,
          "validation_loss": 0.0011359433410689235,
          "train_metric": 0.04554689675569534,
          "validation_metric": 0.0002288674149895087,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.131325845999527
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.030609020323026925,
          "train_loss": 0.03035593219101429,
          "validation_loss": 0.0009501958265900612,
          "train_metric": 0.04610532149672508,
          "validation_metric": 0.0001914004678837955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1413145970000187
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.030464545594441006,
          "train_loss": 0.030529441311955452,
          "validation_loss": 0.0011285392101854086,
          "train_metric": 0.045649196952581406,
          "validation_metric": 0.00022732466459274292,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1511077449995355
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.030594925105106086,
          "train_loss": 0.031162502244114876,
          "validation_loss": 0.0017545822774991393,
          "train_metric": 0.04585970565676689,
          "validation_metric": 0.00039463519351556897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.161095335999562
        }
      ],
      "best": {
        "epoch": 162,
        "updates": 2592
      },
      "best_validation_metric": 0.00018411809287499636,
      "config": {
        "label": "Huber + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "huber",
        "outliers": true,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.013569012229156416
    },
    "92630a9c55d2_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 1,
          "samples_seen": 1024,
          "online_loss": 0.9560257196426392,
          "train_loss": 0.9161807298660278,
          "validation_loss": 0.6619400978088379,
          "train_metric": 0.09227436780929565,
          "validation_metric": 0.06666817516088486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.00904585600073915
        },
        {
          "epoch": 2,
          "updates": 2,
          "samples_seen": 2048,
          "online_loss": 0.9161807298660278,
          "train_loss": 0.8742673993110657,
          "validation_loss": 0.630653440952301,
          "train_metric": 0.08805300295352936,
          "validation_metric": 0.06351710855960846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.019314114999360754
        },
        {
          "epoch": 3,
          "updates": 3,
          "samples_seen": 3072,
          "online_loss": 0.8742673397064209,
          "train_loss": 0.8294147253036499,
          "validation_loss": 0.5974236726760864,
          "train_metric": 0.08353561162948608,
          "validation_metric": 0.06017033010721207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02861860600023647
        },
        {
          "epoch": 4,
          "updates": 4,
          "samples_seen": 4096,
          "online_loss": 0.8294147849082947,
          "train_loss": 0.7809800505638123,
          "validation_loss": 0.5616785883903503,
          "train_metric": 0.07865744829177856,
          "validation_metric": 0.05657021328806877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.037577905000944156
        },
        {
          "epoch": 5,
          "updates": 5,
          "samples_seen": 5120,
          "online_loss": 0.7809799909591675,
          "train_loss": 0.7286146879196167,
          "validation_loss": 0.5231126546859741,
          "train_metric": 0.0733834058046341,
          "validation_metric": 0.05268598347902298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.04615338800067548
        },
        {
          "epoch": 6,
          "updates": 6,
          "samples_seen": 6144,
          "online_loss": 0.7286146879196167,
          "train_loss": 0.6722691059112549,
          "validation_loss": 0.48164600133895874,
          "train_metric": 0.06770848482847214,
          "validation_metric": 0.048509616404771805,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05446923300041817
        },
        {
          "epoch": 7,
          "updates": 7,
          "samples_seen": 7168,
          "online_loss": 0.6722691059112549,
          "train_loss": 0.6122482419013977,
          "validation_loss": 0.4374586343765259,
          "train_metric": 0.06166340038180351,
          "validation_metric": 0.044059231877326965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06309163100013393
        },
        {
          "epoch": 8,
          "updates": 8,
          "samples_seen": 8192,
          "online_loss": 0.6122481822967529,
          "train_loss": 0.5493510365486145,
          "validation_loss": 0.39109256863594055,
          "train_metric": 0.05532862991094589,
          "validation_metric": 0.039389412850141525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0713806179992389
        },
        {
          "epoch": 9,
          "updates": 9,
          "samples_seen": 9216,
          "online_loss": 0.5493510365486145,
          "train_loss": 0.4850430190563202,
          "validation_loss": 0.343569815158844,
          "train_metric": 0.048851754516363144,
          "validation_metric": 0.034603092819452286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0797902600006637
        },
        {
          "epoch": 10,
          "updates": 10,
          "samples_seen": 10240,
          "online_loss": 0.4850430488586426,
          "train_loss": 0.4216090738773346,
          "validation_loss": 0.2965115010738373,
          "train_metric": 0.042462922632694244,
          "validation_metric": 0.029863551259040833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08803174300010141
        },
        {
          "epoch": 11,
          "updates": 11,
          "samples_seen": 11264,
          "online_loss": 0.4216090738773346,
          "train_loss": 0.36223751306533813,
          "validation_loss": 0.2522137463092804,
          "train_metric": 0.03648323193192482,
          "validation_metric": 0.025402043014764786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09658663899972453
        },
        {
          "epoch": 12,
          "updates": 12,
          "samples_seen": 12288,
          "online_loss": 0.36223748326301575,
          "train_loss": 0.3108953833580017,
          "validation_loss": 0.2135777324438095,
          "train_metric": 0.031312245875597,
          "validation_metric": 0.021510768681764603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1049825440004497
        },
        {
          "epoch": 13,
          "updates": 13,
          "samples_seen": 13312,
          "online_loss": 0.3108953833580017,
          "train_loss": 0.2716892957687378,
          "validation_loss": 0.18369662761688232,
          "train_metric": 0.027363549917936325,
          "validation_metric": 0.01850125379860401,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11363218199949188
        },
        {
          "epoch": 14,
          "updates": 14,
          "samples_seen": 14336,
          "online_loss": 0.2716892659664154,
          "train_loss": 0.24716106057167053,
          "validation_loss": 0.1647392064332962,
          "train_metric": 0.024893153458833694,
          "validation_metric": 0.016591928899288177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1226449629994022
        },
        {
          "epoch": 15,
          "updates": 15,
          "samples_seen": 15360,
          "online_loss": 0.24716104567050934,
          "train_loss": 0.23577168583869934,
          "validation_loss": 0.15625663101673126,
          "train_metric": 0.023746062070131302,
          "validation_metric": 0.015737595036625862,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1315583310006332
        },
        {
          "epoch": 16,
          "updates": 16,
          "samples_seen": 16384,
          "online_loss": 0.23577168583869934,
          "train_loss": 0.23156219720840454,
          "validation_loss": 0.1548166573047638,
          "train_metric": 0.023322096094489098,
          "validation_metric": 0.015592565760016441,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24416858100084937
        },
        {
          "epoch": 17,
          "updates": 17,
          "samples_seen": 17408,
          "online_loss": 0.23156219720840454,
          "train_loss": 0.2286653220653534,
          "validation_loss": 0.1568123698234558,
          "train_metric": 0.02303033322095871,
          "validation_metric": 0.015793567523360252,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2520129269996687
        },
        {
          "epoch": 18,
          "updates": 18,
          "samples_seen": 18432,
          "online_loss": 0.2286653220653534,
          "train_loss": 0.22529615461826324,
          "validation_loss": 0.1610189974308014,
          "train_metric": 0.02269100211560726,
          "validation_metric": 0.016217242926359177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25941748100012774
        },
        {
          "epoch": 19,
          "updates": 19,
          "samples_seen": 19456,
          "online_loss": 0.22529613971710205,
          "train_loss": 0.22261163592338562,
          "validation_loss": 0.16769887506961823,
          "train_metric": 0.022420627996325493,
          "validation_metric": 0.016890017315745354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2671747990007134
        },
        {
          "epoch": 20,
          "updates": 20,
          "samples_seen": 20480,
          "online_loss": 0.22261162102222443,
          "train_loss": 0.22140321135520935,
          "validation_loss": 0.1761833280324936,
          "train_metric": 0.022298922762274742,
          "validation_metric": 0.01774454116821289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.27545624599952134
        },
        {
          "epoch": 21,
          "updates": 21,
          "samples_seen": 21504,
          "online_loss": 0.22140321135520935,
          "train_loss": 0.22034846246242523,
          "validation_loss": 0.18385130167007446,
          "train_metric": 0.022192690521478653,
          "validation_metric": 0.01851683109998703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28320126899961906
        },
        {
          "epoch": 22,
          "updates": 22,
          "samples_seen": 22528,
          "online_loss": 0.22034846246242523,
          "train_loss": 0.21711480617523193,
          "validation_loss": 0.1876053810119629,
          "train_metric": 0.02186700887978077,
          "validation_metric": 0.018894927576184273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29090969000026234
        },
        {
          "epoch": 23,
          "updates": 23,
          "samples_seen": 23552,
          "online_loss": 0.21711480617523193,
          "train_loss": 0.2103533148765564,
          "validation_loss": 0.18585535883903503,
          "train_metric": 0.021186014637351036,
          "validation_metric": 0.018718672916293144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2987903310004185
        },
        {
          "epoch": 24,
          "updates": 24,
          "samples_seen": 24576,
          "online_loss": 0.2103532999753952,
          "train_loss": 0.20036010444164276,
          "validation_loss": 0.17898492515087128,
          "train_metric": 0.020179536193609238,
          "validation_metric": 0.01802670769393444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.30671628399977635
        },
        {
          "epoch": 25,
          "updates": 25,
          "samples_seen": 25600,
          "online_loss": 0.20036010444164276,
          "train_loss": 0.18852029740810394,
          "validation_loss": 0.16862811148166656,
          "train_metric": 0.018987072631716728,
          "validation_metric": 0.016983604058623314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.31491426100001263
        },
        {
          "epoch": 26,
          "updates": 26,
          "samples_seen": 26624,
          "online_loss": 0.18852029740810394,
          "train_loss": 0.17649151384830475,
          "validation_loss": 0.15676411986351013,
          "train_metric": 0.017775580286979675,
          "validation_metric": 0.01578870788216591,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3233447359998536
        },
        {
          "epoch": 27,
          "updates": 27,
          "samples_seen": 27648,
          "online_loss": 0.17649152874946594,
          "train_loss": 0.16562947630882263,
          "validation_loss": 0.14510296285152435,
          "train_metric": 0.016681594774127007,
          "validation_metric": 0.014614238403737545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3317649939999683
        },
        {
          "epoch": 28,
          "updates": 28,
          "samples_seen": 28672,
          "online_loss": 0.16562947630882263,
          "train_loss": 0.15673771500587463,
          "validation_loss": 0.13480976223945618,
          "train_metric": 0.015786049887537956,
          "validation_metric": 0.013577544130384922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3402750479999668
        },
        {
          "epoch": 29,
          "updates": 29,
          "samples_seen": 29696,
          "online_loss": 0.15673771500587463,
          "train_loss": 0.15006347000598907,
          "validation_loss": 0.12647628784179688,
          "train_metric": 0.01511384453624487,
          "validation_metric": 0.012738226912915707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34882269199988514
        },
        {
          "epoch": 30,
          "updates": 30,
          "samples_seen": 30720,
          "online_loss": 0.15006348490715027,
          "train_loss": 0.14543038606643677,
          "validation_loss": 0.12022890895605087,
          "train_metric": 0.014647215604782104,
          "validation_metric": 0.012109015136957169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3578098100006173
        },
        {
          "epoch": 31,
          "updates": 31,
          "samples_seen": 31744,
          "online_loss": 0.14543038606643677,
          "train_loss": 0.14241059124469757,
          "validation_loss": 0.1158810555934906,
          "train_metric": 0.014343073591589928,
          "validation_metric": 0.011671116575598717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36694112499935727
        },
        {
          "epoch": 32,
          "updates": 32,
          "samples_seen": 32768,
          "online_loss": 0.14241057634353638,
          "train_loss": 0.14047309756278992,
          "validation_loss": 0.11307072639465332,
          "train_metric": 0.014147937297821045,
          "validation_metric": 0.011388069950044155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37604803899921535
        },
        {
          "epoch": 33,
          "updates": 33,
          "samples_seen": 33792,
          "online_loss": 0.1404731124639511,
          "train_loss": 0.13908520340919495,
          "validation_loss": 0.11136018484830856,
          "train_metric": 0.014008152298629284,
          "validation_metric": 0.01121578924357891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3849584700001287
        },
        {
          "epoch": 34,
          "updates": 34,
          "samples_seen": 34816,
          "online_loss": 0.13908520340919495,
          "train_loss": 0.13777372241020203,
          "validation_loss": 0.1103023961186409,
          "train_metric": 0.013876065611839294,
          "validation_metric": 0.011109251528978348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3970080159997451
        },
        {
          "epoch": 35,
          "updates": 35,
          "samples_seen": 35840,
          "online_loss": 0.13777372241020203,
          "train_loss": 0.13616566359996796,
          "validation_loss": 0.10949002951383591,
          "train_metric": 0.013714107684791088,
          "validation_metric": 0.011027433909475803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4053995369995391
        },
        {
          "epoch": 36,
          "updates": 36,
          "samples_seen": 36864,
          "online_loss": 0.13616566359996796,
          "train_loss": 0.13402269780635834,
          "validation_loss": 0.10860000550746918,
          "train_metric": 0.013498274609446526,
          "validation_metric": 0.010937795042991638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4143038889997115
        },
        {
          "epoch": 37,
          "updates": 37,
          "samples_seen": 37888,
          "online_loss": 0.13402268290519714,
          "train_loss": 0.1312716156244278,
          "validation_loss": 0.10743442177772522,
          "train_metric": 0.013221196830272675,
          "validation_metric": 0.010820401832461357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42312879400014936
        },
        {
          "epoch": 38,
          "updates": 38,
          "samples_seen": 38912,
          "online_loss": 0.1312716156244278,
          "train_loss": 0.12801945209503174,
          "validation_loss": 0.10594778507947922,
          "train_metric": 0.012893651612102985,
          "validation_metric": 0.010670673102140427,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.43184457600000314
        },
        {
          "epoch": 39,
          "updates": 39,
          "samples_seen": 39936,
          "online_loss": 0.12801945209503174,
          "train_loss": 0.12453638017177582,
          "validation_loss": 0.10424444079399109,
          "train_metric": 0.012542849406599998,
          "validation_metric": 0.010499119758605957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44557100599922705
        },
        {
          "epoch": 40,
          "updates": 40,
          "samples_seen": 40960,
          "online_loss": 0.12453639507293701,
          "train_loss": 0.12119479477405548,
          "validation_loss": 0.10253626108169556,
          "train_metric": 0.01220629457384348,
          "validation_metric": 0.010327075608074665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45397396800035494
        },
        {
          "epoch": 41,
          "updates": 41,
          "samples_seen": 41984,
          "online_loss": 0.12119479477405548,
          "train_loss": 0.11836740374565125,
          "validation_loss": 0.10106290131807327,
          "train_metric": 0.011921532452106476,
          "validation_metric": 0.010178686119616032,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4682434000005742
        },
        {
          "epoch": 42,
          "updates": 42,
          "samples_seen": 43008,
          "online_loss": 0.11836740374565125,
          "train_loss": 0.11630696058273315,
          "validation_loss": 0.09999421238899231,
          "train_metric": 0.011714011430740356,
          "validation_metric": 0.010071051307022572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47701811399929284
        },
        {
          "epoch": 43,
          "updates": 43,
          "samples_seen": 44032,
          "online_loss": 0.11630696803331375,
          "train_loss": 0.11504848301410675,
          "validation_loss": 0.09934888780117035,
          "train_metric": 0.011587261222302914,
          "validation_metric": 0.010006056167185307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4856345539992617
        },
        {
          "epoch": 44,
          "updates": 44,
          "samples_seen": 45056,
          "online_loss": 0.11504847556352615,
          "train_loss": 0.11438239365816116,
          "validation_loss": 0.09896872192621231,
          "train_metric": 0.011520175263285637,
          "validation_metric": 0.009967767633497715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.494432453000627
        },
        {
          "epoch": 45,
          "updates": 45,
          "samples_seen": 46080,
          "online_loss": 0.11438238620758057,
          "train_loss": 0.11392681300640106,
          "validation_loss": 0.09857125580310822,
          "train_metric": 0.011474290862679482,
          "validation_metric": 0.009927735663950443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5034258329997101
        },
        {
          "epoch": 46,
          "updates": 46,
          "samples_seen": 47104,
          "online_loss": 0.11392681300640106,
          "train_loss": 0.11327804625034332,
          "validation_loss": 0.09786372631788254,
          "train_metric": 0.011408950202167034,
          "validation_metric": 0.009856476448476315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5125321890009218
        },
        {
          "epoch": 47,
          "updates": 47,
          "samples_seen": 48128,
          "online_loss": 0.11327804625034332,
          "train_loss": 0.11216676980257034,
          "validation_loss": 0.09666233509778976,
          "train_metric": 0.011297025717794895,
          "validation_metric": 0.009735476225614548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5215467559992248
        },
        {
          "epoch": 48,
          "updates": 48,
          "samples_seen": 49152,
          "online_loss": 0.11216676980257034,
          "train_loss": 0.11054156720638275,
          "validation_loss": 0.09495529532432556,
          "train_metric": 0.011133342050015926,
          "validation_metric": 0.009563549421727657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5303051060000143
        },
        {
          "epoch": 49,
          "updates": 49,
          "samples_seen": 50176,
          "online_loss": 0.11054157465696335,
          "train_loss": 0.10854777693748474,
          "validation_loss": 0.09288579970598221,
          "train_metric": 0.01093253493309021,
          "validation_metric": 0.009355118498206139,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5391178270001546
        },
        {
          "epoch": 50,
          "updates": 50,
          "samples_seen": 51200,
          "online_loss": 0.10854779183864594,
          "train_loss": 0.10643056035041809,
          "validation_loss": 0.0906766727566719,
          "train_metric": 0.010719296522438526,
          "validation_metric": 0.009132622741162777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.547937882000042
        },
        {
          "epoch": 51,
          "updates": 51,
          "samples_seen": 52224,
          "online_loss": 0.1064305454492569,
          "train_loss": 0.10442199558019638,
          "validation_loss": 0.0885419249534607,
          "train_metric": 0.010516999289393425,
          "validation_metric": 0.008917619474232197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5565167389995622
        },
        {
          "epoch": 52,
          "updates": 52,
          "samples_seen": 53248,
          "online_loss": 0.10442198812961578,
          "train_loss": 0.10266201198101044,
          "validation_loss": 0.08662351965904236,
          "train_metric": 0.010339743457734585,
          "validation_metric": 0.008724404498934746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5649232689993369
        },
        {
          "epoch": 53,
          "updates": 53,
          "samples_seen": 54272,
          "online_loss": 0.10266203433275223,
          "train_loss": 0.1011735200881958,
          "validation_loss": 0.08496905863285065,
          "train_metric": 0.010189826600253582,
          "validation_metric": 0.008557774126529694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5733820759996888
        },
        {
          "epoch": 54,
          "updates": 54,
          "samples_seen": 55296,
          "online_loss": 0.1011735275387764,
          "train_loss": 0.09988442063331604,
          "validation_loss": 0.08354541659355164,
          "train_metric": 0.010059994645416737,
          "validation_metric": 0.008414389565587044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5821847440001875
        },
        {
          "epoch": 55,
          "updates": 55,
          "samples_seen": 56320,
          "online_loss": 0.09988442808389664,
          "train_loss": 0.09867511689662933,
          "validation_loss": 0.0822724774479866,
          "train_metric": 0.009938196279108524,
          "validation_metric": 0.008286181837320328,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.743202358000417
        },
        {
          "epoch": 56,
          "updates": 56,
          "samples_seen": 57344,
          "online_loss": 0.09867511689662933,
          "train_loss": 0.09742829203605652,
          "validation_loss": 0.08106020092964172,
          "train_metric": 0.009812620468437672,
          "validation_metric": 0.008164087310433388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7624121509998076
        },
        {
          "epoch": 57,
          "updates": 57,
          "samples_seen": 58368,
          "online_loss": 0.09742830693721771,
          "train_loss": 0.09606557339429855,
          "validation_loss": 0.07983757555484772,
          "train_metric": 0.009675372391939163,
          "validation_metric": 0.008040948770940304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.781233773999702
        },
        {
          "epoch": 58,
          "updates": 58,
          "samples_seen": 59392,
          "online_loss": 0.09606557339429855,
          "train_loss": 0.09456409513950348,
          "validation_loss": 0.07856772094964981,
          "train_metric": 0.00952414982020855,
          "validation_metric": 0.007913053967058659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8035625550000987
        },
        {
          "epoch": 59,
          "updates": 59,
          "samples_seen": 60416,
          "online_loss": 0.09456409513950348,
          "train_loss": 0.09295393526554108,
          "validation_loss": 0.07724975794553757,
          "train_metric": 0.009361980482935905,
          "validation_metric": 0.007780313026160002,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8221260720001737
        },
        {
          "epoch": 60,
          "updates": 60,
          "samples_seen": 61440,
          "online_loss": 0.09295394271612167,
          "train_loss": 0.09130071103572845,
          "validation_loss": 0.075909823179245,
          "train_metric": 0.009195473045110703,
          "validation_metric": 0.007645359728485346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8422323119993962
        },
        {
          "epoch": 61,
          "updates": 61,
          "samples_seen": 62464,
          "online_loss": 0.09130071103572845,
          "train_loss": 0.08968067914247513,
          "validation_loss": 0.07458603382110596,
          "train_metric": 0.009032309986650944,
          "validation_metric": 0.007512031123042107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8593628840008023
        },
        {
          "epoch": 62,
          "updates": 62,
          "samples_seen": 63488,
          "online_loss": 0.08968067914247513,
          "train_loss": 0.08815548568964005,
          "validation_loss": 0.07331190258264542,
          "train_metric": 0.008878697641193867,
          "validation_metric": 0.007383706048130989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8807743550005398
        },
        {
          "epoch": 63,
          "updates": 63,
          "samples_seen": 64512,
          "online_loss": 0.08815548568964005,
          "train_loss": 0.08675387501716614,
          "validation_loss": 0.07210339605808258,
          "train_metric": 0.008737532421946526,
          "validation_metric": 0.007261989638209343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8994283160009218
        },
        {
          "epoch": 64,
          "updates": 64,
          "samples_seen": 65536,
          "online_loss": 0.08675387501716614,
          "train_loss": 0.08546563237905502,
          "validation_loss": 0.07095348834991455,
          "train_metric": 0.008607786148786545,
          "validation_metric": 0.007146175019443035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9193547720005881
        },
        {
          "epoch": 65,
          "updates": 65,
          "samples_seen": 66560,
          "online_loss": 0.08546563237905502,
          "train_loss": 0.08424965292215347,
          "validation_loss": 0.06983564049005508,
          "train_metric": 0.00848531723022461,
          "validation_metric": 0.007033590227365494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.938005100000737
        },
        {
          "epoch": 66,
          "updates": 66,
          "samples_seen": 67584,
          "online_loss": 0.08424966037273407,
          "train_loss": 0.08305266499519348,
          "validation_loss": 0.06871530413627625,
          "train_metric": 0.008364761248230934,
          "validation_metric": 0.006920753978192806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9583618439992279
        },
        {
          "epoch": 67,
          "updates": 67,
          "samples_seen": 68608,
          "online_loss": 0.08305266499519348,
          "train_loss": 0.08183129876852036,
          "validation_loss": 0.06756411492824554,
          "train_metric": 0.008241748437285423,
          "validation_metric": 0.006804809905588627,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9760288079996826
        },
        {
          "epoch": 68,
          "updates": 68,
          "samples_seen": 69632,
          "online_loss": 0.08183129876852036,
          "train_loss": 0.08056823909282684,
          "validation_loss": 0.06637135148048401,
          "train_metric": 0.00811453815549612,
          "validation_metric": 0.006684679538011551,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9976929919994291
        },
        {
          "epoch": 69,
          "updates": 69,
          "samples_seen": 70656,
          "online_loss": 0.08056823909282684,
          "train_loss": 0.07927633821964264,
          "validation_loss": 0.06514780223369598,
          "train_metric": 0.007984422147274017,
          "validation_metric": 0.006561446934938431,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0162848939999094
        },
        {
          "epoch": 70,
          "updates": 70,
          "samples_seen": 71680,
          "online_loss": 0.07927633821964264,
          "train_loss": 0.0779900997877121,
          "validation_loss": 0.06392112374305725,
          "train_metric": 0.007854877039790154,
          "validation_metric": 0.0064379009418189526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0347925999994914
        },
        {
          "epoch": 71,
          "updates": 71,
          "samples_seen": 72704,
          "online_loss": 0.0779900923371315,
          "train_loss": 0.07674983143806458,
          "validation_loss": 0.06272576749324799,
          "train_metric": 0.0077299620024859905,
          "validation_metric": 0.00631750887259841,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0534503840008256
        },
        {
          "epoch": 72,
          "updates": 72,
          "samples_seen": 73728,
          "online_loss": 0.07674982398748398,
          "train_loss": 0.07558578252792358,
          "validation_loss": 0.0615919828414917,
          "train_metric": 0.007612723391503096,
          "validation_metric": 0.0062033189460635185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0718948070007173
        },
        {
          "epoch": 73,
          "updates": 73,
          "samples_seen": 74752,
          "online_loss": 0.07558578997850418,
          "train_loss": 0.07450868934392929,
          "validation_loss": 0.06053847447037697,
          "train_metric": 0.007504242472350597,
          "validation_metric": 0.006097213365137577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.094096463999449
        },
        {
          "epoch": 74,
          "updates": 74,
          "samples_seen": 75776,
          "online_loss": 0.0745086818933487,
          "train_loss": 0.07350917160511017,
          "validation_loss": 0.059570372104644775,
          "train_metric": 0.007403573486953974,
          "validation_metric": 0.0059997099451720715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1133773649999057
        },
        {
          "epoch": 75,
          "updates": 75,
          "samples_seen": 76800,
          "online_loss": 0.07350917160511017,
          "train_loss": 0.07256462424993515,
          "validation_loss": 0.058682337403297424,
          "train_metric": 0.007308443076908588,
          "validation_metric": 0.005910269916057587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1316627689993766
        },
        {
          "epoch": 76,
          "updates": 76,
          "samples_seen": 77824,
          "online_loss": 0.07256462424993515,
          "train_loss": 0.07164978981018066,
          "validation_loss": 0.05786418169736862,
          "train_metric": 0.007216304540634155,
          "validation_metric": 0.005827868822962046,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1405150160007906
        },
        {
          "epoch": 77,
          "updates": 77,
          "samples_seen": 78848,
          "online_loss": 0.07164978235960007,
          "train_loss": 0.07074650377035141,
          "validation_loss": 0.0571066215634346,
          "train_metric": 0.007125329691916704,
          "validation_metric": 0.005751570221036673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1488453460005985
        },
        {
          "epoch": 78,
          "updates": 78,
          "samples_seen": 79872,
          "online_loss": 0.070746511220932,
          "train_loss": 0.06984960287809372,
          "validation_loss": 0.05640493705868721,
          "train_metric": 0.0070349955931305885,
          "validation_metric": 0.005680898204445839,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1592820099995151
        },
        {
          "epoch": 79,
          "updates": 79,
          "samples_seen": 80896,
          "online_loss": 0.06984960287809372,
          "train_loss": 0.06896737217903137,
          "validation_loss": 0.055759597569704056,
          "train_metric": 0.006946141831576824,
          "validation_metric": 0.005615903064608574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1677938979992177
        },
        {
          "epoch": 80,
          "updates": 80,
          "samples_seen": 81920,
          "online_loss": 0.06896737217903137,
          "train_loss": 0.06811722368001938,
          "validation_loss": 0.05517415329813957,
          "train_metric": 0.006860516965389252,
          "validation_metric": 0.005556938704103231,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.176036426999417
        },
        {
          "epoch": 81,
          "updates": 81,
          "samples_seen": 82944,
          "online_loss": 0.06811722368001938,
          "train_loss": 0.06731852889060974,
          "validation_loss": 0.05465142801403999,
          "train_metric": 0.006780075840651989,
          "validation_metric": 0.005504291970282793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1844291100005648
        },
        {
          "epoch": 82,
          "updates": 82,
          "samples_seen": 83968,
          "online_loss": 0.06731852889060974,
          "train_loss": 0.06658535450696945,
          "validation_loss": 0.0541897714138031,
          "train_metric": 0.006706233136355877,
          "validation_metric": 0.005457795225083828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1931200330000138
        },
        {
          "epoch": 83,
          "updates": 83,
          "samples_seen": 84992,
          "online_loss": 0.06658535450696945,
          "train_loss": 0.06592170894145966,
          "validation_loss": 0.05378088727593422,
          "train_metric": 0.006639393046498299,
          "validation_metric": 0.005416614003479481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2016662910009472
        },
        {
          "epoch": 84,
          "updates": 84,
          "samples_seen": 86016,
          "online_loss": 0.06592170894145966,
          "train_loss": 0.06532096862792969,
          "validation_loss": 0.05341033264994621,
          "train_metric": 0.006578887812793255,
          "validation_metric": 0.005379293113946915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2103719099995942
        },
        {
          "epoch": 85,
          "updates": 85,
          "samples_seen": 87040,
          "online_loss": 0.06532096117734909,
          "train_loss": 0.06476933509111404,
          "validation_loss": 0.053060635924339294,
          "train_metric": 0.0065233297646045685,
          "validation_metric": 0.005344072822481394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2191066899995349
        },
        {
          "epoch": 86,
          "updates": 86,
          "samples_seen": 88064,
          "online_loss": 0.06476932764053345,
          "train_loss": 0.0642518624663353,
          "validation_loss": 0.05271579325199127,
          "train_metric": 0.006471212022006512,
          "validation_metric": 0.005309341941028833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2276806429999851
        },
        {
          "epoch": 87,
          "updates": 87,
          "samples_seen": 89088,
          "online_loss": 0.0642518550157547,
          "train_loss": 0.06375806778669357,
          "validation_loss": 0.05236562341451645,
          "train_metric": 0.006421479396522045,
          "validation_metric": 0.005274074152112007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2363467609993677
        },
        {
          "epoch": 88,
          "updates": 88,
          "samples_seen": 90112,
          "online_loss": 0.06375806778669357,
          "train_loss": 0.06328494846820831,
          "validation_loss": 0.05200795829296112,
          "train_metric": 0.006373827811330557,
          "validation_metric": 0.005238051526248455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2455414119995112
        },
        {
          "epoch": 89,
          "updates": 89,
          "samples_seen": 91136,
          "online_loss": 0.06328494846820831,
          "train_loss": 0.0628361627459526,
          "validation_loss": 0.05164818465709686,
          "train_metric": 0.006328628398478031,
          "validation_metric": 0.005201816093176603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2544303710001259
        },
        {
          "epoch": 90,
          "updates": 90,
          "samples_seen": 92160,
          "online_loss": 0.0628361701965332,
          "train_loss": 0.062418244779109955,
          "validation_loss": 0.051296256482601166,
          "train_metric": 0.00628653634339571,
          "validation_metric": 0.005166370887309313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2633976319993963
        },
        {
          "epoch": 91,
          "updates": 91,
          "samples_seen": 93184,
          "online_loss": 0.062418244779109955,
          "train_loss": 0.06203577667474747,
          "validation_loss": 0.05096273869276047,
          "train_metric": 0.006248016841709614,
          "validation_metric": 0.005132779944688082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2722587589996692
        },
        {
          "epoch": 92,
          "updates": 92,
          "samples_seen": 94208,
          "online_loss": 0.06203577667474747,
          "train_loss": 0.061688050627708435,
          "validation_loss": 0.0506555512547493,
          "train_metric": 0.006212995387613773,
          "validation_metric": 0.0051018414087593555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2808723010002723
        },
        {
          "epoch": 93,
          "updates": 93,
          "samples_seen": 95232,
          "online_loss": 0.061688054352998734,
          "train_loss": 0.06136844679713249,
          "validation_loss": 0.05037828907370567,
          "train_metric": 0.006180805154144764,
          "validation_metric": 0.005073916167020798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3850411170005827
        },
        {
          "epoch": 94,
          "updates": 94,
          "samples_seen": 96256,
          "online_loss": 0.06136845052242279,
          "train_loss": 0.06106661632657051,
          "validation_loss": 0.05013055354356766,
          "train_metric": 0.0061504049226641655,
          "validation_metric": 0.0050489651039242744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3936389289992803
        },
        {
          "epoch": 95,
          "updates": 95,
          "samples_seen": 97280,
          "online_loss": 0.06106661260128021,
          "train_loss": 0.0607721321284771,
          "validation_loss": 0.04990949109196663,
          "train_metric": 0.00612074788659811,
          "validation_metric": 0.005026700906455517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4020352709994768
        },
        {
          "epoch": 96,
          "updates": 96,
          "samples_seen": 98304,
          "online_loss": 0.060772139579057693,
          "train_loss": 0.06047789379954338,
          "validation_loss": 0.04971146583557129,
          "train_metric": 0.006091112270951271,
          "validation_metric": 0.005006756633520126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.410939679000876
        },
        {
          "epoch": 97,
          "updates": 97,
          "samples_seen": 99328,
          "online_loss": 0.060477904975414276,
          "train_loss": 0.060181625187397,
          "validation_loss": 0.04953300207853317,
          "train_metric": 0.006061272695660591,
          "validation_metric": 0.004988782107830048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4196619360009208
        },
        {
          "epoch": 98,
          "updates": 98,
          "samples_seen": 100352,
          "online_loss": 0.060181621462106705,
          "train_loss": 0.05988515913486481,
          "validation_loss": 0.049370553344488144,
          "train_metric": 0.006031414493918419,
          "validation_metric": 0.00497242109850049,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4283292710006208
        },
        {
          "epoch": 99,
          "updates": 99,
          "samples_seen": 101376,
          "online_loss": 0.05988515913486481,
          "train_loss": 0.05959207937121391,
          "validation_loss": 0.04921960085630417,
          "train_metric": 0.006001895759254694,
          "validation_metric": 0.00495721772313118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.436695153999608
        },
        {
          "epoch": 100,
          "updates": 100,
          "samples_seen": 102400,
          "online_loss": 0.05959208309650421,
          "train_loss": 0.05930487811565399,
          "validation_loss": 0.049073681235313416,
          "train_metric": 0.005972969811409712,
          "validation_metric": 0.00494252098724246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4453412549992208
        },
        {
          "epoch": 101,
          "updates": 101,
          "samples_seen": 103424,
          "online_loss": 0.059304870665073395,
          "train_loss": 0.05902320146560669,
          "validation_loss": 0.04892423376441002,
          "train_metric": 0.00594460079446435,
          "validation_metric": 0.0049274698831140995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.453601442000945
        },
        {
          "epoch": 102,
          "updates": 102,
          "samples_seen": 104448,
          "online_loss": 0.05902320519089699,
          "train_loss": 0.058743856847286224,
          "validation_loss": 0.04876178130507469,
          "train_metric": 0.005916466470807791,
          "validation_metric": 0.0049111079424619675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.46212222900067
        },
        {
          "epoch": 103,
          "updates": 103,
          "samples_seen": 105472,
          "online_loss": 0.05874386057257652,
          "train_loss": 0.05846228823065758,
          "validation_loss": 0.048577819019556046,
          "train_metric": 0.005888108164072037,
          "validation_metric": 0.004892580211162567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4703834480005753
        },
        {
          "epoch": 104,
          "updates": 104,
          "samples_seen": 106496,
          "online_loss": 0.05846228823065758,
          "train_loss": 0.05817470699548721,
          "validation_loss": 0.04836694896221161,
          "train_metric": 0.005859144032001495,
          "validation_metric": 0.004871341399848461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.483856745000594
        },
        {
          "epoch": 105,
          "updates": 105,
          "samples_seen": 107520,
          "online_loss": 0.05817471072077751,
          "train_loss": 0.05787958949804306,
          "validation_loss": 0.04812813177704811,
          "train_metric": 0.005829420872032642,
          "validation_metric": 0.004847289063036442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.492047273000935
        },
        {
          "epoch": 106,
          "updates": 106,
          "samples_seen": 108544,
          "online_loss": 0.05787959322333336,
          "train_loss": 0.05757787078619003,
          "validation_loss": 0.047864656895399094,
          "train_metric": 0.005799032282084227,
          "validation_metric": 0.004820752888917923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5003380140005902
        },
        {
          "epoch": 107,
          "updates": 107,
          "samples_seen": 109568,
          "online_loss": 0.05757787451148033,
          "train_loss": 0.057271845638751984,
          "validation_loss": 0.047582902014255524,
          "train_metric": 0.005768210627138615,
          "validation_metric": 0.004792376421391964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.508594893000918
        },
        {
          "epoch": 108,
          "updates": 108,
          "samples_seen": 110592,
          "online_loss": 0.05727185308933258,
          "train_loss": 0.05696363374590874,
          "validation_loss": 0.04729047045111656,
          "train_metric": 0.0057371677830815315,
          "validation_metric": 0.0047629233449697495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5169412790000933
        },
        {
          "epoch": 109,
          "updates": 109,
          "samples_seen": 111616,
          "online_loss": 0.05696363374590874,
          "train_loss": 0.056654006242752075,
          "validation_loss": 0.046994488686323166,
          "train_metric": 0.0057059843093156815,
          "validation_metric": 0.00473311310634017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.523551034000775
        },
        {
          "epoch": 110,
          "updates": 110,
          "samples_seen": 112640,
          "online_loss": 0.056654006242752075,
          "train_loss": 0.05634216219186783,
          "validation_loss": 0.046700604259967804,
          "train_metric": 0.005674576386809349,
          "validation_metric": 0.004703513812273741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5300351980004052
        },
        {
          "epoch": 111,
          "updates": 111,
          "samples_seen": 113664,
          "online_loss": 0.05634216591715813,
          "train_loss": 0.056026436388492584,
          "validation_loss": 0.04641265422105789,
          "train_metric": 0.005642778240144253,
          "validation_metric": 0.004674512427300215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5360834849998355
        },
        {
          "epoch": 112,
          "updates": 112,
          "samples_seen": 114688,
          "online_loss": 0.056026436388492584,
          "train_loss": 0.05570529028773308,
          "validation_loss": 0.046132851392030716,
          "train_metric": 0.005610432475805283,
          "validation_metric": 0.004646332003176212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.542830591000893
        },
        {
          "epoch": 113,
          "updates": 113,
          "samples_seen": 115712,
          "online_loss": 0.05570528656244278,
          "train_loss": 0.05537811666727066,
          "validation_loss": 0.04586197808384895,
          "train_metric": 0.005577481351792812,
          "validation_metric": 0.0046190498396754265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5487374269996508
        },
        {
          "epoch": 114,
          "updates": 114,
          "samples_seen": 116736,
          "online_loss": 0.05537812039256096,
          "train_loss": 0.05504540354013443,
          "validation_loss": 0.045599520206451416,
          "train_metric": 0.005543971434235573,
          "validation_metric": 0.00459261704236269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.554643395000312
        },
        {
          "epoch": 115,
          "updates": 115,
          "samples_seen": 117760,
          "online_loss": 0.05504540354013443,
          "train_loss": 0.05470822751522064,
          "validation_loss": 0.045343637466430664,
          "train_metric": 0.005510011687874794,
          "validation_metric": 0.0045668454840779305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5610373700001219
        },
        {
          "epoch": 116,
          "updates": 116,
          "samples_seen": 118784,
          "online_loss": 0.054708220064640045,
          "train_loss": 0.05436762049794197,
          "validation_loss": 0.045091256499290466,
          "train_metric": 0.00547570688650012,
          "validation_metric": 0.00454142689704895,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5668834200005222
        },
        {
          "epoch": 117,
          "updates": 117,
          "samples_seen": 119808,
          "online_loss": 0.05436762422323227,
          "train_loss": 0.05402407422661781,
          "validation_loss": 0.04483841732144356,
          "train_metric": 0.005441107787191868,
          "validation_metric": 0.004515961278229952,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.573207478999393
        },
        {
          "epoch": 118,
          "updates": 118,
          "samples_seen": 120832,
          "online_loss": 0.05402407795190811,
          "train_loss": 0.05367743968963623,
          "validation_loss": 0.04458093270659447,
          "train_metric": 0.005406195297837257,
          "validation_metric": 0.0044900281354784966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.579609499000071
        },
        {
          "epoch": 119,
          "updates": 119,
          "samples_seen": 121856,
          "online_loss": 0.05367743968963623,
          "train_loss": 0.05332721769809723,
          "validation_loss": 0.044315192848443985,
          "train_metric": 0.005370921455323696,
          "validation_metric": 0.004463263787329197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5855684060006752
        },
        {
          "epoch": 120,
          "updates": 120,
          "samples_seen": 122880,
          "online_loss": 0.05332722142338753,
          "train_loss": 0.0529729425907135,
          "validation_loss": 0.04403886944055557,
          "train_metric": 0.005335241090506315,
          "validation_metric": 0.00443543354049325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.591747566999402
        },
        {
          "epoch": 121,
          "updates": 121,
          "samples_seen": 123904,
          "online_loss": 0.0529729463160038,
          "train_loss": 0.052614495158195496,
          "validation_loss": 0.04375121369957924,
          "train_metric": 0.0052991388365626335,
          "validation_metric": 0.004406461957842112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.597776619000797
        },
        {
          "epoch": 122,
          "updates": 122,
          "samples_seen": 124928,
          "online_loss": 0.0526144914329052,
          "train_loss": 0.05225200578570366,
          "validation_loss": 0.04345312714576721,
          "train_metric": 0.005262630991637707,
          "validation_metric": 0.004376440308988094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6034221940008138
        },
        {
          "epoch": 123,
          "updates": 123,
          "samples_seen": 125952,
          "online_loss": 0.05225200578570366,
          "train_loss": 0.05188574641942978,
          "validation_loss": 0.04314665123820305,
          "train_metric": 0.00522574270144105,
          "validation_metric": 0.004345572553575039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6094558849999885
        },
        {
          "epoch": 124,
          "updates": 124,
          "samples_seen": 126976,
          "online_loss": 0.05188575014472008,
          "train_loss": 0.05151589214801788,
          "validation_loss": 0.04283452406525612,
          "train_metric": 0.005188492126762867,
          "validation_metric": 0.004314136691391468,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6153111650000938
        },
        {
          "epoch": 125,
          "updates": 125,
          "samples_seen": 128000,
          "online_loss": 0.05151589587330818,
          "train_loss": 0.05114242061972618,
          "validation_loss": 0.04251964017748833,
          "train_metric": 0.005150876939296722,
          "validation_metric": 0.0042824228294193745,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6209973240002
        },
        {
          "epoch": 126,
          "updates": 126,
          "samples_seen": 129024,
          "online_loss": 0.05114242434501648,
          "train_loss": 0.05076514557003975,
          "validation_loss": 0.04220451042056084,
          "train_metric": 0.005112878978252411,
          "validation_metric": 0.004250683356076479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.627267395000672
        },
        {
          "epoch": 127,
          "updates": 127,
          "samples_seen": 130048,
          "online_loss": 0.05076514184474945,
          "train_loss": 0.050383780151605606,
          "validation_loss": 0.041890908032655716,
          "train_metric": 0.005074470303952694,
          "validation_metric": 0.004219098947942257,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6331272120005451
        },
        {
          "epoch": 128,
          "updates": 128,
          "samples_seen": 131072,
          "online_loss": 0.050383783876895905,
          "train_loss": 0.04999801516532898,
          "validation_loss": 0.041579559445381165,
          "train_metric": 0.005035617388784885,
          "validation_metric": 0.004187741316854954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6387896470005217
        },
        {
          "epoch": 129,
          "updates": 129,
          "samples_seen": 132096,
          "online_loss": 0.04999801889061928,
          "train_loss": 0.0496075265109539,
          "validation_loss": 0.04127011448144913,
          "train_metric": 0.004996289033442736,
          "validation_metric": 0.004156574606895447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6447180279992608
        },
        {
          "epoch": 130,
          "updates": 130,
          "samples_seen": 133120,
          "online_loss": 0.0496075265109539,
          "train_loss": 0.0492120236158371,
          "validation_loss": 0.040961213409900665,
          "train_metric": 0.004956455901265144,
          "validation_metric": 0.004125463776290417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.650673681000626
        },
        {
          "epoch": 131,
          "updates": 131,
          "samples_seen": 134144,
          "online_loss": 0.049212031066417694,
          "train_loss": 0.04881126061081886,
          "validation_loss": 0.04065096378326416,
          "train_metric": 0.004916091449558735,
          "validation_metric": 0.0040942165069282055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7460651879991929
        },
        {
          "epoch": 132,
          "updates": 132,
          "samples_seen": 135168,
          "online_loss": 0.04881126061081886,
          "train_loss": 0.048405107110738754,
          "validation_loss": 0.04033726081252098,
          "train_metric": 0.004875184968113899,
          "validation_metric": 0.004062621854245663,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7524342270007764
        },
        {
          "epoch": 133,
          "updates": 133,
          "samples_seen": 136192,
          "online_loss": 0.048405103385448456,
          "train_loss": 0.04799354821443558,
          "validation_loss": 0.040018290281295776,
          "train_metric": 0.0048337336629629135,
          "validation_metric": 0.0040304954163730145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7585391270004038
        },
        {
          "epoch": 134,
          "updates": 134,
          "samples_seen": 137216,
          "online_loss": 0.04799354076385498,
          "train_loss": 0.04757656157016754,
          "validation_loss": 0.03969267010688782,
          "train_metric": 0.004791737999767065,
          "validation_metric": 0.003997700288891792,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7646584159992926
        },
        {
          "epoch": 135,
          "updates": 135,
          "samples_seen": 138240,
          "online_loss": 0.04757656902074814,
          "train_loss": 0.04715413227677345,
          "validation_loss": 0.039359599351882935,
          "train_metric": 0.004749190993607044,
          "validation_metric": 0.003964154981076717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7704742930000066
        },
        {
          "epoch": 136,
          "updates": 136,
          "samples_seen": 139264,
          "online_loss": 0.04715413600206375,
          "train_loss": 0.04672609642148018,
          "validation_loss": 0.03901877626776695,
          "train_metric": 0.004706081468611956,
          "validation_metric": 0.00392982829362154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.776689633999922
        },
        {
          "epoch": 137,
          "updates": 137,
          "samples_seen": 140288,
          "online_loss": 0.04672609642148018,
          "train_loss": 0.0462922677397728,
          "validation_loss": 0.038670431822538376,
          "train_metric": 0.004662388004362583,
          "validation_metric": 0.0038947444409132004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.783139831999506
        },
        {
          "epoch": 138,
          "updates": 138,
          "samples_seen": 141312,
          "online_loss": 0.0462922677397728,
          "train_loss": 0.04585254192352295,
          "validation_loss": 0.03831511363387108,
          "train_metric": 0.004618100356310606,
          "validation_metric": 0.003858958138152957,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7893246919993544
        },
        {
          "epoch": 139,
          "updates": 139,
          "samples_seen": 142336,
          "online_loss": 0.04585254192352295,
          "train_loss": 0.045406971126794815,
          "validation_loss": 0.037953659892082214,
          "train_metric": 0.0045732236467301846,
          "validation_metric": 0.003822553437203169,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7991877639997256
        },
        {
          "epoch": 140,
          "updates": 140,
          "samples_seen": 143360,
          "online_loss": 0.04540696740150452,
          "train_loss": 0.044955700635910034,
          "validation_loss": 0.03758692368865013,
          "train_metric": 0.004527773708105087,
          "validation_metric": 0.0037856174167245626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8049928300006286
        },
        {
          "epoch": 141,
          "updates": 141,
          "samples_seen": 144384,
          "online_loss": 0.044955700635910034,
          "train_loss": 0.04449890926480293,
          "validation_loss": 0.037215523421764374,
          "train_metric": 0.004481767304241657,
          "validation_metric": 0.0037482113111764193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8140261320004356
        },
        {
          "epoch": 142,
          "updates": 142,
          "samples_seen": 145408,
          "online_loss": 0.044498905539512634,
          "train_loss": 0.0440366230905056,
          "validation_loss": 0.03683977574110031,
          "train_metric": 0.004435207229107618,
          "validation_metric": 0.003710367251187563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8202475670004787
        },
        {
          "epoch": 143,
          "updates": 143,
          "samples_seen": 146432,
          "online_loss": 0.0440366193652153,
          "train_loss": 0.043568823486566544,
          "validation_loss": 0.036459796130657196,
          "train_metric": 0.004388092551380396,
          "validation_metric": 0.0036720966454595327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8299650000008114
        },
        {
          "epoch": 144,
          "updates": 144,
          "samples_seen": 147456,
          "online_loss": 0.04356881603598595,
          "train_loss": 0.043095484375953674,
          "validation_loss": 0.03607569634914398,
          "train_metric": 0.0043404195457696915,
          "validation_metric": 0.003633412066847086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8368240290001268
        },
        {
          "epoch": 145,
          "updates": 145,
          "samples_seen": 148480,
          "online_loss": 0.043095484375953674,
          "train_loss": 0.04261673614382744,
          "validation_loss": 0.03568778932094574,
          "train_metric": 0.004292201716452837,
          "validation_metric": 0.0035943433176726103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8455726160009363
        },
        {
          "epoch": 146,
          "updates": 146,
          "samples_seen": 149504,
          "online_loss": 0.04261673241853714,
          "train_loss": 0.04213280975818634,
          "validation_loss": 0.03529653698205948,
          "train_metric": 0.004243462346494198,
          "validation_metric": 0.0035549381282180548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8517504920000647
        },
        {
          "epoch": 147,
          "updates": 147,
          "samples_seen": 150528,
          "online_loss": 0.04213280975818634,
          "train_loss": 0.04164402186870575,
          "validation_loss": 0.034902423620224,
          "train_metric": 0.004194233100861311,
          "validation_metric": 0.0035152442287653685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8608178850008699
        },
        {
          "epoch": 148,
          "updates": 148,
          "samples_seen": 151552,
          "online_loss": 0.04164402186870575,
          "train_loss": 0.04115065187215805,
          "validation_loss": 0.03450560197234154,
          "train_metric": 0.004144543316215277,
          "validation_metric": 0.00347527745179832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.866778087000057
        },
        {
          "epoch": 149,
          "updates": 149,
          "samples_seen": 152576,
          "online_loss": 0.04115065187215805,
          "train_loss": 0.040652986615896225,
          "validation_loss": 0.034105800092220306,
          "train_metric": 0.004094419535249472,
          "validation_metric": 0.003435011487454176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8738806979999936
        },
        {
          "epoch": 150,
          "updates": 150,
          "samples_seen": 153600,
          "online_loss": 0.04065297916531563,
          "train_loss": 0.040151309221982956,
          "validation_loss": 0.03370228782296181,
          "train_metric": 0.004043893422931433,
          "validation_metric": 0.003394371597096324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.880519299000298
        },
        {
          "epoch": 151,
          "updates": 151,
          "samples_seen": 154624,
          "online_loss": 0.040151312947273254,
          "train_loss": 0.03964599221944809,
          "validation_loss": 0.03329404816031456,
          "train_metric": 0.003992999438196421,
          "validation_metric": 0.003353254869580269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8863259200006723
        },
        {
          "epoch": 152,
          "updates": 152,
          "samples_seen": 155648,
          "online_loss": 0.03964599221944809,
          "train_loss": 0.03913738951086998,
          "validation_loss": 0.03288005292415619,
          "train_metric": 0.003941774368286133,
          "validation_metric": 0.0033115586265921593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8925977500002773
        },
        {
          "epoch": 153,
          "updates": 153,
          "samples_seen": 156672,
          "online_loss": 0.03913738951086998,
          "train_loss": 0.03862586244940758,
          "validation_loss": 0.03245968371629715,
          "train_metric": 0.003890255931764841,
          "validation_metric": 0.0032692207023501396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8993821179992665
        },
        {
          "epoch": 154,
          "updates": 154,
          "samples_seen": 157696,
          "online_loss": 0.038625869899988174,
          "train_loss": 0.03811182081699371,
          "validation_loss": 0.03203314542770386,
          "train_metric": 0.003838482778519392,
          "validation_metric": 0.003226260654628277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9051837170009094
        },
        {
          "epoch": 155,
          "updates": 155,
          "samples_seen": 158720,
          "online_loss": 0.038111817091703415,
          "train_loss": 0.03759568929672241,
          "validation_loss": 0.03160158172249794,
          "train_metric": 0.003786499612033367,
          "validation_metric": 0.003182795364409685,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9114941740008362
        },
        {
          "epoch": 156,
          "updates": 156,
          "samples_seen": 159744,
          "online_loss": 0.037595681846141815,
          "train_loss": 0.03707795590162277,
          "validation_loss": 0.031166914850473404,
          "train_metric": 0.003734356025233865,
          "validation_metric": 0.0031390171498060226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.917531331999271
        },
        {
          "epoch": 157,
          "updates": 157,
          "samples_seen": 160768,
          "online_loss": 0.03707795590162277,
          "train_loss": 0.0365590900182724,
          "validation_loss": 0.0307312048971653,
          "train_metric": 0.003682097652927041,
          "validation_metric": 0.0030951346270740032,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9232809570003155
        },
        {
          "epoch": 158,
          "updates": 158,
          "samples_seen": 161792,
          "online_loss": 0.0365590900182724,
          "train_loss": 0.036039549857378006,
          "validation_loss": 0.030295955017209053,
          "train_metric": 0.0036297712940722704,
          "validation_metric": 0.00305129773914814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.929710635000447
        },
        {
          "epoch": 159,
          "updates": 159,
          "samples_seen": 162816,
          "online_loss": 0.036039549857378006,
          "train_loss": 0.03551984950900078,
          "validation_loss": 0.029861686751246452,
          "train_metric": 0.0035774291027337313,
          "validation_metric": 0.0030075598042458296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9354271200008952
        },
        {
          "epoch": 160,
          "updates": 160,
          "samples_seen": 163840,
          "online_loss": 0.03551984950900078,
          "train_loss": 0.035000547766685486,
          "validation_loss": 0.029427718371152878,
          "train_metric": 0.0035251271910965443,
          "validation_metric": 0.0029638519044965506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9417255810003553
        },
        {
          "epoch": 161,
          "updates": 161,
          "samples_seen": 164864,
          "online_loss": 0.035000551491975784,
          "train_loss": 0.03448221832513809,
          "validation_loss": 0.028992759063839912,
          "train_metric": 0.0034729233011603355,
          "validation_metric": 0.002920044120401144,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9477025190008135
        },
        {
          "epoch": 162,
          "updates": 162,
          "samples_seen": 165888,
          "online_loss": 0.03448222205042839,
          "train_loss": 0.033965401351451874,
          "validation_loss": 0.02855565771460533,
          "train_metric": 0.003420870751142502,
          "validation_metric": 0.002876021433621645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9535244390008302
        },
        {
          "epoch": 163,
          "updates": 163,
          "samples_seen": 166912,
          "online_loss": 0.03396540507674217,
          "train_loss": 0.033450670540332794,
          "validation_loss": 0.028116250410676003,
          "train_metric": 0.0033690291456878185,
          "validation_metric": 0.002831765916198492,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9594550500005425
        },
        {
          "epoch": 164,
          "updates": 164,
          "samples_seen": 167936,
          "online_loss": 0.033450666815042496,
          "train_loss": 0.0329386405646801,
          "validation_loss": 0.027675678953528404,
          "train_metric": 0.003317459486424923,
          "validation_metric": 0.0027873930521309376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.965646339000159
        },
        {
          "epoch": 165,
          "updates": 165,
          "samples_seen": 168960,
          "online_loss": 0.0329386442899704,
          "train_loss": 0.032429955899715424,
          "validation_loss": 0.027235977351665497,
          "train_metric": 0.003266226267442107,
          "validation_metric": 0.0027431081980466843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9714027690006333
        },
        {
          "epoch": 166,
          "updates": 166,
          "samples_seen": 169984,
          "online_loss": 0.032429955899715424,
          "train_loss": 0.031925223767757416,
          "validation_loss": 0.026799123734235764,
          "train_metric": 0.0032153914216905832,
          "validation_metric": 0.002699109259992838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9777130069996929
        },
        {
          "epoch": 167,
          "updates": 167,
          "samples_seen": 171008,
          "online_loss": 0.03192522004246712,
          "train_loss": 0.03142505884170532,
          "validation_loss": 0.026366285979747772,
          "train_metric": 0.0031650171149522066,
          "validation_metric": 0.002655515680089593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.983582323000519
        },
        {
          "epoch": 168,
          "updates": 168,
          "samples_seen": 172032,
          "online_loss": 0.03142506256699562,
          "train_loss": 0.030930064618587494,
          "validation_loss": 0.025937583297491074,
          "train_metric": 0.003115162719041109,
          "validation_metric": 0.0026123386342078447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0682980569999927
        },
        {
          "epoch": 169,
          "updates": 169,
          "samples_seen": 173056,
          "online_loss": 0.030930066481232643,
          "train_loss": 0.030440794304013252,
          "validation_loss": 0.0255126953125,
          "train_metric": 0.003065885277464986,
          "validation_metric": 0.002569545293226838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0742211760007194
        },
        {
          "epoch": 170,
          "updates": 170,
          "samples_seen": 174080,
          "online_loss": 0.030440792441368103,
          "train_loss": 0.02995777502655983,
          "validation_loss": 0.025091754272580147,
          "train_metric": 0.003017237177118659,
          "validation_metric": 0.0025271496269851923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.07996076299969
        },
        {
          "epoch": 171,
          "updates": 171,
          "samples_seen": 175104,
          "online_loss": 0.02995777502655983,
          "train_loss": 0.02948158048093319,
          "validation_loss": 0.02467595785856247,
          "train_metric": 0.0029692770913243294,
          "validation_metric": 0.0024852720089256763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0856432439995842
        },
        {
          "epoch": 172,
          "updates": 172,
          "samples_seen": 176128,
          "online_loss": 0.02948158234357834,
          "train_loss": 0.029012786224484444,
          "validation_loss": 0.024266991764307022,
          "train_metric": 0.0029220618307590485,
          "validation_metric": 0.002444082871079445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.091411503000927
        },
        {
          "epoch": 173,
          "updates": 173,
          "samples_seen": 177152,
          "online_loss": 0.029012786224484444,
          "train_loss": 0.028551889583468437,
          "validation_loss": 0.023865897208452225,
          "train_metric": 0.002875641919672489,
          "validation_metric": 0.002403685823082924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0974065690006682
        },
        {
          "epoch": 174,
          "updates": 174,
          "samples_seen": 178176,
          "online_loss": 0.028551891446113586,
          "train_loss": 0.028099365532398224,
          "validation_loss": 0.023472392931580544,
          "train_metric": 0.002830065321177244,
          "validation_metric": 0.002364053390920162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1034765770000377
        },
        {
          "epoch": 175,
          "updates": 175,
          "samples_seen": 179200,
          "online_loss": 0.028099369257688522,
          "train_loss": 0.027655644342303276,
          "validation_loss": 0.02308523654937744,
          "train_metric": 0.002785374876111746,
          "validation_metric": 0.0023250600788742304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.109783747999245
        },
        {
          "epoch": 176,
          "updates": 176,
          "samples_seen": 180224,
          "online_loss": 0.027655644342303276,
          "train_loss": 0.027221055701375008,
          "validation_loss": 0.022703608497977257,
          "train_metric": 0.0027416052762418985,
          "validation_metric": 0.0022866243962198496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.115619203999813
        },
        {
          "epoch": 177,
          "updates": 177,
          "samples_seen": 181248,
          "online_loss": 0.02722105383872986,
          "train_loss": 0.026795951649546623,
          "validation_loss": 0.02232828177511692,
          "train_metric": 0.0026987900491803885,
          "validation_metric": 0.0022488227114081383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.12141300499934
        },
        {
          "epoch": 178,
          "updates": 178,
          "samples_seen": 182272,
          "online_loss": 0.026795949786901474,
          "train_loss": 0.026380637660622597,
          "validation_loss": 0.021961314603686333,
          "train_metric": 0.0026569608598947525,
          "validation_metric": 0.0022118634078651667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.127476485000443
        },
        {
          "epoch": 179,
          "updates": 179,
          "samples_seen": 183296,
          "online_loss": 0.026380637660622597,
          "train_loss": 0.025975331664085388,
          "validation_loss": 0.021604424342513084,
          "train_metric": 0.002616140292957425,
          "validation_metric": 0.002175918547436595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.133368655000595
        },
        {
          "epoch": 180,
          "updates": 180,
          "samples_seen": 184320,
          "online_loss": 0.025975331664085388,
          "train_loss": 0.02558022178709507,
          "validation_loss": 0.021257640793919563,
          "train_metric": 0.0025763465091586113,
          "validation_metric": 0.0021409918554127216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1389551280008163
        },
        {
          "epoch": 181,
          "updates": 181,
          "samples_seen": 185344,
          "online_loss": 0.02558022178709507,
          "train_loss": 0.025195438414812088,
          "validation_loss": 0.020919619128108025,
          "train_metric": 0.002537592314183712,
          "validation_metric": 0.002106947358697653,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1447551290002593
        },
        {
          "epoch": 182,
          "updates": 182,
          "samples_seen": 186368,
          "online_loss": 0.02519543655216694,
          "train_loss": 0.024821080267429352,
          "validation_loss": 0.020589401945471764,
          "train_metric": 0.002499887952581048,
          "validation_metric": 0.002073689131066203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1505863399997907
        },
        {
          "epoch": 183,
          "updates": 183,
          "samples_seen": 187392,
          "online_loss": 0.024821078404784203,
          "train_loss": 0.024457227438688278,
          "validation_loss": 0.020267533138394356,
          "train_metric": 0.002463242504745722,
          "validation_metric": 0.002041271887719631,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.156684211000538
        },
        {
          "epoch": 184,
          "updates": 184,
          "samples_seen": 188416,
          "online_loss": 0.024457227438688278,
          "train_loss": 0.024103917181491852,
          "validation_loss": 0.01995517686009407,
          "train_metric": 0.00242765829898417,
          "validation_metric": 0.0020098122768104076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1629646019991924
        },
        {
          "epoch": 185,
          "updates": 185,
          "samples_seen": 189440,
          "online_loss": 0.024103917181491852,
          "train_loss": 0.023761125281453133,
          "validation_loss": 0.019652405753731728,
          "train_metric": 0.002393133472651243,
          "validation_metric": 0.0019793184474110603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1687896830007958
        },
        {
          "epoch": 186,
          "updates": 186,
          "samples_seen": 190464,
          "online_loss": 0.023761127144098282,
          "train_loss": 0.02342882566154003,
          "validation_loss": 0.019358089193701744,
          "train_metric": 0.002359665697440505,
          "validation_metric": 0.0019496759632602334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1749359980003646
        },
        {
          "epoch": 187,
          "updates": 187,
          "samples_seen": 191488,
          "online_loss": 0.02342882752418518,
          "train_loss": 0.023106925189495087,
          "validation_loss": 0.019071459770202637,
          "train_metric": 0.0023272449616342783,
          "validation_metric": 0.0019208076409995556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.180805895000958
        },
        {
          "epoch": 188,
          "updates": 188,
          "samples_seen": 192512,
          "online_loss": 0.023106925189495087,
          "train_loss": 0.02279531955718994,
          "validation_loss": 0.018793577328324318,
          "train_metric": 0.0022958610206842422,
          "validation_metric": 0.001892820349894464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1864979799993307
        },
        {
          "epoch": 189,
          "updates": 189,
          "samples_seen": 193536,
          "online_loss": 0.02279532141983509,
          "train_loss": 0.02249387465417385,
          "validation_loss": 0.018526069819927216,
          "train_metric": 0.002265501068904996,
          "validation_metric": 0.0018658781191334128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.192558151000412
        },
        {
          "epoch": 190,
          "updates": 190,
          "samples_seen": 194560,
          "online_loss": 0.022493876516819,
          "train_loss": 0.022202426567673683,
          "validation_loss": 0.018268991261720657,
          "train_metric": 0.0022361474111676216,
          "validation_metric": 0.0018399859545752406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.198306861000674
        },
        {
          "epoch": 191,
          "updates": 191,
          "samples_seen": 195584,
          "online_loss": 0.022202428430318832,
          "train_loss": 0.021920789033174515,
          "validation_loss": 0.01802077516913414,
          "train_metric": 0.0022077816538512707,
          "validation_metric": 0.001814986695535481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2040736510007264
        },
        {
          "epoch": 192,
          "updates": 192,
          "samples_seen": 196608,
          "online_loss": 0.021920789033174515,
          "train_loss": 0.02164873480796814,
          "validation_loss": 0.01778046041727066,
          "train_metric": 0.0021803812123835087,
          "validation_metric": 0.001790782785974443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2103462340000988
        },
        {
          "epoch": 193,
          "updates": 193,
          "samples_seen": 197632,
          "online_loss": 0.02164873480796814,
          "train_loss": 0.021386023610830307,
          "validation_loss": 0.017548788338899612,
          "train_metric": 0.0021539221052080393,
          "validation_metric": 0.0017674496630206704,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2160735620000196
        },
        {
          "epoch": 194,
          "updates": 194,
          "samples_seen": 198656,
          "online_loss": 0.021386023610830307,
          "train_loss": 0.021132394671440125,
          "validation_loss": 0.017326459288597107,
          "train_metric": 0.0021283773239701986,
          "validation_metric": 0.0017450572922825813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2224623460006114
        },
        {
          "epoch": 195,
          "updates": 195,
          "samples_seen": 199680,
          "online_loss": 0.021132394671440125,
          "train_loss": 0.020887546241283417,
          "validation_loss": 0.017112668603658676,
          "train_metric": 0.002103717066347599,
          "validation_metric": 0.0017235249979421496,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2289407749995007
        },
        {
          "epoch": 196,
          "updates": 196,
          "samples_seen": 200704,
          "online_loss": 0.020887546241283417,
          "train_loss": 0.02065117098391056,
          "validation_loss": 0.016906360164284706,
          "train_metric": 0.002079910133033991,
          "validation_metric": 0.0017027466092258692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.234582731000046
        },
        {
          "epoch": 197,
          "updates": 197,
          "samples_seen": 201728,
          "online_loss": 0.02065116912126541,
          "train_loss": 0.020422937348484993,
          "validation_loss": 0.016708076000213623,
          "train_metric": 0.0020569232292473316,
          "validation_metric": 0.0016827763756737113,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2405139279999275
        },
        {
          "epoch": 198,
          "updates": 198,
          "samples_seen": 202752,
          "online_loss": 0.020422935485839844,
          "train_loss": 0.020202511921525,
          "validation_loss": 0.016518671065568924,
          "train_metric": 0.002034723060205579,
          "validation_metric": 0.0016637000953778625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.246321861999604
        },
        {
          "epoch": 199,
          "updates": 199,
          "samples_seen": 203776,
          "online_loss": 0.020202510058879852,
          "train_loss": 0.019989557564258575,
          "validation_loss": 0.016337184235453606,
          "train_metric": 0.0020132751669734716,
          "validation_metric": 0.00164542137645185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.251972772000954
        },
        {
          "epoch": 200,
          "updates": 200,
          "samples_seen": 204800,
          "online_loss": 0.019989555701613426,
          "train_loss": 0.01978372223675251,
          "validation_loss": 0.016161827370524406,
          "train_metric": 0.0019925436936318874,
          "validation_metric": 0.0016277600079774857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2576637090005534
        },
        {
          "epoch": 201,
          "updates": 201,
          "samples_seen": 205824,
          "online_loss": 0.01978371851146221,
          "train_loss": 0.019584650173783302,
          "validation_loss": 0.015992432832717896,
          "train_metric": 0.001972494414076209,
          "validation_metric": 0.0016106993425637484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2640472710008908
        },
        {
          "epoch": 202,
          "updates": 202,
          "samples_seen": 206848,
          "online_loss": 0.01958465203642845,
          "train_loss": 0.019391996785998344,
          "validation_loss": 0.01582961343228817,
          "train_metric": 0.001953090657480061,
          "validation_metric": 0.0015943007310852408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2703900569995312
        },
        {
          "epoch": 203,
          "updates": 203,
          "samples_seen": 207872,
          "online_loss": 0.019391993060708046,
          "train_loss": 0.01920541562139988,
          "validation_loss": 0.015672655776143074,
          "train_metric": 0.0019342991290614009,
          "validation_metric": 0.0015784924617037177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2767243680009415
        },
        {
          "epoch": 204,
          "updates": 204,
          "samples_seen": 208896,
          "online_loss": 0.01920541748404503,
          "train_loss": 0.01902456395328045,
          "validation_loss": 0.015520447865128517,
          "train_metric": 0.0019160843221470714,
          "validation_metric": 0.0015631626592949033,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.282729114000176
        },
        {
          "epoch": 205,
          "updates": 205,
          "samples_seen": 209920,
          "online_loss": 0.01902456395328045,
          "train_loss": 0.018849104642868042,
          "validation_loss": 0.015373315662145615,
          "train_metric": 0.0018984127091243863,
          "validation_metric": 0.0015483441529795527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.288411998000811
        },
        {
          "epoch": 206,
          "updates": 206,
          "samples_seen": 210944,
          "online_loss": 0.018849104642868042,
          "train_loss": 0.018678706139326096,
          "validation_loss": 0.015231634490191936,
          "train_metric": 0.0018812509952113032,
          "validation_metric": 0.0015340744284912944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3770723739999085
        },
        {
          "epoch": 207,
          "updates": 207,
          "samples_seen": 211968,
          "online_loss": 0.018678706139326096,
          "train_loss": 0.01851305365562439,
          "validation_loss": 0.015094072557985783,
          "train_metric": 0.0018645668169483542,
          "validation_metric": 0.0015202196082100272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3832857580000564
        },
        {
          "epoch": 208,
          "updates": 208,
          "samples_seen": 212992,
          "online_loss": 0.01851305179297924,
          "train_loss": 0.018351847305893898,
          "validation_loss": 0.014959624037146568,
          "train_metric": 0.0018483306048437953,
          "validation_metric": 0.0015066786436364055,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.389048125000045
        },
        {
          "epoch": 209,
          "updates": 209,
          "samples_seen": 214016,
          "online_loss": 0.01835184544324875,
          "train_loss": 0.018194779753684998,
          "validation_loss": 0.01482880488038063,
          "train_metric": 0.0018325112760066986,
          "validation_metric": 0.0014935029903426766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3954020630008017
        },
        {
          "epoch": 210,
          "updates": 210,
          "samples_seen": 215040,
          "online_loss": 0.018194779753684998,
          "train_loss": 0.01804158464074135,
          "validation_loss": 0.014701683074235916,
          "train_metric": 0.0018170821713283658,
          "validation_metric": 0.0014806996332481503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.40127535099964
        },
        {
          "epoch": 211,
          "updates": 211,
          "samples_seen": 216064,
          "online_loss": 0.01804158464074135,
          "train_loss": 0.017891990020871162,
          "validation_loss": 0.014577329158782959,
          "train_metric": 0.0018020154675468802,
          "validation_metric": 0.001468175440095365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.407709860999603
        },
        {
          "epoch": 212,
          "updates": 212,
          "samples_seen": 217088,
          "online_loss": 0.017891988158226013,
          "train_loss": 0.01774575188755989,
          "validation_loss": 0.014455876313149929,
          "train_metric": 0.0017872871831059456,
          "validation_metric": 0.0014559428673237562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.41717456600054
        },
        {
          "epoch": 213,
          "updates": 213,
          "samples_seen": 218112,
          "online_loss": 0.01774575188755989,
          "train_loss": 0.01760263927280903,
          "validation_loss": 0.014337806031107903,
          "train_metric": 0.0017728733364492655,
          "validation_metric": 0.0014440515078604221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4258943400000135
        },
        {
          "epoch": 214,
          "updates": 214,
          "samples_seen": 219136,
          "online_loss": 0.017602641135454178,
          "train_loss": 0.017462443560361862,
          "validation_loss": 0.01422226894646883,
          "train_metric": 0.0017587533220648766,
          "validation_metric": 0.0014324148651212454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4321455959998275
        },
        {
          "epoch": 215,
          "updates": 215,
          "samples_seen": 220160,
          "online_loss": 0.01746244542300701,
          "train_loss": 0.017324956133961678,
          "validation_loss": 0.014108626171946526,
          "train_metric": 0.0017449060687795281,
          "validation_metric": 0.0014209692599251866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4377438600004098
        },
        {
          "epoch": 216,
          "updates": 216,
          "samples_seen": 221184,
          "online_loss": 0.017324956133961678,
          "train_loss": 0.01719001494348049,
          "validation_loss": 0.013997439295053482,
          "train_metric": 0.0017313151620328426,
          "validation_metric": 0.0014097709208726883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.445511758000066
        },
        {
          "epoch": 217,
          "updates": 217,
          "samples_seen": 222208,
          "online_loss": 0.017190013080835342,
          "train_loss": 0.01705744117498398,
          "validation_loss": 0.013888632878661156,
          "train_metric": 0.0017179630231112242,
          "validation_metric": 0.0013988121645525098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.451509652000823
        },
        {
          "epoch": 218,
          "updates": 218,
          "samples_seen": 223232,
          "online_loss": 0.01705744117498398,
          "train_loss": 0.016927087679505348,
          "validation_loss": 0.013781637884676456,
          "train_metric": 0.0017048345180228353,
          "validation_metric": 0.0013880361802875996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4607615349996195
        },
        {
          "epoch": 219,
          "updates": 219,
          "samples_seen": 224256,
          "online_loss": 0.016927089542150497,
          "train_loss": 0.0167988333851099,
          "validation_loss": 0.013676902279257774,
          "train_metric": 0.0016919169574975967,
          "validation_metric": 0.0013774875551462173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4669317309999315
        },
        {
          "epoch": 220,
          "updates": 220,
          "samples_seen": 225280,
          "online_loss": 0.0167988333851099,
          "train_loss": 0.016672546043992043,
          "validation_loss": 0.013574469834566116,
          "train_metric": 0.0016791976522654295,
          "validation_metric": 0.0013671707129105926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4753716769992025
        },
        {
          "epoch": 221,
          "updates": 221,
          "samples_seen": 226304,
          "online_loss": 0.016672546043992043,
          "train_loss": 0.016548123210668564,
          "validation_loss": 0.013473549857735634,
          "train_metric": 0.0016666664741933346,
          "validation_metric": 0.0013570066075772047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.481581064999773
        },
        {
          "epoch": 222,
          "updates": 222,
          "samples_seen": 227328,
          "online_loss": 0.016548123210668564,
          "train_loss": 0.016425468027591705,
          "validation_loss": 0.0133742680773139,
          "train_metric": 0.0016543129459023476,
          "validation_metric": 0.0013470071135088801,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.487109072999374
        },
        {
          "epoch": 223,
          "updates": 223,
          "samples_seen": 228352,
          "online_loss": 0.016425466164946556,
          "train_loss": 0.01630450040102005,
          "validation_loss": 0.013276929967105389,
          "train_metric": 0.0016421295003965497,
          "validation_metric": 0.0013372038956731558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4967647539997415
        },
        {
          "epoch": 224,
          "updates": 224,
          "samples_seen": 229376,
          "online_loss": 0.01630450040102005,
          "train_loss": 0.016185149550437927,
          "validation_loss": 0.013181084766983986,
          "train_metric": 0.0016301089199259877,
          "validation_metric": 0.0013275506207719445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5037028179995104
        },
        {
          "epoch": 225,
          "updates": 225,
          "samples_seen": 230400,
          "online_loss": 0.016185151413083076,
          "train_loss": 0.01606736145913601,
          "validation_loss": 0.013086820021271706,
          "train_metric": 0.001618245616555214,
          "validation_metric": 0.0013180566020309925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5123530979999487
        },
        {
          "epoch": 226,
          "updates": 226,
          "samples_seen": 231424,
          "online_loss": 0.01606736145913601,
          "train_loss": 0.015951069071888924,
          "validation_loss": 0.012994443997740746,
          "train_metric": 0.0016065333038568497,
          "validation_metric": 0.0013087530387565494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5188364349996846
        },
        {
          "epoch": 227,
          "updates": 227,
          "samples_seen": 232448,
          "online_loss": 0.015951069071888924,
          "train_loss": 0.01583625003695488,
          "validation_loss": 0.012903437949717045,
          "train_metric": 0.0015949688386172056,
          "validation_metric": 0.0012995870783925056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5245743470004527
        },
        {
          "epoch": 228,
          "updates": 228,
          "samples_seen": 233472,
          "online_loss": 0.01583624817430973,
          "train_loss": 0.01572285033762455,
          "validation_loss": 0.012813728302717209,
          "train_metric": 0.0015835477970540524,
          "validation_metric": 0.0012905518524348736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.530703476999406
        },
        {
          "epoch": 229,
          "updates": 229,
          "samples_seen": 234496,
          "online_loss": 0.01572285033762455,
          "train_loss": 0.015610852278769016,
          "validation_loss": 0.012725596316158772,
          "train_metric": 0.0015722678508609533,
          "validation_metric": 0.0012816755333915353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.536593594999431
        },
        {
          "epoch": 230,
          "updates": 230,
          "samples_seen": 235520,
          "online_loss": 0.015610852278769016,
          "train_loss": 0.015500236302614212,
          "validation_loss": 0.012638724409043789,
          "train_metric": 0.0015611270209774375,
          "validation_metric": 0.0012729258742183447,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5428956560008373
        },
        {
          "epoch": 231,
          "updates": 231,
          "samples_seen": 236544,
          "online_loss": 0.015500234439969063,
          "train_loss": 0.015390983782708645,
          "validation_loss": 0.01255321130156517,
          "train_metric": 0.0015501236775889993,
          "validation_metric": 0.001264313468709588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.548766144000183
        },
        {
          "epoch": 232,
          "updates": 232,
          "samples_seen": 237568,
          "online_loss": 0.015390986576676369,
          "train_loss": 0.015283076092600822,
          "validation_loss": 0.012469255365431309,
          "train_metric": 0.001539255608804524,
          "validation_metric": 0.0012558576418086886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.554263866000838
        },
        {
          "epoch": 233,
          "updates": 233,
          "samples_seen": 238592,
          "online_loss": 0.015283077023923397,
          "train_loss": 0.015176510438323021,
          "validation_loss": 0.012386498041450977,
          "train_metric": 0.001528522465378046,
          "validation_metric": 0.0012475226540118456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.560812744000941
        },
        {
          "epoch": 234,
          "updates": 234,
          "samples_seen": 239616,
          "online_loss": 0.015176510438323021,
          "train_loss": 0.015071277506649494,
          "validation_loss": 0.012304961681365967,
          "train_metric": 0.0015179237816482782,
          "validation_metric": 0.0012393107172101736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.566423673000827
        },
        {
          "epoch": 235,
          "updates": 235,
          "samples_seen": 240640,
          "online_loss": 0.015071275644004345,
          "train_loss": 0.014967373572289944,
          "validation_loss": 0.012224791571497917,
          "train_metric": 0.0015074589755386114,
          "validation_metric": 0.001231236383318901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5723414740004955
        },
        {
          "epoch": 236,
          "updates": 236,
          "samples_seen": 241664,
          "online_loss": 0.014967373572289944,
          "train_loss": 0.014864782802760601,
          "validation_loss": 0.01214572973549366,
          "train_metric": 0.00149712641723454,
          "validation_metric": 0.001223273342475295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.578653343000042
        },
        {
          "epoch": 237,
          "updates": 237,
          "samples_seen": 242688,
          "online_loss": 0.014864781871438026,
          "train_loss": 0.01476352196186781,
          "validation_loss": 0.01206795871257782,
          "train_metric": 0.0014869278529658914,
          "validation_metric": 0.0012154405703768134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.584384843999942
        },
        {
          "epoch": 238,
          "updates": 238,
          "samples_seen": 243712,
          "online_loss": 0.014763521030545235,
          "train_loss": 0.014663580805063248,
          "validation_loss": 0.011991524137556553,
          "train_metric": 0.0014768620021641254,
          "validation_metric": 0.0012077424908056855,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.590335278000566
        },
        {
          "epoch": 239,
          "updates": 239,
          "samples_seen": 244736,
          "online_loss": 0.014663581736385822,
          "train_loss": 0.014564956538379192,
          "validation_loss": 0.011916174553334713,
          "train_metric": 0.001466928981244564,
          "validation_metric": 0.0012001533759757876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.596633003000534
        },
        {
          "epoch": 240,
          "updates": 240,
          "samples_seen": 245760,
          "online_loss": 0.014564955607056618,
          "train_loss": 0.01446764450520277,
          "validation_loss": 0.011842071078717709,
          "train_metric": 0.00145712832454592,
          "validation_metric": 0.001192689873278141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6023804300002666
        },
        {
          "epoch": 241,
          "updates": 241,
          "samples_seen": 246784,
          "online_loss": 0.01446764450520277,
          "train_loss": 0.014371640980243683,
          "validation_loss": 0.011769124306738377,
          "train_metric": 0.0014474589843302965,
          "validation_metric": 0.0011853432515636086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.608509428999241
        },
        {
          "epoch": 242,
          "updates": 242,
          "samples_seen": 247808,
          "online_loss": 0.014371640048921108,
          "train_loss": 0.014276951551437378,
          "validation_loss": 0.011697265319526196,
          "train_metric": 0.001437922241166234,
          "validation_metric": 0.0011781057110056281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6143357750006544
        },
        {
          "epoch": 243,
          "updates": 243,
          "samples_seen": 248832,
          "online_loss": 0.014276951551437378,
          "train_loss": 0.014183557592332363,
          "validation_loss": 0.011626663617789745,
          "train_metric": 0.0014285158831626177,
          "validation_metric": 0.001170995063148439,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6999056660006318
        },
        {
          "epoch": 244,
          "updates": 244,
          "samples_seen": 249856,
          "online_loss": 0.014183559454977512,
          "train_loss": 0.01409146562218666,
          "validation_loss": 0.011557161808013916,
          "train_metric": 0.0014192407252267003,
          "validation_metric": 0.0011639950098469853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7058401230005984
        },
        {
          "epoch": 245,
          "updates": 245,
          "samples_seen": 250880,
          "online_loss": 0.01409146562218666,
          "train_loss": 0.01400065328925848,
          "validation_loss": 0.011488780379295349,
          "train_metric": 0.0014100947882980108,
          "validation_metric": 0.001157107762992382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.712124739000501
        },
        {
          "epoch": 246,
          "updates": 246,
          "samples_seen": 251904,
          "online_loss": 0.01400065515190363,
          "train_loss": 0.013911128044128418,
          "validation_loss": 0.011421524919569492,
          "train_metric": 0.0014010778395459056,
          "validation_metric": 0.0011503342539072037,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.717959793000773
        },
        {
          "epoch": 247,
          "updates": 247,
          "samples_seen": 252928,
          "online_loss": 0.013911128975450993,
          "train_loss": 0.013822861947119236,
          "validation_loss": 0.01135528925806284,
          "train_metric": 0.0013921877834945917,
          "validation_metric": 0.001143663190305233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7235474930002965
        },
        {
          "epoch": 248,
          "updates": 248,
          "samples_seen": 253952,
          "online_loss": 0.013822861947119236,
          "train_loss": 0.013735854998230934,
          "validation_loss": 0.01129014790058136,
          "train_metric": 0.0013834249693900347,
          "validation_metric": 0.0011371024884283543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.729792341000575
        },
        {
          "epoch": 249,
          "updates": 249,
          "samples_seen": 254976,
          "online_loss": 0.01373585406690836,
          "train_loss": 0.01365008857101202,
          "validation_loss": 0.011226068250834942,
          "train_metric": 0.0013747869525104761,
          "validation_metric": 0.0011306486558169127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7357357170003525
        },
        {
          "epoch": 250,
          "updates": 250,
          "samples_seen": 256000,
          "online_loss": 0.01365008857101202,
          "train_loss": 0.013565543107688427,
          "validation_loss": 0.011162974871695042,
          "train_metric": 0.0013662718702107668,
          "validation_metric": 0.0011242938926443458,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7416493090004224
        },
        {
          "epoch": 251,
          "updates": 251,
          "samples_seen": 257024,
          "online_loss": 0.013565544039011002,
          "train_loss": 0.013482214882969856,
          "validation_loss": 0.011100923642516136,
          "train_metric": 0.0013578791404142976,
          "validation_metric": 0.0011180444853380322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7475328740001714
        },
        {
          "epoch": 252,
          "updates": 252,
          "samples_seen": 258048,
          "online_loss": 0.013482213951647282,
          "train_loss": 0.013400077819824219,
          "validation_loss": 0.011039791628718376,
          "train_metric": 0.0013496067840605974,
          "validation_metric": 0.0011118873953819275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7532254559991998
        },
        {
          "epoch": 253,
          "updates": 253,
          "samples_seen": 259072,
          "online_loss": 0.013400077819824219,
          "train_loss": 0.013319115154445171,
          "validation_loss": 0.01097963098436594,
          "train_metric": 0.001341452356427908,
          "validation_metric": 0.001105828327126801,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7592163640001672
        },
        {
          "epoch": 254,
          "updates": 254,
          "samples_seen": 260096,
          "online_loss": 0.013319114223122597,
          "train_loss": 0.013239311054348946,
          "validation_loss": 0.010920384898781776,
          "train_metric": 0.0013334149261936545,
          "validation_metric": 0.0010998613433912396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7653676429999905
        },
        {
          "epoch": 255,
          "updates": 255,
          "samples_seen": 261120,
          "online_loss": 0.013239310123026371,
          "train_loss": 0.013160645961761475,
          "validation_loss": 0.010862023569643497,
          "train_metric": 0.001325491932220757,
          "validation_metric": 0.0010939831845462322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.771357595000154
        },
        {
          "epoch": 256,
          "updates": 256,
          "samples_seen": 262144,
          "online_loss": 0.0131606450304389,
          "train_loss": 0.013083091005682945,
          "validation_loss": 0.010804543271660805,
          "train_metric": 0.0013176811626181006,
          "validation_metric": 0.0010881940834224224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7776276059994416
        },
        {
          "epoch": 257,
          "updates": 257,
          "samples_seen": 263168,
          "online_loss": 0.013083091005682945,
          "train_loss": 0.013006645254790783,
          "validation_loss": 0.010747889056801796,
          "train_metric": 0.001309981569647789,
          "validation_metric": 0.001082488102838397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.783490042000267
        },
        {
          "epoch": 258,
          "updates": 258,
          "samples_seen": 264192,
          "online_loss": 0.013006644323468208,
          "train_loss": 0.01293126866221428,
          "validation_loss": 0.010692053474485874,
          "train_metric": 0.0013023900100961328,
          "validation_metric": 0.0010768645443022251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.789093376000892
        },
        {
          "epoch": 259,
          "updates": 259,
          "samples_seen": 265216,
          "online_loss": 0.012931269593536854,
          "train_loss": 0.01285694632679224,
          "validation_loss": 0.010637006722390652,
          "train_metric": 0.0012949046213179827,
          "validation_metric": 0.0010713203810155392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.795376299000054
        },
        {
          "epoch": 260,
          "updates": 260,
          "samples_seen": 266240,
          "online_loss": 0.012856947258114815,
          "train_loss": 0.012783662416040897,
          "validation_loss": 0.010582711547613144,
          "train_metric": 0.0012875237734988332,
          "validation_metric": 0.0010658518876880407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8011542090007424
        },
        {
          "epoch": 261,
          "updates": 261,
          "samples_seen": 267264,
          "online_loss": 0.012783662416040897,
          "train_loss": 0.01271138247102499,
          "validation_loss": 0.010529173538088799,
          "train_metric": 0.0012802437413483858,
          "validation_metric": 0.001060459646396339,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8069864509998297
        },
        {
          "epoch": 262,
          "updates": 262,
          "samples_seen": 268288,
          "online_loss": 0.01271138060837984,
          "train_loss": 0.012640092521905899,
          "validation_loss": 0.01047633495181799,
          "train_metric": 0.0012730640592053533,
          "validation_metric": 0.0010551379527896643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8129051809992234
        },
        {
          "epoch": 263,
          "updates": 263,
          "samples_seen": 269312,
          "online_loss": 0.012640093453228474,
          "train_loss": 0.01256977953016758,
          "validation_loss": 0.010424201376736164,
          "train_metric": 0.0012659822823479772,
          "validation_metric": 0.0010498873889446259,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.818517260999215
        },
        {
          "epoch": 264,
          "updates": 264,
          "samples_seen": 270336,
          "online_loss": 0.012569780461490154,
          "train_loss": 0.012500399723649025,
          "validation_loss": 0.01037272997200489,
          "train_metric": 0.0012589945690706372,
          "validation_metric": 0.0010447032982483506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.824370620999616
        },
        {
          "epoch": 265,
          "updates": 265,
          "samples_seen": 271360,
          "online_loss": 0.012500401586294174,
          "train_loss": 0.012431953102350235,
          "validation_loss": 0.010321916081011295,
          "train_metric": 0.0012521008029580116,
          "validation_metric": 0.0010395855642855167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8301876700006687
        },
        {
          "epoch": 266,
          "updates": 266,
          "samples_seen": 272384,
          "online_loss": 0.012431954964995384,
          "train_loss": 0.012364410795271397,
          "validation_loss": 0.010271732695400715,
          "train_metric": 0.0012452981900423765,
          "validation_metric": 0.0010345311602577567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8359569420008484
        },
        {
          "epoch": 267,
          "updates": 267,
          "samples_seen": 273408,
          "online_loss": 0.012364409863948822,
          "train_loss": 0.012297743931412697,
          "validation_loss": 0.010222158394753933,
          "train_metric": 0.0012385838199406862,
          "validation_metric": 0.001029538456350565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.842102970000269
        },
        {
          "epoch": 268,
          "updates": 268,
          "samples_seen": 274432,
          "online_loss": 0.012297744862735271,
          "train_loss": 0.012231932953000069,
          "validation_loss": 0.01017316710203886,
          "train_metric": 0.0012319553643465042,
          "validation_metric": 0.0010246041929349303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.847886302999541
        },
        {
          "epoch": 269,
          "updates": 269,
          "samples_seen": 275456,
          "online_loss": 0.012231932021677494,
          "train_loss": 0.012166965752840042,
          "validation_loss": 0.0101247513666749,
          "train_metric": 0.0012254121247678995,
          "validation_metric": 0.0010197279043495655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.85354603199994
        },
        {
          "epoch": 270,
          "updates": 270,
          "samples_seen": 276480,
          "online_loss": 0.012166964821517467,
          "train_loss": 0.012102819979190826,
          "validation_loss": 0.010076900944113731,
          "train_metric": 0.0012189517728984356,
          "validation_metric": 0.001014908542856574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.859702723999362
        },
        {
          "epoch": 271,
          "updates": 271,
          "samples_seen": 277504,
          "online_loss": 0.012102819979190826,
          "train_loss": 0.012039457447826862,
          "validation_loss": 0.0100295664742589,
          "train_metric": 0.0012125701177865267,
          "validation_metric": 0.001010141451843083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8656075859998964
        },
        {
          "epoch": 272,
          "updates": 272,
          "samples_seen": 278528,
          "online_loss": 0.012039458379149437,
          "train_loss": 0.011976887471973896,
          "validation_loss": 0.009982755407691002,
          "train_metric": 0.001206268323585391,
          "validation_metric": 0.001005426631309092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8711940850007522
        },
        {
          "epoch": 273,
          "updates": 273,
          "samples_seen": 279552,
          "online_loss": 0.01197688840329647,
          "train_loss": 0.011915059760212898,
          "validation_loss": 0.009936421178281307,
          "train_metric": 0.0012000411516055465,
          "validation_metric": 0.0010007600067183375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8773131420002755
        },
        {
          "epoch": 274,
          "updates": 274,
          "samples_seen": 280576,
          "online_loss": 0.011915058828890324,
          "train_loss": 0.011853977106511593,
          "validation_loss": 0.009890591725707054,
          "train_metric": 0.001193889300338924,
          "validation_metric": 0.0009961441392078996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8829629449992353
        },
        {
          "epoch": 275,
          "updates": 275,
          "samples_seen": 281600,
          "online_loss": 0.011853978037834167,
          "train_loss": 0.011793596670031548,
          "validation_loss": 0.009845202788710594,
          "train_metric": 0.0011878079967573285,
          "validation_metric": 0.0009915727423503995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.888520726000934
        },
        {
          "epoch": 276,
          "updates": 276,
          "samples_seen": 282624,
          "online_loss": 0.011793597601354122,
          "train_loss": 0.011733927763998508,
          "validation_loss": 0.009800274856388569,
          "train_metric": 0.0011817981721833348,
          "validation_metric": 0.0009870477952063084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8944950889999745
        },
        {
          "epoch": 277,
          "updates": 277,
          "samples_seen": 283648,
          "online_loss": 0.011733926832675934,
          "train_loss": 0.011674927547574043,
          "validation_loss": 0.009755740873515606,
          "train_metric": 0.0011758559849113226,
          "validation_metric": 0.000982562662102282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.900388905
        },
        {
          "epoch": 278,
          "updates": 278,
          "samples_seen": 284672,
          "online_loss": 0.011674928478896618,
          "train_loss": 0.011616590432822704,
          "validation_loss": 0.00971164833754301,
          "train_metric": 0.0011699805036187172,
          "validation_metric": 0.0009781216504052281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.90616014599982
        },
        {
          "epoch": 279,
          "updates": 279,
          "samples_seen": 285696,
          "online_loss": 0.011616591364145279,
          "train_loss": 0.011558886617422104,
          "validation_loss": 0.009667918086051941,
          "train_metric": 0.001164168817922473,
          "validation_metric": 0.0009737172513268888,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9121577620007884
        },
        {
          "epoch": 280,
          "updates": 280,
          "samples_seen": 286720,
          "online_loss": 0.011558886617422104,
          "train_loss": 0.011501805856823921,
          "validation_loss": 0.009624623693525791,
          "train_metric": 0.0011584198800846934,
          "validation_metric": 0.0009693568572402,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.917866359999607
        },
        {
          "epoch": 281,
          "updates": 281,
          "samples_seen": 287744,
          "online_loss": 0.011501805856823921,
          "train_loss": 0.01144532673060894,
          "validation_loss": 0.009581646881997585,
          "train_metric": 0.001152731478214264,
          "validation_metric": 0.0009650283609516919,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9237218289999873
        },
        {
          "epoch": 282,
          "updates": 282,
          "samples_seen": 288768,
          "online_loss": 0.01144532673060894,
          "train_loss": 0.01138943899422884,
          "validation_loss": 0.009539092890918255,
          "train_metric": 0.001147102564573288,
          "validation_metric": 0.0009607425308786333,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9295814050001354
        },
        {
          "epoch": 283,
          "updates": 283,
          "samples_seen": 289792,
          "online_loss": 0.01138943899422884,
          "train_loss": 0.0113341324031353,
          "validation_loss": 0.009496835991740227,
          "train_metric": 0.0011415320914238691,
          "validation_metric": 0.0009564865613356233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.016938840999501
        },
        {
          "epoch": 284,
          "updates": 284,
          "samples_seen": 290816,
          "online_loss": 0.011334131471812725,
          "train_loss": 0.011279396712779999,
          "validation_loss": 0.0094550009816885,
          "train_metric": 0.0011360195931047201,
          "validation_metric": 0.0009522730833850801,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0232104700007767
        },
        {
          "epoch": 285,
          "updates": 285,
          "samples_seen": 291840,
          "online_loss": 0.011279396712779999,
          "train_loss": 0.011225294321775436,
          "validation_loss": 0.009413606487214565,
          "train_metric": 0.0011305705411359668,
          "validation_metric": 0.000948103959672153,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0295848030000343
        },
        {
          "epoch": 286,
          "updates": 286,
          "samples_seen": 292864,
          "online_loss": 0.011225294321775436,
          "train_loss": 0.011171985417604446,
          "validation_loss": 0.00937261525541544,
          "train_metric": 0.0011252014664933085,
          "validation_metric": 0.0009439754649065435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.035017410000364
        },
        {
          "epoch": 287,
          "updates": 287,
          "samples_seen": 293888,
          "online_loss": 0.011171986348927021,
          "train_loss": 0.01112006139010191,
          "validation_loss": 0.009333410300314426,
          "train_metric": 0.0011199719738215208,
          "validation_metric": 0.0009400270646438003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.040707804999329
        },
        {
          "epoch": 288,
          "updates": 288,
          "samples_seen": 294912,
          "online_loss": 0.01112006139010191,
          "train_loss": 0.011071468703448772,
          "validation_loss": 0.009295758791267872,
          "train_metric": 0.0011150777572765946,
          "validation_metric": 0.000936234719119966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0466642009996576
        },
        {
          "epoch": 289,
          "updates": 289,
          "samples_seen": 295936,
          "online_loss": 0.011071467772126198,
          "train_loss": 0.011032209731638432,
          "validation_loss": 0.009270659647881985,
          "train_metric": 0.001111123594455421,
          "validation_metric": 0.000933706876821816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0524178399991797
        },
        {
          "epoch": 290,
          "updates": 290,
          "samples_seen": 296960,
          "online_loss": 0.011032208800315857,
          "train_loss": 0.01101807039231062,
          "validation_loss": 0.00926258135586977,
          "train_metric": 0.0011096998350694776,
          "validation_metric": 0.0009328933083452284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0582811799995397
        },
        {
          "epoch": 291,
          "updates": 291,
          "samples_seen": 297984,
          "online_loss": 0.01101807039231062,
          "train_loss": 0.01105087623000145,
          "validation_loss": 0.009315301664173603,
          "train_metric": 0.0011130038183182478,
          "validation_metric": 0.000938202952966094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.063616679999541
        },
        {
          "epoch": 292,
          "updates": 292,
          "samples_seen": 299008,
          "online_loss": 0.011050877161324024,
          "train_loss": 0.011097170412540436,
          "validation_loss": 0.009355859830975533,
          "train_metric": 0.0011176662519574165,
          "validation_metric": 0.0009422879666090012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0689354869991803
        },
        {
          "epoch": 293,
          "updates": 293,
          "samples_seen": 300032,
          "online_loss": 0.011097169481217861,
          "train_loss": 0.011017989367246628,
          "validation_loss": 0.009306702762842178,
          "train_metric": 0.0011096915695816278,
          "validation_metric": 0.0009373371140100062,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.073939538999184
        },
        {
          "epoch": 294,
          "updates": 294,
          "samples_seen": 301056,
          "online_loss": 0.011017989367246628,
          "train_loss": 0.01081245206296444,
          "validation_loss": 0.009098373353481293,
          "train_metric": 0.0010889905970543623,
          "validation_metric": 0.0009163548820652068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0797321610007202
        },
        {
          "epoch": 295,
          "updates": 295,
          "samples_seen": 302080,
          "online_loss": 0.010812451131641865,
          "train_loss": 0.010736918076872826,
          "validation_loss": 0.009035415947437286,
          "train_metric": 0.0010813830886036158,
          "validation_metric": 0.0009100139723159373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0855920939993666
        },
        {
          "epoch": 296,
          "updates": 296,
          "samples_seen": 303104,
          "online_loss": 0.010736917145550251,
          "train_loss": 0.01080072671175003,
          "validation_loss": 0.009120979346334934,
          "train_metric": 0.0010878096800297499,
          "validation_metric": 0.0009186315583065152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0909463280004275
        },
        {
          "epoch": 297,
          "updates": 297,
          "samples_seen": 304128,
          "online_loss": 0.010800727643072605,
          "train_loss": 0.010751890018582344,
          "validation_loss": 0.009060953743755817,
          "train_metric": 0.0010828911326825619,
          "validation_metric": 0.0009125859942287207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.096103096000661
        },
        {
          "epoch": 298,
          "updates": 298,
          "samples_seen": 305152,
          "online_loss": 0.010751890949904919,
          "train_loss": 0.010601373389363289,
          "validation_loss": 0.008935630321502686,
          "train_metric": 0.001067731762304902,
          "validation_metric": 0.0008999639539979398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.101787379000598
        },
        {
          "epoch": 299,
          "updates": 299,
          "samples_seen": 306176,
          "online_loss": 0.010601373389363289,
          "train_loss": 0.010574838146567345,
          "validation_loss": 0.008920644409954548,
          "train_metric": 0.0010650589829310775,
          "validation_metric": 0.0008984546293504536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.107674037999459
        },
        {
          "epoch": 300,
          "updates": 300,
          "samples_seen": 307200,
          "online_loss": 0.010574838146567345,
          "train_loss": 0.010599175468087196,
          "validation_loss": 0.008936634287238121,
          "train_metric": 0.0010675102239474654,
          "validation_metric": 0.0009000651771202683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.113333285000408
        },
        {
          "epoch": 301,
          "updates": 301,
          "samples_seen": 308224,
          "online_loss": 0.010599175468087196,
          "train_loss": 0.010505600832402706,
          "validation_loss": 0.008872766979038715,
          "train_metric": 0.0010580855887383223,
          "validation_metric": 0.0008936325903050601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.119091295000544
        },
        {
          "epoch": 302,
          "updates": 302,
          "samples_seen": 309248,
          "online_loss": 0.010505600832402706,
          "train_loss": 0.010417626239359379,
          "validation_loss": 0.00878792442381382,
          "train_metric": 0.0010492251021787524,
          "validation_metric": 0.0008850875310599804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.125218429000597
        },
        {
          "epoch": 303,
          "updates": 303,
          "samples_seen": 310272,
          "online_loss": 0.010417627170681953,
          "train_loss": 0.01042686216533184,
          "validation_loss": 0.008795334957540035,
          "train_metric": 0.0010501553770154715,
          "validation_metric": 0.0008858339278958738,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1303352110007836
        },
        {
          "epoch": 304,
          "updates": 304,
          "samples_seen": 311296,
          "online_loss": 0.010426861234009266,
          "train_loss": 0.010390405543148518,
          "validation_loss": 0.008787287399172783,
          "train_metric": 0.001046483521349728,
          "validation_metric": 0.0008850234444253147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.136033688999305
        },
        {
          "epoch": 305,
          "updates": 305,
          "samples_seen": 312320,
          "online_loss": 0.010390405543148518,
          "train_loss": 0.010296872816979885,
          "validation_loss": 0.008691019378602505,
          "train_metric": 0.0010370634263381362,
          "validation_metric": 0.0008753277943469584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.141988178000247
        },
        {
          "epoch": 306,
          "updates": 306,
          "samples_seen": 313344,
          "online_loss": 0.01029687374830246,
          "train_loss": 0.010269378311932087,
          "validation_loss": 0.008669773116707802,
          "train_metric": 0.001034294138662517,
          "validation_metric": 0.0008731878479011357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1480987960003404
        },
        {
          "epoch": 307,
          "updates": 307,
          "samples_seen": 314368,
          "online_loss": 0.010269378311932087,
          "train_loss": 0.010258140042424202,
          "validation_loss": 0.008683186955749989,
          "train_metric": 0.0010331624653190374,
          "validation_metric": 0.0008745387895032763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.153401376999682
        },
        {
          "epoch": 308,
          "updates": 308,
          "samples_seen": 315392,
          "online_loss": 0.010258140973746777,
          "train_loss": 0.01018638163805008,
          "validation_loss": 0.008604709059000015,
          "train_metric": 0.0010259350528940558,
          "validation_metric": 0.0008666347712278366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1594653030006157
        },
        {
          "epoch": 309,
          "updates": 309,
          "samples_seen": 316416,
          "online_loss": 0.01018638163805008,
          "train_loss": 0.0101321991533041,
          "validation_loss": 0.00856311060488224,
          "train_metric": 0.0010204779682680964,
          "validation_metric": 0.000862445100210607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.165544444000261
        },
        {
          "epoch": 310,
          "updates": 310,
          "samples_seen": 317440,
          "online_loss": 0.0101321991533041,
          "train_loss": 0.010119913145899773,
          "validation_loss": 0.008570505306124687,
          "train_metric": 0.001019240589812398,
          "validation_metric": 0.0008631899254396558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1708008629993856
        },
        {
          "epoch": 311,
          "updates": 311,
          "samples_seen": 318464,
          "online_loss": 0.010119913145899773,
          "train_loss": 0.010073830373585224,
          "validation_loss": 0.008516929112374783,
          "train_metric": 0.00101459922734648,
          "validation_metric": 0.0008577939006499946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.176921071999459
        },
        {
          "epoch": 312,
          "updates": 312,
          "samples_seen": 319488,
          "online_loss": 0.010073830373585224,
          "train_loss": 0.010010900907218456,
          "validation_loss": 0.008472004905343056,
          "train_metric": 0.001008261344395578,
          "validation_metric": 0.0008532694191671908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1827199529998325
        },
        {
          "epoch": 313,
          "updates": 313,
          "samples_seen": 320512,
          "online_loss": 0.010010900907218456,
          "train_loss": 0.009984558448195457,
          "validation_loss": 0.008459164761006832,
          "train_metric": 0.001005608239211142,
          "validation_metric": 0.0008519761613570154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.188535237999531
        },
        {
          "epoch": 314,
          "updates": 314,
          "samples_seen": 321536,
          "online_loss": 0.009984559379518032,
          "train_loss": 0.00995534285902977,
          "validation_loss": 0.008423496969044209,
          "train_metric": 0.0010026657255366445,
          "validation_metric": 0.0008483837591484189,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.195020420000219
        },
        {
          "epoch": 315,
          "updates": 315,
          "samples_seen": 322560,
          "online_loss": 0.009955343790352345,
          "train_loss": 0.009898434393107891,
          "validation_loss": 0.008388032205402851,
          "train_metric": 0.0009969340171664953,
          "validation_metric": 0.0008448117878288031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2008846330008964
        },
        {
          "epoch": 316,
          "updates": 316,
          "samples_seen": 323584,
          "online_loss": 0.009898433461785316,
          "train_loss": 0.009857302531599998,
          "validation_loss": 0.008354474790394306,
          "train_metric": 0.0009927913779392838,
          "validation_metric": 0.0008414320182055235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.206962095000563
        },
        {
          "epoch": 317,
          "updates": 317,
          "samples_seen": 324608,
          "online_loss": 0.009857302531599998,
          "train_loss": 0.009832232259213924,
          "validation_loss": 0.008326785638928413,
          "train_metric": 0.0009902663296088576,
          "validation_metric": 0.0008386434055864811,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.213132303999373
        },
        {
          "epoch": 318,
          "updates": 318,
          "samples_seen": 325632,
          "online_loss": 0.009832233190536499,
          "train_loss": 0.00978793017566204,
          "validation_loss": 0.008304623886942863,
          "train_metric": 0.0009858044795691967,
          "validation_metric": 0.0008364113746210933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.219075598000927
        },
        {
          "epoch": 319,
          "updates": 319,
          "samples_seen": 326656,
          "online_loss": 0.00978793017566204,
          "train_loss": 0.009739385917782784,
          "validation_loss": 0.008257566951215267,
          "train_metric": 0.000980915268883109,
          "validation_metric": 0.0008316718158312142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.225018863000514
        },
        {
          "epoch": 320,
          "updates": 320,
          "samples_seen": 327680,
          "online_loss": 0.009739386849105358,
          "train_loss": 0.009708731435239315,
          "validation_loss": 0.008229514583945274,
          "train_metric": 0.0009778279345482588,
          "validation_metric": 0.0008288464741781354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.315316193000399
        },
        {
          "epoch": 321,
          "updates": 321,
          "samples_seen": 328704,
          "online_loss": 0.009708731435239315,
          "train_loss": 0.009674850851297379,
          "validation_loss": 0.008216078393161297,
          "train_metric": 0.0009744155104272068,
          "validation_metric": 0.0008274932624772191,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3214874589993997
        },
        {
          "epoch": 322,
          "updates": 322,
          "samples_seen": 329728,
          "online_loss": 0.00967484898865223,
          "train_loss": 0.009628215804696083,
          "validation_loss": 0.008166844956576824,
          "train_metric": 0.0009697186178527772,
          "validation_metric": 0.0008225347264669836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.331310638999639
        },
        {
          "epoch": 323,
          "updates": 323,
          "samples_seen": 330752,
          "online_loss": 0.009628215804696083,
          "train_loss": 0.009589431807398796,
          "validation_loss": 0.008136371150612831,
          "train_metric": 0.000965812592767179,
          "validation_metric": 0.0008194655529223382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.337855664000017
        },
        {
          "epoch": 324,
          "updates": 324,
          "samples_seen": 331776,
          "online_loss": 0.009589433670043945,
          "train_loss": 0.009558582678437233,
          "validation_loss": 0.008121336810290813,
          "train_metric": 0.0009627054096199572,
          "validation_metric": 0.0008179512806236744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3472304360002454
        },
        {
          "epoch": 325,
          "updates": 325,
          "samples_seen": 332800,
          "online_loss": 0.009558582678437233,
          "train_loss": 0.009519043378531933,
          "validation_loss": 0.008077571168541908,
          "train_metric": 0.0009587232489138842,
          "validation_metric": 0.0008135434472933412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3532397100007074
        },
        {
          "epoch": 326,
          "updates": 326,
          "samples_seen": 333824,
          "online_loss": 0.009519042447209358,
          "train_loss": 0.009476329199969769,
          "validation_loss": 0.008048299700021744,
          "train_metric": 0.0009544212371110916,
          "validation_metric": 0.000810595287475735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.363323720999688
        },
        {
          "epoch": 327,
          "updates": 327,
          "samples_seen": 334848,
          "online_loss": 0.009476329199969769,
          "train_loss": 0.00944205280393362,
          "validation_loss": 0.008024157024919987,
          "train_metric": 0.0009509689407423139,
          "validation_metric": 0.0008081637788563967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3692530949992943
        },
        {
          "epoch": 328,
          "updates": 328,
          "samples_seen": 335872,
          "online_loss": 0.009442051872611046,
          "train_loss": 0.00940813310444355,
          "validation_loss": 0.007987575605511665,
          "train_metric": 0.0009475527331233025,
          "validation_metric": 0.000804479350335896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.377322128999367
        },
        {
          "epoch": 329,
          "updates": 329,
          "samples_seen": 336896,
          "online_loss": 0.0094081349670887,
          "train_loss": 0.009367621503770351,
          "validation_loss": 0.007963025011122227,
          "train_metric": 0.0009434724925085902,
          "validation_metric": 0.0008020067471079528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3833988240003237
        },
        {
          "epoch": 330,
          "updates": 330,
          "samples_seen": 337920,
          "online_loss": 0.009367620572447777,
          "train_loss": 0.009328952990472317,
          "validation_loss": 0.007928638719022274,
          "train_metric": 0.0009395779343321919,
          "validation_metric": 0.0007985435076989233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3891468759993586
        },
        {
          "epoch": 331,
          "updates": 331,
          "samples_seen": 338944,
          "online_loss": 0.009328952059149742,
          "train_loss": 0.009295446798205376,
          "validation_loss": 0.007896795868873596,
          "train_metric": 0.0009362033451907337,
          "validation_metric": 0.0007953362073749304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.398909088000437
        },
        {
          "epoch": 332,
          "updates": 332,
          "samples_seen": 339968,
          "online_loss": 0.00929544772952795,
          "train_loss": 0.009259497746825218,
          "validation_loss": 0.007876589894294739,
          "train_metric": 0.0009325827122665942,
          "validation_metric": 0.0007933011511340737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.405134132000967
        },
        {
          "epoch": 333,
          "updates": 333,
          "samples_seen": 340992,
          "online_loss": 0.009259498678147793,
          "train_loss": 0.009220229461789131,
          "validation_loss": 0.007837343029677868,
          "train_metric": 0.0009286276763305068,
          "validation_metric": 0.000789348385296762,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.413941707000049
        },
        {
          "epoch": 334,
          "updates": 334,
          "samples_seen": 342016,
          "online_loss": 0.009220229461789131,
          "train_loss": 0.009183898568153381,
          "validation_loss": 0.00780781963840127,
          "train_metric": 0.0009249686845578253,
          "validation_metric": 0.0007863750797696412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.419864023999253
        },
        {
          "epoch": 335,
          "updates": 335,
          "samples_seen": 343040,
          "online_loss": 0.009183898568153381,
          "train_loss": 0.009149959310889244,
          "validation_loss": 0.007786274421960115,
          "train_metric": 0.0009215503232553601,
          "validation_metric": 0.0007842050981707871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4303028300000733
        },
        {
          "epoch": 336,
          "updates": 336,
          "samples_seen": 344064,
          "online_loss": 0.009149959310889244,
          "train_loss": 0.00911349430680275,
          "validation_loss": 0.0077485935762524605,
          "train_metric": 0.0009178777690976858,
          "validation_metric": 0.000780409958679229,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.436588717999257
        },
        {
          "epoch": 337,
          "updates": 337,
          "samples_seen": 345088,
          "online_loss": 0.00911349430680275,
          "train_loss": 0.009075745940208435,
          "validation_loss": 0.0077221146784722805,
          "train_metric": 0.0009140759357251227,
          "validation_metric": 0.0007777430582791567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4429091420006444
        },
        {
          "epoch": 338,
          "updates": 338,
          "samples_seen": 346112,
          "online_loss": 0.00907574687153101,
          "train_loss": 0.00904047954827547,
          "validation_loss": 0.007694223430007696,
          "train_metric": 0.0009105239296332002,
          "validation_metric": 0.0007749339565634727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4486795299999358
        },
        {
          "epoch": 339,
          "updates": 339,
          "samples_seen": 347136,
          "online_loss": 0.00904047954827547,
          "train_loss": 0.009006165899336338,
          "validation_loss": 0.007660509552806616,
          "train_metric": 0.0009070679661817849,
          "validation_metric": 0.0007715384708717465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4542822610001167
        },
        {
          "epoch": 340,
          "updates": 340,
          "samples_seen": 348160,
          "online_loss": 0.009006165899336338,
          "train_loss": 0.008969972841441631,
          "validation_loss": 0.007637253031134605,
          "train_metric": 0.0009034228278324008,
          "validation_metric": 0.0007691961945965886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.460561948000759
        },
        {
          "epoch": 341,
          "updates": 341,
          "samples_seen": 349184,
          "online_loss": 0.008969972841441631,
          "train_loss": 0.008933423087000847,
          "validation_loss": 0.007603557780385017,
          "train_metric": 0.0008997416007332504,
          "validation_metric": 0.0007658025133423507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4665335890003917
        },
        {
          "epoch": 342,
          "updates": 342,
          "samples_seen": 350208,
          "online_loss": 0.008933422155678272,
          "train_loss": 0.008898633532226086,
          "validation_loss": 0.007573622278869152,
          "train_metric": 0.0008962376741692424,
          "validation_metric": 0.0007627874729223549,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.472470794000401
        },
        {
          "epoch": 343,
          "updates": 343,
          "samples_seen": 351232,
          "online_loss": 0.008898632600903511,
          "train_loss": 0.008864246308803558,
          "validation_loss": 0.007550555747002363,
          "train_metric": 0.0008927744929678738,
          "validation_metric": 0.0007604642887599766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4788281740002276
        },
        {
          "epoch": 344,
          "updates": 344,
          "samples_seen": 352256,
          "online_loss": 0.008864246308803558,
          "train_loss": 0.008828646503388882,
          "validation_loss": 0.007515530101954937,
          "train_metric": 0.0008891889592632651,
          "validation_metric": 0.000756936555262655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4844059829993057
        },
        {
          "epoch": 345,
          "updates": 345,
          "samples_seen": 353280,
          "online_loss": 0.008828646503388882,
          "train_loss": 0.008792945183813572,
          "validation_loss": 0.0074888537637889385,
          "train_metric": 0.0008855932974256575,
          "validation_metric": 0.000754249922465533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.490218024000569
        },
        {
          "epoch": 346,
          "updates": 346,
          "samples_seen": 354304,
          "online_loss": 0.008792944252490997,
          "train_loss": 0.008758436888456345,
          "validation_loss": 0.007461829110980034,
          "train_metric": 0.0008821177762001753,
          "validation_metric": 0.0007515280740335584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4964683550006157
        },
        {
          "epoch": 347,
          "updates": 347,
          "samples_seen": 355328,
          "online_loss": 0.008758436888456345,
          "train_loss": 0.008724218234419823,
          "validation_loss": 0.007429366931319237,
          "train_metric": 0.000878671242389828,
          "validation_metric": 0.0007482586661353707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.502307318000021
        },
        {
          "epoch": 348,
          "updates": 348,
          "samples_seen": 356352,
          "online_loss": 0.008724217303097248,
          "train_loss": 0.008689246140420437,
          "validation_loss": 0.00740527268499136,
          "train_metric": 0.0008751490386202931,
          "validation_metric": 0.0007458319305442274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5084212140009186
        },
        {
          "epoch": 349,
          "updates": 349,
          "samples_seen": 357376,
          "online_loss": 0.008689245209097862,
          "train_loss": 0.008654196746647358,
          "validation_loss": 0.0073736729100346565,
          "train_metric": 0.0008716190350241959,
          "validation_metric": 0.0007426493102684617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5150592850004614
        },
        {
          "epoch": 350,
          "updates": 350,
          "samples_seen": 358400,
          "online_loss": 0.008654195815324783,
          "train_loss": 0.008619943633675575,
          "validation_loss": 0.007344632409512997,
          "train_metric": 0.0008681690669618547,
          "validation_metric": 0.0007397244335152209,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5206915109993133
        },
        {
          "epoch": 351,
          "updates": 351,
          "samples_seen": 359424,
          "online_loss": 0.008619942702353,
          "train_loss": 0.00858599878847599,
          "validation_loss": 0.007320243865251541,
          "train_metric": 0.0008647502982057631,
          "validation_metric": 0.0007372681866399944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.526867000000493
        },
        {
          "epoch": 352,
          "updates": 352,
          "samples_seen": 360448,
          "online_loss": 0.008585997857153416,
          "train_loss": 0.008551636710762978,
          "validation_loss": 0.007287680171430111,
          "train_metric": 0.000861289503518492,
          "validation_metric": 0.0007339884177781641,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5325121549994947
        },
        {
          "epoch": 353,
          "updates": 353,
          "samples_seen": 361472,
          "online_loss": 0.008551637642085552,
          "train_loss": 0.008517151698470116,
          "validation_loss": 0.007261710707098246,
          "train_metric": 0.0008578163688071072,
          "validation_metric": 0.0007313729147426784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5383879080000042
        },
        {
          "epoch": 354,
          "updates": 354,
          "samples_seen": 362496,
          "online_loss": 0.008517151698470116,
          "train_loss": 0.008483167737722397,
          "validation_loss": 0.007233730051666498,
          "train_metric": 0.0008543935837224126,
          "validation_metric": 0.000728554732631892,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5447433890003595
        },
        {
          "epoch": 355,
          "updates": 355,
          "samples_seen": 363520,
          "online_loss": 0.008483167737722397,
          "train_loss": 0.008449537679553032,
          "validation_loss": 0.007203337736427784,
          "train_metric": 0.0008510064799338579,
          "validation_metric": 0.0007254937663674355,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.550479854000514
        },
        {
          "epoch": 356,
          "updates": 356,
          "samples_seen": 364544,
          "online_loss": 0.008449537679553032,
          "train_loss": 0.008415733464062214,
          "validation_loss": 0.0071789417415857315,
          "train_metric": 0.0008476019138470292,
          "validation_metric": 0.0007230367045849562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.556060291000904
        },
        {
          "epoch": 357,
          "updates": 357,
          "samples_seen": 365568,
          "online_loss": 0.008415733464062214,
          "train_loss": 0.008381789550185204,
          "validation_loss": 0.0071480232290923595,
          "train_metric": 0.0008441830868832767,
          "validation_metric": 0.0007199227693490684,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5621401250009512
        },
        {
          "epoch": 358,
          "updates": 358,
          "samples_seen": 366592,
          "online_loss": 0.008381789550185204,
          "train_loss": 0.008348115719854832,
          "validation_loss": 0.007120681926608086,
          "train_metric": 0.0008407916175201535,
          "validation_metric": 0.0007171690231189132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.652861357000802
        },
        {
          "epoch": 359,
          "updates": 359,
          "samples_seen": 367616,
          "online_loss": 0.008348115719854832,
          "train_loss": 0.00831479113548994,
          "validation_loss": 0.007094590459018946,
          "train_metric": 0.0008374352473765612,
          "validation_metric": 0.0007145411800593138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6589893929995014
        },
        {
          "epoch": 360,
          "updates": 360,
          "samples_seen": 368640,
          "online_loss": 0.00831479113548994,
          "train_loss": 0.008281494490802288,
          "validation_loss": 0.0070640528574585915,
          "train_metric": 0.0008340817876160145,
          "validation_metric": 0.0007114655454643071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6651972380004736
        },
        {
          "epoch": 361,
          "updates": 361,
          "samples_seen": 369664,
          "online_loss": 0.008281493559479713,
          "train_loss": 0.008248092606663704,
          "validation_loss": 0.007038996554911137,
          "train_metric": 0.0008307177340611815,
          "validation_metric": 0.0007089420105330646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.671265368000604
        },
        {
          "epoch": 362,
          "updates": 362,
          "samples_seen": 370688,
          "online_loss": 0.008248093537986279,
          "train_loss": 0.0082147978246212,
          "validation_loss": 0.007009865250438452,
          "train_metric": 0.0008273643325082958,
          "validation_metric": 0.0007060079951770604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.677171408000504
        },
        {
          "epoch": 363,
          "updates": 363,
          "samples_seen": 371712,
          "online_loss": 0.0082147978246212,
          "train_loss": 0.008181772194802761,
          "validation_loss": 0.006981836166232824,
          "train_metric": 0.0008240381139330566,
          "validation_metric": 0.0007031849818304181,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.682934882999689
        },
        {
          "epoch": 364,
          "updates": 364,
          "samples_seen": 372736,
          "online_loss": 0.008181773126125336,
          "train_loss": 0.008148910477757454,
          "validation_loss": 0.006956398021429777,
          "train_metric": 0.0008207284263335168,
          "validation_metric": 0.0007006230298429728,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6886974989993178
        },
        {
          "epoch": 365,
          "updates": 365,
          "samples_seen": 373760,
          "online_loss": 0.008148910477757454,
          "train_loss": 0.008116043172776699,
          "validation_loss": 0.006926482077687979,
          "train_metric": 0.0008174180402420461,
          "validation_metric": 0.0006976099102757871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.695051130000138
        },
        {
          "epoch": 366,
          "updates": 366,
          "samples_seen": 374784,
          "online_loss": 0.008116043172776699,
          "train_loss": 0.008083184249699116,
          "validation_loss": 0.006900908891111612,
          "train_metric": 0.0008141087018884718,
          "validation_metric": 0.0006950342212803662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7008293149992824
        },
        {
          "epoch": 367,
          "updates": 367,
          "samples_seen": 375808,
          "online_loss": 0.008083184249699116,
          "train_loss": 0.008050497621297836,
          "validation_loss": 0.006873002275824547,
          "train_metric": 0.0008108167094178498,
          "validation_metric": 0.0006922235479578376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7068842659991788
        },
        {
          "epoch": 368,
          "updates": 368,
          "samples_seen": 376832,
          "online_loss": 0.008050497621297836,
          "train_loss": 0.008018006570637226,
          "validation_loss": 0.006844908930361271,
          "train_metric": 0.0008075441583059728,
          "validation_metric": 0.0006893941899761558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7129280019998987
        },
        {
          "epoch": 369,
          "updates": 369,
          "samples_seen": 377856,
          "online_loss": 0.008018005639314651,
          "train_loss": 0.007985618896782398,
          "validation_loss": 0.006819637026637793,
          "train_metric": 0.000804282259196043,
          "validation_metric": 0.0006868488271720707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7186820189999708
        },
        {
          "epoch": 370,
          "updates": 370,
          "samples_seen": 378880,
          "online_loss": 0.007985617965459824,
          "train_loss": 0.007953251712024212,
          "validation_loss": 0.006790517829358578,
          "train_metric": 0.0008010223391465843,
          "validation_metric": 0.0006839160341769457,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.724545066999781
        },
        {
          "epoch": 371,
          "updates": 371,
          "samples_seen": 379904,
          "online_loss": 0.007953251712024212,
          "train_loss": 0.007920962758362293,
          "validation_loss": 0.00676477886736393,
          "train_metric": 0.0007977703935466707,
          "validation_metric": 0.0006813236977905035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7305590729993128
        },
        {
          "epoch": 372,
          "updates": 372,
          "samples_seen": 380928,
          "online_loss": 0.007920962758362293,
          "train_loss": 0.007888827472925186,
          "validation_loss": 0.006737541873008013,
          "train_metric": 0.0007945338729768991,
          "validation_metric": 0.0006785806035622954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7368284219992347
        },
        {
          "epoch": 373,
          "updates": 373,
          "samples_seen": 381952,
          "online_loss": 0.007888827472925186,
          "train_loss": 0.007856844924390316,
          "validation_loss": 0.006709812209010124,
          "train_metric": 0.0007913126610219479,
          "validation_metric": 0.0006757876253686845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.742828284999632
        },
        {
          "epoch": 374,
          "updates": 374,
          "samples_seen": 382976,
          "online_loss": 0.007856844924390316,
          "train_loss": 0.007824954576790333,
          "validation_loss": 0.006684564985334873,
          "train_metric": 0.0007881008205004036,
          "validation_metric": 0.0006732449983246624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.748631636000937
        },
        {
          "epoch": 375,
          "updates": 375,
          "samples_seen": 384000,
          "online_loss": 0.007824954576790333,
          "train_loss": 0.00779313687235117,
          "validation_loss": 0.00665618758648634,
          "train_metric": 0.0007848961977288127,
          "validation_metric": 0.0006703867693431675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.75454266899942
        },
        {
          "epoch": 376,
          "updates": 376,
          "samples_seen": 385024,
          "online_loss": 0.007793136406689882,
          "train_loss": 0.007761402986943722,
          "validation_loss": 0.006630581337958574,
          "train_metric": 0.0007816999568603933,
          "validation_metric": 0.0006678078789263964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.760905159999311
        },
        {
          "epoch": 377,
          "updates": 377,
          "samples_seen": 386048,
          "online_loss": 0.0077614025212824345,
          "train_loss": 0.00772979948669672,
          "validation_loss": 0.006603706162422895,
          "train_metric": 0.0007785171619616449,
          "validation_metric": 0.0006651011062785983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.766765657999713
        },
        {
          "epoch": 378,
          "updates": 378,
          "samples_seen": 387072,
          "online_loss": 0.007729799021035433,
          "train_loss": 0.0076983352191746235,
          "validation_loss": 0.006576582323759794,
          "train_metric": 0.0007753480458632112,
          "validation_metric": 0.0006623693043366075,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.772622656999374
        },
        {
          "epoch": 379,
          "updates": 379,
          "samples_seen": 388096,
          "online_loss": 0.007698334287852049,
          "train_loss": 0.007666975725442171,
          "validation_loss": 0.006551343481987715,
          "train_metric": 0.0007721898145973682,
          "validation_metric": 0.0006598272011615336,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.778977334000956
        },
        {
          "epoch": 380,
          "updates": 380,
          "samples_seen": 389120,
          "online_loss": 0.007666975725442171,
          "train_loss": 0.00763569725677371,
          "validation_loss": 0.006523622665554285,
          "train_metric": 0.0007690393831580877,
          "validation_metric": 0.00065703532891348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7846876269995846
        },
        {
          "epoch": 381,
          "updates": 381,
          "samples_seen": 390144,
          "online_loss": 0.007635696791112423,
          "train_loss": 0.007604505401104689,
          "validation_loss": 0.006498339585959911,
          "train_metric": 0.0007658979156985879,
          "validation_metric": 0.0006544889765791595,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.790703198999836
        },
        {
          "epoch": 382,
          "updates": 382,
          "samples_seen": 391168,
          "online_loss": 0.007604505401104689,
          "train_loss": 0.007573440670967102,
          "validation_loss": 0.006471645087003708,
          "train_metric": 0.0007627691957168281,
          "validation_metric": 0.0006518004811368883,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.796861500999512
        },
        {
          "epoch": 383,
          "updates": 383,
          "samples_seen": 392192,
          "online_loss": 0.007573440670967102,
          "train_loss": 0.007542495150119066,
          "validation_loss": 0.006445253733545542,
          "train_metric": 0.0007596524083055556,
          "validation_metric": 0.0006491424283012748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.80279008800062
        },
        {
          "epoch": 384,
          "updates": 384,
          "samples_seen": 393216,
          "online_loss": 0.007542494684457779,
          "train_loss": 0.007511670235544443,
          "validation_loss": 0.006419970653951168,
          "train_metric": 0.0007565479027107358,
          "validation_metric": 0.0006465960177592933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8089923900006397
        },
        {
          "epoch": 385,
          "updates": 385,
          "samples_seen": 394240,
          "online_loss": 0.007511670235544443,
          "train_loss": 0.007480933330953121,
          "validation_loss": 0.00639291200786829,
          "train_metric": 0.0007534523028880358,
          "validation_metric": 0.0006438706768676639,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8148604960006196
        },
        {
          "epoch": 386,
          "updates": 386,
          "samples_seen": 395264,
          "online_loss": 0.007480933330953121,
          "train_loss": 0.007450300734490156,
          "validation_loss": 0.006367967929691076,
          "train_metric": 0.0007503668894059956,
          "validation_metric": 0.0006413584342226386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8209907559994463
        },
        {
          "epoch": 387,
          "updates": 387,
          "samples_seen": 396288,
          "online_loss": 0.007450300734490156,
          "train_loss": 0.007419774308800697,
          "validation_loss": 0.006341410335153341,
          "train_metric": 0.0007472924771718681,
          "validation_metric": 0.0006386835011653602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8271650269998645
        },
        {
          "epoch": 388,
          "updates": 388,
          "samples_seen": 397312,
          "online_loss": 0.007419774308800697,
          "train_loss": 0.0073893568478524685,
          "validation_loss": 0.006315832957625389,
          "train_metric": 0.0007442290079779923,
          "validation_metric": 0.0006361075211316347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8330716630007373
        },
        {
          "epoch": 389,
          "updates": 389,
          "samples_seen": 398336,
          "online_loss": 0.007389357313513756,
          "train_loss": 0.007359063718467951,
          "validation_loss": 0.006290399469435215,
          "train_metric": 0.0007411779370158911,
          "validation_metric": 0.0006335459183901548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8394374849995074
        },
        {
          "epoch": 390,
          "updates": 390,
          "samples_seen": 399360,
          "online_loss": 0.007359063718467951,
          "train_loss": 0.007328876294195652,
          "validation_loss": 0.006264133844524622,
          "train_metric": 0.0007381375762633979,
          "validation_metric": 0.0006309006130322814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8455831310002395
        },
        {
          "epoch": 391,
          "updates": 391,
          "samples_seen": 400384,
          "online_loss": 0.007328876294195652,
          "train_loss": 0.007298794109374285,
          "validation_loss": 0.006239399779587984,
          "train_metric": 0.000735107809305191,
          "validation_metric": 0.0006284094997681677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8513722620009503
        },
        {
          "epoch": 392,
          "updates": 392,
          "samples_seen": 401408,
          "online_loss": 0.007298794109374285,
          "train_loss": 0.007268812041729689,
          "validation_loss": 0.00621312391012907,
          "train_metric": 0.0007320881122723222,
          "validation_metric": 0.0006257630884647369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8572001440006716
        },
        {
          "epoch": 393,
          "updates": 393,
          "samples_seen": 402432,
          "online_loss": 0.007268811576068401,
          "train_loss": 0.007238939870148897,
          "validation_loss": 0.00618830556049943,
          "train_metric": 0.0007290794746950269,
          "validation_metric": 0.0006232634768821299,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.8630687500008207
        },
        {
          "epoch": 394,
          "updates": 394,
          "samples_seen": 403456,
          "online_loss": 0.0072389403358101845,
          "train_loss": 0.0072091808542609215,
          "validation_loss": 0.006162720732390881,
          "train_metric": 0.0007260822458192706,
          "validation_metric": 0.0006206866237334907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.868831746000069
        },
        {
          "epoch": 395,
          "updates": 395,
          "samples_seen": 404480,
          "online_loss": 0.007209180388599634,
          "train_loss": 0.007179534062743187,
          "validation_loss": 0.006137385498732328,
          "train_metric": 0.0007230963674373925,
          "validation_metric": 0.000618135032709688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.949516261000099
        },
        {
          "epoch": 396,
          "updates": 396,
          "samples_seen": 405504,
          "online_loss": 0.007179534528404474,
          "train_loss": 0.00714999670162797,
          "validation_loss": 0.00611260486766696,
          "train_metric": 0.000720121490303427,
          "validation_metric": 0.0006156391464173794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9555721300002915
        },
        {
          "epoch": 397,
          "updates": 397,
          "samples_seen": 406528,
          "online_loss": 0.00714999670162797,
          "train_loss": 0.007120572961866856,
          "validation_loss": 0.006086932960897684,
          "train_metric": 0.0007171579636633396,
          "validation_metric": 0.0006130535621196032,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.961885271000938
        },
        {
          "epoch": 398,
          "updates": 398,
          "samples_seen": 407552,
          "online_loss": 0.007120572496205568,
          "train_loss": 0.007091253064572811,
          "validation_loss": 0.00606258399784565,
          "train_metric": 0.0007142050890251994,
          "validation_metric": 0.0006106012733653188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9676590760009276
        },
        {
          "epoch": 399,
          "updates": 399,
          "samples_seen": 408576,
          "online_loss": 0.007091253064572811,
          "train_loss": 0.007062043063342571,
          "validation_loss": 0.006037057377398014,
          "train_metric": 0.00071126309921965,
          "validation_metric": 0.0006080302409827709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9734608529997786
        },
        {
          "epoch": 400,
          "updates": 400,
          "samples_seen": 409600,
          "online_loss": 0.007062042597681284,
          "train_loss": 0.007032935973256826,
          "validation_loss": 0.006012636236846447,
          "train_metric": 0.0007083315867930651,
          "validation_metric": 0.0006055706180632114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9798149460002605
        },
        {
          "epoch": 401,
          "updates": 401,
          "samples_seen": 410624,
          "online_loss": 0.007032935973256826,
          "train_loss": 0.007003942970186472,
          "validation_loss": 0.005987630225718021,
          "train_metric": 0.0007054114248603582,
          "validation_metric": 0.0006030522054061294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9857287800004997
        },
        {
          "epoch": 402,
          "updates": 402,
          "samples_seen": 411648,
          "online_loss": 0.007003942970186472,
          "train_loss": 0.006975062191486359,
          "validation_loss": 0.005962934345006943,
          "train_metric": 0.0007025027880445123,
          "validation_metric": 0.0006005648174323142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9917986130003555
        },
        {
          "epoch": 403,
          "updates": 403,
          "samples_seen": 412672,
          "online_loss": 0.006975062191486359,
          "train_loss": 0.006946297828108072,
          "validation_loss": 0.005938550923019648,
          "train_metric": 0.0006996057345531881,
          "validation_metric": 0.0005981091526336968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.9977343770005973
        },
        {
          "epoch": 404,
          "updates": 404,
          "samples_seen": 413696,
          "online_loss": 0.00694629829376936,
          "train_loss": 0.006917642429471016,
          "validation_loss": 0.0059136394411325455,
          "train_metric": 0.0006967196241021156,
          "validation_metric": 0.0005956000532023609,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.003367793000507
        },
        {
          "epoch": 405,
          "updates": 405,
          "samples_seen": 414720,
          "online_loss": 0.006917642895132303,
          "train_loss": 0.006889088079333305,
          "validation_loss": 0.005889653693884611,
          "train_metric": 0.0006938436417840421,
          "validation_metric": 0.000593184435274452,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.009545879000143
        },
        {
          "epoch": 406,
          "updates": 406,
          "samples_seen": 415744,
          "online_loss": 0.006889087148010731,
          "train_loss": 0.00686064874753356,
          "validation_loss": 0.005864798091351986,
          "train_metric": 0.000690979475621134,
          "validation_metric": 0.0005906810401938856,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.015371023999251
        },
        {
          "epoch": 407,
          "updates": 407,
          "samples_seen": 416768,
          "online_loss": 0.006860649213194847,
          "train_loss": 0.006832318380475044,
          "validation_loss": 0.005840970668941736,
          "train_metric": 0.000688126077875495,
          "validation_metric": 0.0005882811965420842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.021466813999723
        },
        {
          "epoch": 408,
          "updates": 408,
          "samples_seen": 417792,
          "online_loss": 0.006832317914813757,
          "train_loss": 0.006804101634770632,
          "validation_loss": 0.005816394463181496,
          "train_metric": 0.0006852842052467167,
          "validation_metric": 0.0005858059739693999,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.027652986000248
        },
        {
          "epoch": 409,
          "updates": 409,
          "samples_seen": 418816,
          "online_loss": 0.006804102100431919,
          "train_loss": 0.006775992922484875,
          "validation_loss": 0.005792529322206974,
          "train_metric": 0.0006824532756581903,
          "validation_metric": 0.0005834022304043174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.033537133000209
        },
        {
          "epoch": 410,
          "updates": 410,
          "samples_seen": 419840,
          "online_loss": 0.006775992922484875,
          "train_loss": 0.006747990846633911,
          "validation_loss": 0.005768367554992437,
          "train_metric": 0.00067963293986395,
          "validation_metric": 0.0005809689173474908,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.039341424000668
        },
        {
          "epoch": 411,
          "updates": 411,
          "samples_seen": 420864,
          "online_loss": 0.006747990846633911,
          "train_loss": 0.0067201037891209126,
          "validation_loss": 0.0057444158010184765,
          "train_metric": 0.0006768241291865706,
          "validation_metric": 0.000578556559048593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.045470204000594
        },
        {
          "epoch": 412,
          "updates": 412,
          "samples_seen": 421888,
          "online_loss": 0.006720102857798338,
          "train_loss": 0.006692328955978155,
          "validation_loss": 0.005720663815736771,
          "train_metric": 0.0006740268436260521,
          "validation_metric": 0.0005761642823927104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.051328169000044
        },
        {
          "epoch": 413,
          "updates": 413,
          "samples_seen": 422912,
          "online_loss": 0.006692328490316868,
          "train_loss": 0.006664664950221777,
          "validation_loss": 0.005696664098650217,
          "train_metric": 0.0006712406175211072,
          "validation_metric": 0.0005737471510656178,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.057140142000208
        },
        {
          "epoch": 414,
          "updates": 414,
          "samples_seen": 423936,
          "online_loss": 0.006664664018899202,
          "train_loss": 0.006637110374867916,
          "validation_loss": 0.005673250649124384,
          "train_metric": 0.0006684653926640749,
          "validation_metric": 0.0005713889840990305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.062974735999887
        },
        {
          "epoch": 415,
          "updates": 415,
          "samples_seen": 424960,
          "online_loss": 0.006637110374867916,
          "train_loss": 0.006609667092561722,
          "validation_loss": 0.005649292375892401,
          "train_metric": 0.0006657014600932598,
          "validation_metric": 0.0005689761019311845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.068536535000021
        },
        {
          "epoch": 416,
          "updates": 416,
          "samples_seen": 425984,
          "online_loss": 0.0066096666269004345,
          "train_loss": 0.006582328118383884,
          "validation_loss": 0.005626128055155277,
          "train_metric": 0.0006629479466937482,
          "validation_metric": 0.0005666430224664509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.074210725000739
        },
        {
          "epoch": 417,
          "updates": 417,
          "samples_seen": 427008,
          "online_loss": 0.00658232718706131,
          "train_loss": 0.006555108819156885,
          "validation_loss": 0.005602285265922546,
          "train_metric": 0.0006602065404877067,
          "validation_metric": 0.0005642416654154658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.080041167000672
        },
        {
          "epoch": 418,
          "updates": 418,
          "samples_seen": 428032,
          "online_loss": 0.006555108819156885,
          "train_loss": 0.006528000347316265,
          "validation_loss": 0.005579311866313219,
          "train_metric": 0.0006574763683602214,
          "validation_metric": 0.0005619278526864946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0858769660007965
        },
        {
          "epoch": 419,
          "updates": 419,
          "samples_seen": 429056,
          "online_loss": 0.006528000347316265,
          "train_loss": 0.00650099478662014,
          "validation_loss": 0.0055556464940309525,
          "train_metric": 0.0006547564407810569,
          "validation_metric": 0.000559544307179749,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.0919970810009545
        },
        {
          "epoch": 420,
          "updates": 420,
          "samples_seen": 430080,
          "online_loss": 0.006500994320958853,
          "train_loss": 0.00647410424426198,
          "validation_loss": 0.005532821174710989,
          "train_metric": 0.0006520480383187532,
          "validation_metric": 0.0005572455120272934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.097903828000199
        },
        {
          "epoch": 421,
          "updates": 421,
          "samples_seen": 431104,
          "online_loss": 0.006474103778600693,
          "train_loss": 0.006447324994951487,
          "validation_loss": 0.005509353242814541,
          "train_metric": 0.0006493509281426668,
          "validation_metric": 0.0005548819317482412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.103607713999736
        },
        {
          "epoch": 422,
          "updates": 422,
          "samples_seen": 432128,
          "online_loss": 0.006447325460612774,
          "train_loss": 0.006420657504349947,
          "validation_loss": 0.005486657842993736,
          "train_metric": 0.0006466651102527976,
          "validation_metric": 0.0005525961169041693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.10954979500093
        },
        {
          "epoch": 423,
          "updates": 423,
          "samples_seen": 433152,
          "online_loss": 0.00642065703868866,
          "train_loss": 0.006394094787538052,
          "validation_loss": 0.005463392939418554,
          "train_metric": 0.0006439897697418928,
          "validation_metric": 0.0005502529675140977,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.115383059999658
        },
        {
          "epoch": 424,
          "updates": 424,
          "samples_seen": 434176,
          "online_loss": 0.006394095253199339,
          "train_loss": 0.006367648486047983,
          "validation_loss": 0.0054408591240644455,
          "train_metric": 0.0006413261871784925,
          "validation_metric": 0.0005479835090227425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.121257950000654
        },
        {
          "epoch": 425,
          "updates": 425,
          "samples_seen": 435200,
          "online_loss": 0.0063676489517092705,
          "train_loss": 0.006341305561363697,
          "validation_loss": 0.005417742300778627,
          "train_metric": 0.0006386730819940567,
          "validation_metric": 0.0005456551443785429,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.127397923999524
        },
        {
          "epoch": 426,
          "updates": 426,
          "samples_seen": 436224,
          "online_loss": 0.00634130509570241,
          "train_loss": 0.0063150785863399506,
          "validation_loss": 0.005395415239036083,
          "train_metric": 0.0006360316183418036,
          "validation_metric": 0.0005434065242297947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.133224954999605
        },
        {
          "epoch": 427,
          "updates": 427,
          "samples_seen": 437248,
          "online_loss": 0.0063150785863399506,
          "train_loss": 0.00628896476700902,
          "validation_loss": 0.005372406914830208,
          "train_metric": 0.0006334014469757676,
          "validation_metric": 0.0005410891608335078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.13896511799976
        },
        {
          "epoch": 428,
          "updates": 428,
          "samples_seen": 438272,
          "online_loss": 0.0062889643013477325,
          "train_loss": 0.006262956187129021,
          "validation_loss": 0.00535034341737628,
          "train_metric": 0.0006307819858193398,
          "validation_metric": 0.0005388670251704752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.1450764990004245
        },
        {
          "epoch": 429,
          "updates": 429,
          "samples_seen": 439296,
          "online_loss": 0.006262956652790308,
          "train_loss": 0.006237058900296688,
          "validation_loss": 0.005327374208718538,
          "train_metric": 0.0006281737587414682,
          "validation_metric": 0.0005365536198951304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.151043685000332
        },
        {
          "epoch": 430,
          "updates": 430,
          "samples_seen": 440320,
          "online_loss": 0.006237058900296688,
          "train_loss": 0.006211273837834597,
          "validation_loss": 0.005305679980665445,
          "train_metric": 0.0006255767657421529,
          "validation_metric": 0.0005343686789274216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.157296719000442
        },
        {
          "epoch": 431,
          "updates": 431,
          "samples_seen": 441344,
          "online_loss": 0.006211273837834597,
          "train_loss": 0.006185600534081459,
          "validation_loss": 0.005282602272927761,
          "train_metric": 0.0006229911232367158,
          "validation_metric": 0.0005320443888194859,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.163268968999546
        },
        {
          "epoch": 432,
          "updates": 432,
          "samples_seen": 442368,
          "online_loss": 0.006185601465404034,
          "train_loss": 0.006160035729408264,
          "validation_loss": 0.005261456593871117,
          "train_metric": 0.0006204161909408867,
          "validation_metric": 0.000529914628714323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.168840128000738
        },
        {
          "epoch": 433,
          "updates": 433,
          "samples_seen": 443392,
          "online_loss": 0.006160035729408264,
          "train_loss": 0.006134594790637493,
          "validation_loss": 0.005238051991909742,
          "train_metric": 0.0006178538897074759,
          "validation_metric": 0.0005275573930703104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.17459738800062
        },
        {
          "epoch": 434,
          "updates": 434,
          "samples_seen": 444416,
          "online_loss": 0.006134594324976206,
          "train_loss": 0.0061092684045434,
          "validation_loss": 0.005217799451202154,
          "train_metric": 0.0006153031717985868,
          "validation_metric": 0.0005255176220089197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.18077798000013
        },
        {
          "epoch": 435,
          "updates": 435,
          "samples_seen": 445440,
          "online_loss": 0.0061092684045434,
          "train_loss": 0.006084077525883913,
          "validation_loss": 0.005193622782826424,
          "train_metric": 0.0006127659580670297,
          "validation_metric": 0.0005230826791375875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.275526826999339
        },
        {
          "epoch": 436,
          "updates": 436,
          "samples_seen": 446464,
          "online_loss": 0.006084077525883913,
          "train_loss": 0.00605903472751379,
          "validation_loss": 0.005174973979592323,
          "train_metric": 0.0006102437619119883,
          "validation_metric": 0.0005212044343352318,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.281997976999264
        },
        {
          "epoch": 437,
          "updates": 437,
          "samples_seen": 447488,
          "online_loss": 0.0060590351931750774,
          "train_loss": 0.006034188438206911,
          "validation_loss": 0.0051491414196789265,
          "train_metric": 0.0006077412399463356,
          "validation_metric": 0.0005186025518923998,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.287846638999326
        },
        {
          "epoch": 438,
          "updates": 438,
          "samples_seen": 448512,
          "online_loss": 0.0060341875068843365,
          "train_loss": 0.006009622011333704,
          "validation_loss": 0.005133690312504768,
          "train_metric": 0.0006052671815268695,
          "validation_metric": 0.0005170464864932001,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.294153240000014
        },
        {
          "epoch": 439,
          "updates": 439,
          "samples_seen": 449536,
          "online_loss": 0.006009621545672417,
          "train_loss": 0.005985528230667114,
          "validation_loss": 0.005104547366499901,
          "train_metric": 0.000602840562351048,
          "validation_metric": 0.0005141113651916385,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.299995494999166
        },
        {
          "epoch": 440,
          "updates": 440,
          "samples_seen": 450560,
          "online_loss": 0.0059855286963284016,
          "train_loss": 0.005962285213172436,
          "validation_loss": 0.005096100736409426,
          "train_metric": 0.0006004995084367692,
          "validation_metric": 0.0005132606020197272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.306073174999256
        },
        {
          "epoch": 441,
          "updates": 441,
          "samples_seen": 451584,
          "online_loss": 0.005962285213172436,
          "train_loss": 0.005940768867731094,
          "validation_loss": 0.005060903262346983,
          "train_metric": 0.0005983323790132999,
          "validation_metric": 0.0005097156390547752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.312572647999332
        },
        {
          "epoch": 442,
          "updates": 442,
          "samples_seen": 452608,
          "online_loss": 0.00594076793640852,
          "train_loss": 0.005922776181250811,
          "validation_loss": 0.005069847218692303,
          "train_metric": 0.0005965203163214028,
          "validation_metric": 0.0005106164608150721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.318020426000658
        },
        {
          "epoch": 443,
          "updates": 443,
          "samples_seen": 453632,
          "online_loss": 0.005922776646912098,
          "train_loss": 0.005912355612963438,
          "validation_loss": 0.005026886239647865,
          "train_metric": 0.000595470832195133,
          "validation_metric": 0.0005062896525487304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.323918354999478
        },
        {
          "epoch": 444,
          "updates": 444,
          "samples_seen": 454656,
          "online_loss": 0.005912355612963438,
          "train_loss": 0.005917022470384836,
          "validation_loss": 0.0050839330069720745,
          "train_metric": 0.0005959408008493483,
          "validation_metric": 0.0005120350979268551,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.329500605999783
        },
        {
          "epoch": 445,
          "updates": 445,
          "samples_seen": 455680,
          "online_loss": 0.005917022004723549,
          "train_loss": 0.005951241124421358,
          "validation_loss": 0.005045617464929819,
          "train_metric": 0.0005993872764520347,
          "validation_metric": 0.0005081761046312749,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.3349190700009785
        },
        {
          "epoch": 446,
          "updates": 446,
          "samples_seen": 456704,
          "online_loss": 0.005951240658760071,
          "train_loss": 0.006025376729667187,
          "validation_loss": 0.0052184490486979485,
          "train_metric": 0.0006068538641557097,
          "validation_metric": 0.0005255831638351083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.340060287000597
        },
        {
          "epoch": 447,
          "updates": 447,
          "samples_seen": 457728,
          "online_loss": 0.006025377195328474,
          "train_loss": 0.006128461565822363,
          "validation_loss": 0.005188141018152237,
          "train_metric": 0.0006172362482175231,
          "validation_metric": 0.0005225305212661624,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.345790714000032
        },
        {
          "epoch": 448,
          "updates": 448,
          "samples_seen": 458752,
          "online_loss": 0.006128461565822363,
          "train_loss": 0.00615021213889122,
          "validation_loss": 0.00535599747672677,
          "train_metric": 0.0006194269517436624,
          "validation_metric": 0.0005394364707171917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.351114131000941
        },
        {
          "epoch": 449,
          "updates": 449,
          "samples_seen": 459776,
          "online_loss": 0.00615021213889122,
          "train_loss": 0.006005745846778154,
          "validation_loss": 0.005080681759864092,
          "train_metric": 0.0006048767827451229,
          "validation_metric": 0.0005117076216265559,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.3565647110008285
        },
        {
          "epoch": 450,
          "updates": 450,
          "samples_seen": 460800,
          "online_loss": 0.006005745846778154,
          "train_loss": 0.005780122242867947,
          "validation_loss": 0.004952471703290939,
          "train_metric": 0.000582152686547488,
          "validation_metric": 0.0004987947759218514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.362573046999387
        },
        {
          "epoch": 451,
          "updates": 451,
          "samples_seen": 461824,
          "online_loss": 0.005780121777206659,
          "train_loss": 0.005717719905078411,
          "validation_loss": 0.004879597574472427,
          "train_metric": 0.0005758678889833391,
          "validation_metric": 0.0004914551391266286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.368403316000695
        },
        {
          "epoch": 452,
          "updates": 452,
          "samples_seen": 462848,
          "online_loss": 0.005717719905078411,
          "train_loss": 0.00582148227840662,
          "validation_loss": 0.004924903158098459,
          "train_metric": 0.0005863183178007603,
          "validation_metric": 0.0004960181540809572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.373605962000511
        },
        {
          "epoch": 453,
          "updates": 453,
          "samples_seen": 463872,
          "online_loss": 0.005821483675390482,
          "train_loss": 0.005864258855581284,
          "validation_loss": 0.005064626224339008,
          "train_metric": 0.0005906266742385924,
          "validation_metric": 0.0005100906128063798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.379381022999951
        },
        {
          "epoch": 454,
          "updates": 454,
          "samples_seen": 464896,
          "online_loss": 0.005864258855581284,
          "train_loss": 0.005745203234255314,
          "validation_loss": 0.004861348774284124,
          "train_metric": 0.0005786358378827572,
          "validation_metric": 0.000489617174025625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.385169397000936
        },
        {
          "epoch": 455,
          "updates": 455,
          "samples_seen": 465920,
          "online_loss": 0.005745203699916601,
          "train_loss": 0.005626029800623655,
          "validation_loss": 0.004794072359800339,
          "train_metric": 0.0005666330689564347,
          "validation_metric": 0.00048284136573784053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.391185303000384
        },
        {
          "epoch": 456,
          "updates": 456,
          "samples_seen": 466944,
          "online_loss": 0.005626029334962368,
          "train_loss": 0.005653627682477236,
          "validation_loss": 0.004846238996833563,
          "train_metric": 0.0005694127758033574,
          "validation_metric": 0.0004880955384578556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.396724611000536
        },
        {
          "epoch": 457,
          "updates": 457,
          "samples_seen": 467968,
          "online_loss": 0.005653628148138523,
          "train_loss": 0.005709425080567598,
          "validation_loss": 0.004836018197238445,
          "train_metric": 0.0005750323762185872,
          "validation_metric": 0.00048706604866310954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.401878095000939
        },
        {
          "epoch": 458,
          "updates": 458,
          "samples_seen": 468992,
          "online_loss": 0.005709424614906311,
          "train_loss": 0.00564693845808506,
          "validation_loss": 0.004850100725889206,
          "train_metric": 0.0005687389057129622,
          "validation_metric": 0.0004884844529442489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.4069373839993204
        },
        {
          "epoch": 459,
          "updates": 459,
          "samples_seen": 470016,
          "online_loss": 0.00564693845808506,
          "train_loss": 0.0055503747425973415,
          "validation_loss": 0.004717221483588219,
          "train_metric": 0.0005590134533122182,
          "validation_metric": 0.0004751012602355331,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.413275225999314
        },
        {
          "epoch": 460,
          "updates": 460,
          "samples_seen": 471040,
          "online_loss": 0.0055503747425973415,
          "train_loss": 0.005548297893255949,
          "validation_loss": 0.0047110700979828835,
          "train_metric": 0.0005588041967712343,
          "validation_metric": 0.0004744817561004311,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.419240618000913
        },
        {
          "epoch": 461,
          "updates": 461,
          "samples_seen": 472064,
          "online_loss": 0.005548297427594662,
          "train_loss": 0.005583335179835558,
          "validation_loss": 0.004796434193849564,
          "train_metric": 0.0005623329780064523,
          "validation_metric": 0.0004830793186556548,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.424861370000144
        },
        {
          "epoch": 462,
          "updates": 462,
          "samples_seen": 473088,
          "online_loss": 0.005583334714174271,
          "train_loss": 0.005545533262193203,
          "validation_loss": 0.004700015764683485,
          "train_metric": 0.0005585258477367461,
          "validation_metric": 0.0004733683599624783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.430626772000323
        },
        {
          "epoch": 463,
          "updates": 463,
          "samples_seen": 474112,
          "online_loss": 0.005545533262193203,
          "train_loss": 0.0054727597162127495,
          "validation_loss": 0.004667599219828844,
          "train_metric": 0.0005511963390745223,
          "validation_metric": 0.00047010352136567235,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.436568156999783
        },
        {
          "epoch": 464,
          "updates": 464,
          "samples_seen": 475136,
          "online_loss": 0.005472759250551462,
          "train_loss": 0.005460731219500303,
          "validation_loss": 0.004664131440222263,
          "train_metric": 0.0005499848630279303,
          "validation_metric": 0.00046975421719253063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.442532668999775
        },
        {
          "epoch": 465,
          "updates": 465,
          "samples_seen": 476160,
          "online_loss": 0.005460730753839016,
          "train_loss": 0.00547958817332983,
          "validation_loss": 0.0046471599489450455,
          "train_metric": 0.0005518840625882149,
          "validation_metric": 0.0004680449783336371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.44876237400058
        },
        {
          "epoch": 466,
          "updates": 466,
          "samples_seen": 477184,
          "online_loss": 0.005479587707668543,
          "train_loss": 0.005450176540762186,
          "validation_loss": 0.00467004906386137,
          "train_metric": 0.0005489217583090067,
          "validation_metric": 0.0004703502345364541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.454198637000445
        },
        {
          "epoch": 467,
          "updates": 467,
          "samples_seen": 478208,
          "online_loss": 0.005450176075100899,
          "train_loss": 0.00539557496085763,
          "validation_loss": 0.004586576949805021,
          "train_metric": 0.0005434225313365459,
          "validation_metric": 0.00046194318565540016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.460469284000283
        },
        {
          "epoch": 468,
          "updates": 468,
          "samples_seen": 479232,
          "online_loss": 0.0053955744951963425,
          "train_loss": 0.005379514768719673,
          "validation_loss": 0.004572260659188032,
          "train_metric": 0.0005418050568550825,
          "validation_metric": 0.0004605012945830822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.4662411530007375
        },
        {
          "epoch": 469,
          "updates": 469,
          "samples_seen": 480256,
          "online_loss": 0.005379514768719673,
          "train_loss": 0.0053860838524997234,
          "validation_loss": 0.004614884965121746,
          "train_metric": 0.0005424665869213641,
          "validation_metric": 0.00046479437150992453,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.471602179000911
        },
        {
          "epoch": 470,
          "updates": 470,
          "samples_seen": 481280,
          "online_loss": 0.005386083386838436,
          "train_loss": 0.0053630699403584,
          "validation_loss": 0.0045541757717728615,
          "train_metric": 0.0005401488160714507,
          "validation_metric": 0.00045867988956160843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.4778767860007065
        },
        {
          "epoch": 471,
          "updates": 471,
          "samples_seen": 482304,
          "online_loss": 0.005363069474697113,
          "train_loss": 0.005320402327924967,
          "validation_loss": 0.004541447851806879,
          "train_metric": 0.0005358514608815312,
          "validation_metric": 0.00045739798224531114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.483629432999805
        },
        {
          "epoch": 472,
          "updates": 472,
          "samples_seen": 483328,
          "online_loss": 0.005320402793586254,
          "train_loss": 0.005301722325384617,
          "validation_loss": 0.004524496383965015,
          "train_metric": 0.0005339700728654861,
          "validation_metric": 0.00045569075155071914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.565843934000441
        },
        {
          "epoch": 473,
          "updates": 473,
          "samples_seen": 484352,
          "online_loss": 0.005301722325384617,
          "train_loss": 0.00530044361948967,
          "validation_loss": 0.0045011816546320915,
          "train_metric": 0.0005338412593118846,
          "validation_metric": 0.00045334253809414804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.571776266000597
        },
        {
          "epoch": 474,
          "updates": 474,
          "samples_seen": 485376,
          "online_loss": 0.00530044361948967,
          "train_loss": 0.005281118676066399,
          "validation_loss": 0.004519022069871426,
          "train_metric": 0.000531894969753921,
          "validation_metric": 0.0004551393212750554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.577341419999357
        },
        {
          "epoch": 475,
          "updates": 475,
          "samples_seen": 486400,
          "online_loss": 0.005281119607388973,
          "train_loss": 0.005246984772384167,
          "validation_loss": 0.004462258890271187,
          "train_metric": 0.0005284571088850498,
          "validation_metric": 0.00044942236854694784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.583577145000163
        },
        {
          "epoch": 476,
          "updates": 476,
          "samples_seen": 487424,
          "online_loss": 0.005246985238045454,
          "train_loss": 0.005226347129791975,
          "validation_loss": 0.00444546015933156,
          "train_metric": 0.0005263785715214908,
          "validation_metric": 0.00044773047557100654,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.589825466999173
        },
        {
          "epoch": 477,
          "updates": 477,
          "samples_seen": 488448,
          "online_loss": 0.005226346664130688,
          "train_loss": 0.005219349171966314,
          "validation_loss": 0.004458371549844742,
          "train_metric": 0.0005256737349554896,
          "validation_metric": 0.0004490308929234743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.595490346000588
        },
        {
          "epoch": 478,
          "updates": 478,
          "samples_seen": 489472,
          "online_loss": 0.005219349637627602,
          "train_loss": 0.0052032470703125,
          "validation_loss": 0.004417611751705408,
          "train_metric": 0.0005240520695224404,
          "validation_metric": 0.00044492565211839974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.601111837000644
        },
        {
          "epoch": 479,
          "updates": 479,
          "samples_seen": 490496,
          "online_loss": 0.0052032470703125,
          "train_loss": 0.0051751588471233845,
          "validation_loss": 0.004413962364196777,
          "train_metric": 0.0005212230025790632,
          "validation_metric": 0.0004445580707397312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.606865634999849
        },
        {
          "epoch": 480,
          "updates": 480,
          "samples_seen": 491520,
          "online_loss": 0.005175158381462097,
          "train_loss": 0.005153283942490816,
          "validation_loss": 0.0043878378346562386,
          "train_metric": 0.0005190200172364712,
          "validation_metric": 0.00044192696805112064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.613310165999792
        },
        {
          "epoch": 481,
          "updates": 481,
          "samples_seen": 492544,
          "online_loss": 0.005153284408152103,
          "train_loss": 0.005142100155353546,
          "validation_loss": 0.004366682842373848,
          "train_metric": 0.0005178934661671519,
          "validation_metric": 0.000439796393038705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.619083651999972
        },
        {
          "epoch": 482,
          "updates": 482,
          "samples_seen": 493568,
          "online_loss": 0.005142100155353546,
          "train_loss": 0.005127749405801296,
          "validation_loss": 0.0043765646405518055,
          "train_metric": 0.0005164481117390096,
          "validation_metric": 0.00044079151120968163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.624704620999182
        },
        {
          "epoch": 483,
          "updates": 483,
          "samples_seen": 494592,
          "online_loss": 0.0051277498714625835,
          "train_loss": 0.005104556679725647,
          "validation_loss": 0.004336550831794739,
          "train_metric": 0.0005141122383065522,
          "validation_metric": 0.0004367615038063377,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.630736273000366
        },
        {
          "epoch": 484,
          "updates": 484,
          "samples_seen": 495616,
          "online_loss": 0.00510455621406436,
          "train_loss": 0.005082395393401384,
          "validation_loss": 0.004324785899370909,
          "train_metric": 0.0005118803237564862,
          "validation_metric": 0.0004355765995569527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.63663527800054
        },
        {
          "epoch": 485,
          "updates": 485,
          "samples_seen": 496640,
          "online_loss": 0.005082395859062672,
          "train_loss": 0.005067810416221619,
          "validation_loss": 0.004317300394177437,
          "train_metric": 0.0005104112788103521,
          "validation_metric": 0.0004348226939328015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.643243602000439
        },
        {
          "epoch": 486,
          "updates": 486,
          "samples_seen": 497664,
          "online_loss": 0.005067810416221619,
          "train_loss": 0.005054162815213203,
          "validation_loss": 0.004289817065000534,
          "train_metric": 0.0005090368213132024,
          "validation_metric": 0.00043205474503338337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.652875824000148
        },
        {
          "epoch": 487,
          "updates": 487,
          "samples_seen": 498688,
          "online_loss": 0.005054162815213203,
          "train_loss": 0.005034736357629299,
          "validation_loss": 0.004290166310966015,
          "train_metric": 0.0005070801707915962,
          "validation_metric": 0.0004320899024605751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.6586362029993325
        },
        {
          "epoch": 488,
          "updates": 488,
          "samples_seen": 499712,
          "online_loss": 0.005034736357629299,
          "train_loss": 0.005013400688767433,
          "validation_loss": 0.004260144662111998,
          "train_metric": 0.0005049313767813146,
          "validation_metric": 0.00042906621820293367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.667878656000539
        },
        {
          "epoch": 489,
          "updates": 489,
          "samples_seen": 500736,
          "online_loss": 0.005013400688767433,
          "train_loss": 0.0049963234923779964,
          "validation_loss": 0.004244146402925253,
          "train_metric": 0.0005032113986089826,
          "validation_metric": 0.0004274549428373575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.67637394600024
        },
        {
          "epoch": 490,
          "updates": 490,
          "samples_seen": 501760,
          "online_loss": 0.0049963234923779964,
          "train_loss": 0.0049820938147604465,
          "validation_loss": 0.004241675604134798,
          "train_metric": 0.0005017782095819712,
          "validation_metric": 0.00042720595956780016,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.682206871000744
        },
        {
          "epoch": 491,
          "updates": 491,
          "samples_seen": 502784,
          "online_loss": 0.004982093349099159,
          "train_loss": 0.0049654110334813595,
          "validation_loss": 0.004212919156998396,
          "train_metric": 0.000500097987242043,
          "validation_metric": 0.00042430980829522014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.687868756000171
        },
        {
          "epoch": 492,
          "updates": 492,
          "samples_seen": 503808,
          "online_loss": 0.004965410567820072,
          "train_loss": 0.004945795983076096,
          "validation_loss": 0.0042075044475495815,
          "train_metric": 0.0004981224774383008,
          "validation_metric": 0.0004237644898239523,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.694353844000943
        },
        {
          "epoch": 493,
          "updates": 493,
          "samples_seen": 504832,
          "online_loss": 0.004945795983076096,
          "train_loss": 0.004927377682179213,
          "validation_loss": 0.004186879377812147,
          "train_metric": 0.0004962674574926496,
          "validation_metric": 0.00042168714571744204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.700246348000292
        },
        {
          "epoch": 494,
          "updates": 494,
          "samples_seen": 505856,
          "online_loss": 0.0049273781478405,
          "train_loss": 0.004911820404231548,
          "validation_loss": 0.004168929997831583,
          "train_metric": 0.000494700507260859,
          "validation_metric": 0.00041987933218479156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.705997392000427
        },
        {
          "epoch": 495,
          "updates": 495,
          "samples_seen": 506880,
          "online_loss": 0.004911820404231548,
          "train_loss": 0.004896415863186121,
          "validation_loss": 0.00416621845215559,
          "train_metric": 0.0004931490402668715,
          "validation_metric": 0.00041960630915127695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.71199644599983
        },
        {
          "epoch": 496,
          "updates": 496,
          "samples_seen": 507904,
          "online_loss": 0.004896415397524834,
          "train_loss": 0.00487891212105751,
          "validation_loss": 0.004140087869018316,
          "train_metric": 0.0004913861630484462,
          "validation_metric": 0.0004169745370745659,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.717715782000596
        },
        {
          "epoch": 497,
          "updates": 497,
          "samples_seen": 508928,
          "online_loss": 0.00487891212105751,
          "train_loss": 0.004860554821789265,
          "validation_loss": 0.004131512250751257,
          "train_metric": 0.0004895373131148517,
          "validation_metric": 0.0004161108226981014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.723431235999669
        },
        {
          "epoch": 498,
          "updates": 498,
          "samples_seen": 509952,
          "online_loss": 0.004860554821789265,
          "train_loss": 0.0048436657525599,
          "validation_loss": 0.004115728661417961,
          "train_metric": 0.0004878362233284861,
          "validation_metric": 0.00041452108416706324,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.729581722000148
        },
        {
          "epoch": 499,
          "updates": 499,
          "samples_seen": 510976,
          "online_loss": 0.004843664821237326,
          "train_loss": 0.004828166216611862,
          "validation_loss": 0.004097010940313339,
          "train_metric": 0.0004862751520704478,
          "validation_metric": 0.0004126358835492283,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.735574321000968
        },
        {
          "epoch": 500,
          "updates": 500,
          "samples_seen": 512000,
          "online_loss": 0.004828165750950575,
          "train_loss": 0.004812204744666815,
          "validation_loss": 0.0040921238251030445,
          "train_metric": 0.0004846676310990006,
          "validation_metric": 0.00041214373777620494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.741612997999255
        },
        {
          "epoch": 501,
          "updates": 501,
          "samples_seen": 513024,
          "online_loss": 0.004812205210328102,
          "train_loss": 0.004795031622052193,
          "validation_loss": 0.004068543668836355,
          "train_metric": 0.000482937932247296,
          "validation_metric": 0.00040976880700327456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.74762085600014
        },
        {
          "epoch": 502,
          "updates": 502,
          "samples_seen": 514048,
          "online_loss": 0.004795031156390905,
          "train_loss": 0.004777717869728804,
          "validation_loss": 0.004058061633259058,
          "train_metric": 0.0004811942344531417,
          "validation_metric": 0.00040871312376111746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.753545363999365
        },
        {
          "epoch": 503,
          "updates": 503,
          "samples_seen": 515072,
          "online_loss": 0.004777717404067516,
          "train_loss": 0.0047614192590117455,
          "validation_loss": 0.004043733701109886,
          "train_metric": 0.000479552720207721,
          "validation_metric": 0.0004072700394317508,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.759530679000818
        },
        {
          "epoch": 504,
          "updates": 504,
          "samples_seen": 516096,
          "online_loss": 0.0047614192590117455,
          "train_loss": 0.004745827056467533,
          "validation_loss": 0.004025453235954046,
          "train_metric": 0.0004779822484124452,
          "validation_metric": 0.00040542890201322734,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.765460824000911
        },
        {
          "epoch": 505,
          "updates": 505,
          "samples_seen": 517120,
          "online_loss": 0.0047458261251449585,
          "train_loss": 0.004729886539280415,
          "validation_loss": 0.004018636420369148,
          "train_metric": 0.00047637676470912993,
          "validation_metric": 0.00040474237175658345,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.771778536000056
        },
        {
          "epoch": 506,
          "updates": 506,
          "samples_seen": 518144,
          "online_loss": 0.004729886073619127,
          "train_loss": 0.004713280126452446,
          "validation_loss": 0.0039973026141524315,
          "train_metric": 0.00047470437129959464,
          "validation_metric": 0.0004025937232654542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.77791841600083
        },
        {
          "epoch": 507,
          "updates": 507,
          "samples_seen": 519168,
          "online_loss": 0.004713280126452446,
          "train_loss": 0.004696644842624664,
          "validation_loss": 0.003986472263932228,
          "train_metric": 0.00047302889288403094,
          "validation_metric": 0.0004015028243884444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.783805262999522
        },
        {
          "epoch": 508,
          "updates": 508,
          "samples_seen": 520192,
          "online_loss": 0.004696644842624664,
          "train_loss": 0.004680612124502659,
          "validation_loss": 0.003972415812313557,
          "train_metric": 0.0004714141250588,
          "validation_metric": 0.0004000871267635375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.789553725000587
        },
        {
          "epoch": 509,
          "updates": 509,
          "samples_seen": 521216,
          "online_loss": 0.004680612124502659,
          "train_loss": 0.004665039945393801,
          "validation_loss": 0.003955496940761805,
          "train_metric": 0.0004698457196354866,
          "validation_metric": 0.00039838309749029577,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.7955188890009595
        },
        {
          "epoch": 510,
          "updates": 510,
          "samples_seen": 522240,
          "online_loss": 0.004665039945393801,
          "train_loss": 0.004649346694350243,
          "validation_loss": 0.003947170916944742,
          "train_metric": 0.0004682652361225337,
          "validation_metric": 0.0003975445870310068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.93143663400042
        },
        {
          "epoch": 511,
          "updates": 511,
          "samples_seen": 523264,
          "online_loss": 0.00464934716001153,
          "train_loss": 0.004633293952792883,
          "validation_loss": 0.003927561454474926,
          "train_metric": 0.0004666483437176794,
          "validation_metric": 0.0003955696302000433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.96149793999939
        },
        {
          "epoch": 512,
          "updates": 512,
          "samples_seen": 524288,
          "online_loss": 0.004633293952792883,
          "train_loss": 0.004617157392203808,
          "validation_loss": 0.0039169578813016415,
          "train_metric": 0.00046502327313646674,
          "validation_metric": 0.0003945016651414335,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 4.991343386000153
        },
        {
          "epoch": 513,
          "updates": 513,
          "samples_seen": 525312,
          "online_loss": 0.004617157857865095,
          "train_loss": 0.004601328168064356,
          "validation_loss": 0.0039022162090986967,
          "train_metric": 0.00046342884888872504,
          "validation_metric": 0.00039301690412685275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.019041240999286
        },
        {
          "epoch": 514,
          "updates": 514,
          "samples_seen": 526336,
          "online_loss": 0.0046013277024030685,
          "train_loss": 0.00458584725856781,
          "validation_loss": 0.0038867127150297165,
          "train_metric": 0.00046186972758732736,
          "validation_metric": 0.000391455483622849,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.047468045999267
        },
        {
          "epoch": 515,
          "updates": 515,
          "samples_seen": 527360,
          "online_loss": 0.00458584725856781,
          "train_loss": 0.00457043806090951,
          "validation_loss": 0.0038768097292631865,
          "train_metric": 0.0004603177949320525,
          "validation_metric": 0.0003904580953530967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.07664760499938
        },
        {
          "epoch": 516,
          "updates": 516,
          "samples_seen": 528384,
          "online_loss": 0.004570438526570797,
          "train_loss": 0.004554860759526491,
          "validation_loss": 0.003858624491840601,
          "train_metric": 0.0004587489238474518,
          "validation_metric": 0.00038862653309479356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.105919481000456
        },
        {
          "epoch": 517,
          "updates": 517,
          "samples_seen": 529408,
          "online_loss": 0.004554860293865204,
          "train_loss": 0.004539182409644127,
          "validation_loss": 0.0038483955431729555,
          "train_metric": 0.0004571698373183608,
          "validation_metric": 0.000387596373911947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.135534093000388
        },
        {
          "epoch": 518,
          "updates": 518,
          "samples_seen": 530432,
          "online_loss": 0.004539182409644127,
          "train_loss": 0.004523598589003086,
          "validation_loss": 0.003832841757684946,
          "train_metric": 0.0004556002968456596,
          "validation_metric": 0.0003860298020299524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.164732418999847
        },
        {
          "epoch": 519,
          "updates": 519,
          "samples_seen": 531456,
          "online_loss": 0.004523598589003086,
          "train_loss": 0.004508246202021837,
          "validation_loss": 0.003819046774879098,
          "train_metric": 0.0004540540394373238,
          "validation_metric": 0.0003846404142677784,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.193049370000153
        },
        {
          "epoch": 520,
          "updates": 520,
          "samples_seen": 532480,
          "online_loss": 0.00450824573636055,
          "train_loss": 0.00449305959045887,
          "validation_loss": 0.003807678585872054,
          "train_metric": 0.00045252442942000926,
          "validation_metric": 0.00038349549868144095,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.221215785999448
        },
        {
          "epoch": 521,
          "updates": 521,
          "samples_seen": 533504,
          "online_loss": 0.004493059124797583,
          "train_loss": 0.0044778864830732346,
          "validation_loss": 0.003791246097534895,
          "train_metric": 0.00045099633280187845,
          "validation_metric": 0.00038184039294719696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.250204584999665
        },
        {
          "epoch": 522,
          "updates": 522,
          "samples_seen": 534528,
          "online_loss": 0.004477886017411947,
          "train_loss": 0.004462640266865492,
          "validation_loss": 0.0037811649963259697,
          "train_metric": 0.00044946084381081164,
          "validation_metric": 0.0003808251058217138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.294799160999901
        },
        {
          "epoch": 523,
          "updates": 523,
          "samples_seen": 535552,
          "online_loss": 0.004462640732526779,
          "train_loss": 0.004447391256690025,
          "validation_loss": 0.00376520911231637,
          "train_metric": 0.000447924918262288,
          "validation_metric": 0.0003792181087192148,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.361281953999423
        },
        {
          "epoch": 524,
          "updates": 524,
          "samples_seen": 536576,
          "online_loss": 0.004447390791028738,
          "train_loss": 0.004432233050465584,
          "validation_loss": 0.003753083758056164,
          "train_metric": 0.00044639827683568,
          "validation_metric": 0.0003779967955779284,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4046721890008484
        },
        {
          "epoch": 525,
          "updates": 525,
          "samples_seen": 537600,
          "online_loss": 0.0044322325848042965,
          "train_loss": 0.0044172173365950584,
          "validation_loss": 0.00374008622020483,
          "train_metric": 0.0004448859835974872,
          "validation_metric": 0.0003766877925954759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.433978521999961
        },
        {
          "epoch": 526,
          "updates": 526,
          "samples_seen": 538624,
          "online_loss": 0.0044172173365950584,
          "train_loss": 0.004402318969368935,
          "validation_loss": 0.003725519170984626,
          "train_metric": 0.000443385390099138,
          "validation_metric": 0.0003752206393983215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.4625920449998375
        },
        {
          "epoch": 527,
          "updates": 527,
          "samples_seen": 539648,
          "online_loss": 0.004402318503707647,
          "train_loss": 0.004387451335787773,
          "validation_loss": 0.003714723279699683,
          "train_metric": 0.00044188799802213907,
          "validation_metric": 0.0003741332911886275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.492037801999686
        },
        {
          "epoch": 528,
          "updates": 528,
          "samples_seen": 540672,
          "online_loss": 0.004387451335787773,
          "train_loss": 0.0043725850991904736,
          "validation_loss": 0.003699085209518671,
          "train_metric": 0.0004403907514642924,
          "validation_metric": 0.0003725583082996309,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.525356744999954
        },
        {
          "epoch": 529,
          "updates": 529,
          "samples_seen": 541696,
          "online_loss": 0.004372584633529186,
          "train_loss": 0.004357730038464069,
          "validation_loss": 0.0036880786065012217,
          "train_metric": 0.0004388946108520031,
          "validation_metric": 0.00037144977250136435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.554385227000239
        },
        {
          "epoch": 530,
          "updates": 530,
          "samples_seen": 542720,
          "online_loss": 0.004357729572802782,
          "train_loss": 0.004342934116721153,
          "validation_loss": 0.0036737467162311077,
          "train_metric": 0.00043740440742112696,
          "validation_metric": 0.00037000630982220173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.583794472000591
        },
        {
          "epoch": 531,
          "updates": 531,
          "samples_seen": 543744,
          "online_loss": 0.004342934116721153,
          "train_loss": 0.004328228998929262,
          "validation_loss": 0.0036611375398933887,
          "train_metric": 0.00043592340080067515,
          "validation_metric": 0.00036873636418022215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.612539599000229
        },
        {
          "epoch": 532,
          "updates": 532,
          "samples_seen": 544768,
          "online_loss": 0.004328228533267975,
          "train_loss": 0.004313625395298004,
          "validation_loss": 0.003648902289569378,
          "train_metric": 0.0004344525805208832,
          "validation_metric": 0.00036750410799868405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.640486078000322
        },
        {
          "epoch": 533,
          "updates": 533,
          "samples_seen": 545792,
          "online_loss": 0.004313625395298004,
          "train_loss": 0.004299085587263107,
          "validation_loss": 0.00363471289165318,
          "train_metric": 0.00043298816308379173,
          "validation_metric": 0.0003660749353002757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.668294469000102
        },
        {
          "epoch": 534,
          "updates": 534,
          "samples_seen": 546816,
          "online_loss": 0.00429908512160182,
          "train_loss": 0.004284575115889311,
          "validation_loss": 0.003623755881562829,
          "train_metric": 0.0004315267433412373,
          "validation_metric": 0.00036497143446467817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.697405517999869
        },
        {
          "epoch": 535,
          "updates": 535,
          "samples_seen": 547840,
          "online_loss": 0.004284575115889311,
          "train_loss": 0.00427010003477335,
          "validation_loss": 0.003609247272834182,
          "train_metric": 0.0004300688742659986,
          "validation_metric": 0.00036351013113744557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.726406482999664
        },
        {
          "epoch": 536,
          "updates": 536,
          "samples_seen": 548864,
          "online_loss": 0.004270100500434637,
          "train_loss": 0.004255660343915224,
          "validation_loss": 0.003598044393584132,
          "train_metric": 0.00042861452675424516,
          "validation_metric": 0.00036238189204595983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.754887860999588
        },
        {
          "epoch": 537,
          "updates": 537,
          "samples_seen": 549888,
          "online_loss": 0.004255660343915224,
          "train_loss": 0.004241280257701874,
          "validation_loss": 0.00358438934199512,
          "train_metric": 0.0004271661746315658,
          "validation_metric": 0.0003610065905377269,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.783069768000132
        },
        {
          "epoch": 538,
          "updates": 538,
          "samples_seen": 550912,
          "online_loss": 0.0042412797920405865,
          "train_loss": 0.00422696303576231,
          "validation_loss": 0.00357212801463902,
          "train_metric": 0.00042572422535158694,
          "validation_metric": 0.0003597716276999563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.8110953779996635
        },
        {
          "epoch": 539,
          "updates": 539,
          "samples_seen": 551936,
          "online_loss": 0.004226962570101023,
          "train_loss": 0.004212722647935152,
          "validation_loss": 0.0035599111579358578,
          "train_metric": 0.00042429004679434,
          "validation_metric": 0.0003585412341635674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.83900978299971
        },
        {
          "epoch": 540,
          "updates": 540,
          "samples_seen": 552960,
          "online_loss": 0.004212722182273865,
          "train_loss": 0.004198549780994654,
          "validation_loss": 0.0035465965047478676,
          "train_metric": 0.00042286262032575905,
          "validation_metric": 0.0003572002169676125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.867482211000606
        },
        {
          "epoch": 541,
          "updates": 541,
          "samples_seen": 553984,
          "online_loss": 0.004198550246655941,
          "train_loss": 0.0041844225488603115,
          "validation_loss": 0.00353529816493392,
          "train_metric": 0.0004214397631585598,
          "validation_metric": 0.00035606231540441513,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.8965817209991656
        },
        {
          "epoch": 542,
          "updates": 542,
          "samples_seen": 555008,
          "online_loss": 0.0041844225488603115,
          "train_loss": 0.004170337226241827,
          "validation_loss": 0.003521573729813099,
          "train_metric": 0.00042002112604677677,
          "validation_metric": 0.0003546799998730421,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.9253907870006515
        },
        {
          "epoch": 543,
          "updates": 543,
          "samples_seen": 556032,
          "online_loss": 0.004170337691903114,
          "train_loss": 0.0041562956757843494,
          "validation_loss": 0.0035104858689010143,
          "train_metric": 0.00041860685450956225,
          "validation_metric": 0.00035356328589841723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 5.962403089000873
        },
        {
          "epoch": 544,
          "updates": 544,
          "samples_seen": 557056,
          "online_loss": 0.0041562956757843494,
          "train_loss": 0.004142304882407188,
          "validation_loss": 0.003497124183923006,
          "train_metric": 0.00041719782166182995,
          "validation_metric": 0.00035221752477809787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.004440351000085
        },
        {
          "epoch": 545,
          "updates": 545,
          "samples_seen": 558080,
          "online_loss": 0.004142305348068476,
          "train_loss": 0.0041283671744167805,
          "validation_loss": 0.003485568333417177,
          "train_metric": 0.00041579402750357985,
          "validation_metric": 0.00035105369170196354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.023888407000413
        },
        {
          "epoch": 546,
          "updates": 546,
          "samples_seen": 559104,
          "online_loss": 0.0041283671744167805,
          "train_loss": 0.004114488605409861,
          "validation_loss": 0.0034730173647403717,
          "train_metric": 0.0004143962578382343,
          "validation_metric": 0.0003497895668260753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.03430305800066
        },
        {
          "epoch": 547,
          "updates": 547,
          "samples_seen": 560128,
          "online_loss": 0.004114488605409861,
          "train_loss": 0.004100666847079992,
          "validation_loss": 0.0034608454443514347,
          "train_metric": 0.0004130041925236583,
          "validation_metric": 0.00034856368438340724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.127934744999948
        },
        {
          "epoch": 548,
          "updates": 548,
          "samples_seen": 561152,
          "online_loss": 0.00410066731274128,
          "train_loss": 0.004086906090378761,
          "validation_loss": 0.003449148265644908,
          "train_metric": 0.0004116182099096477,
          "validation_metric": 0.00034738556132651865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.134273614999984
        },
        {
          "epoch": 549,
          "updates": 549,
          "samples_seen": 562176,
          "online_loss": 0.004086906090378761,
          "train_loss": 0.004073194228112698,
          "validation_loss": 0.003436448285356164,
          "train_metric": 0.00041023720405064523,
          "validation_metric": 0.0003461065061856061,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.140713525999672
        },
        {
          "epoch": 550,
          "updates": 550,
          "samples_seen": 563200,
          "online_loss": 0.00407319376245141,
          "train_loss": 0.004059534519910812,
          "validation_loss": 0.0034252656623721123,
          "train_metric": 0.000408861436881125,
          "validation_metric": 0.00034498018794693053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.146749301000455
        },
        {
          "epoch": 551,
          "updates": 551,
          "samples_seen": 564224,
          "online_loss": 0.004059534054249525,
          "train_loss": 0.004045926034450531,
          "validation_loss": 0.0034124080557376146,
          "train_metric": 0.0004074909084010869,
          "validation_metric": 0.0003436852421145886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.152870054000232
        },
        {
          "epoch": 552,
          "updates": 552,
          "samples_seen": 565248,
          "online_loss": 0.004045926034450531,
          "train_loss": 0.004032364580780268,
          "validation_loss": 0.003401408204808831,
          "train_metric": 0.0004061250074300915,
          "validation_metric": 0.00034257740480825305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.159282259000975
        },
        {
          "epoch": 553,
          "updates": 553,
          "samples_seen": 566272,
          "online_loss": 0.00403236411511898,
          "train_loss": 0.004018852021545172,
          "validation_loss": 0.003388694254681468,
          "train_metric": 0.00040476402500644326,
          "validation_metric": 0.0003412968653719872,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.165588628000478
        },
        {
          "epoch": 554,
          "updates": 554,
          "samples_seen": 567296,
          "online_loss": 0.004018851090222597,
          "train_loss": 0.0040053860284388065,
          "validation_loss": 0.003377559594810009,
          "train_metric": 0.000403407757403329,
          "validation_metric": 0.0003401754074729979,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.171741272000872
        },
        {
          "epoch": 555,
          "updates": 555,
          "samples_seen": 568320,
          "online_loss": 0.0040053860284388065,
          "train_loss": 0.003991978708654642,
          "validation_loss": 0.003365229582414031,
          "train_metric": 0.0004020574560854584,
          "validation_metric": 0.0003389335470274091,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.178134945999773
        },
        {
          "epoch": 556,
          "updates": 556,
          "samples_seen": 569344,
          "online_loss": 0.003991978242993355,
          "train_loss": 0.003978621680289507,
          "validation_loss": 0.003353868378326297,
          "train_metric": 0.0004007121897302568,
          "validation_metric": 0.00033778930082917213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.184572333000688
        },
        {
          "epoch": 557,
          "updates": 557,
          "samples_seen": 570368,
          "online_loss": 0.003978621680289507,
          "train_loss": 0.003965316340327263,
          "validation_loss": 0.0033419965766370296,
          "train_metric": 0.00039937213296070695,
          "validation_metric": 0.00033659368637017906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.194077193000339
        },
        {
          "epoch": 558,
          "updates": 558,
          "samples_seen": 571392,
          "online_loss": 0.003965316340327263,
          "train_loss": 0.003952061291784048,
          "validation_loss": 0.003330356441438198,
          "train_metric": 0.0003980371984653175,
          "validation_metric": 0.00033542126766406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.2010218989998975
        },
        {
          "epoch": 559,
          "updates": 559,
          "samples_seen": 572416,
          "online_loss": 0.003952061757445335,
          "train_loss": 0.003938859794288874,
          "validation_loss": 0.003318916307762265,
          "train_metric": 0.0003967075317632407,
          "validation_metric": 0.00033426916343159974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.207116096000391
        },
        {
          "epoch": 560,
          "updates": 560,
          "samples_seen": 573440,
          "online_loss": 0.003938859794288874,
          "train_loss": 0.003925707656890154,
          "validation_loss": 0.003307092934846878,
          "train_metric": 0.00039538287092000246,
          "validation_metric": 0.0003330782928969711,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.216666192000048
        },
        {
          "epoch": 561,
          "updates": 561,
          "samples_seen": 574464,
          "online_loss": 0.003925707656890154,
          "train_loss": 0.0039126081392169,
          "validation_loss": 0.0032959873788058758,
          "train_metric": 0.0003940635360777378,
          "validation_metric": 0.0003319598617963493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.222554632999163
        },
        {
          "epoch": 562,
          "updates": 562,
          "samples_seen": 575488,
          "online_loss": 0.0039126076735556126,
          "train_loss": 0.003899557515978813,
          "validation_loss": 0.003284041304141283,
          "train_metric": 0.0003927490906789899,
          "validation_metric": 0.00033075656392611563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.230119866999303
        },
        {
          "epoch": 563,
          "updates": 563,
          "samples_seen": 576512,
          "online_loss": 0.003899556817486882,
          "train_loss": 0.0038865571841597557,
          "validation_loss": 0.0032731876708567142,
          "train_metric": 0.0003914397966582328,
          "validation_metric": 0.00032966348226182163,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.235945184000229
        },
        {
          "epoch": 564,
          "updates": 564,
          "samples_seen": 577536,
          "online_loss": 0.0038865571841597557,
          "train_loss": 0.0038736043497920036,
          "validation_loss": 0.0032612078357487917,
          "train_metric": 0.0003901352465618402,
          "validation_metric": 0.0003284568665549159,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.24236851599926
        },
        {
          "epoch": 565,
          "updates": 565,
          "samples_seen": 578560,
          "online_loss": 0.0038736043497920036,
          "train_loss": 0.003860700409859419,
          "validation_loss": 0.0032505313865840435,
          "train_metric": 0.0003888356441166252,
          "validation_metric": 0.0003273816837463528,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.248132489999989
        },
        {
          "epoch": 566,
          "updates": 566,
          "samples_seen": 579584,
          "online_loss": 0.0038607006426900625,
          "train_loss": 0.003847845597192645,
          "validation_loss": 0.0032385417725890875,
          "train_metric": 0.0003875409020110965,
          "validation_metric": 0.00032617407850921154,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.2537006700003985
        },
        {
          "epoch": 567,
          "updates": 567,
          "samples_seen": 580608,
          "online_loss": 0.0038478458300232887,
          "train_loss": 0.0038350410759449005,
          "validation_loss": 0.0032280373852699995,
          "train_metric": 0.0003862512530758977,
          "validation_metric": 0.00032511609606444836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.260165529000005
        },
        {
          "epoch": 568,
          "updates": 568,
          "samples_seen": 581632,
          "online_loss": 0.003835040843114257,
          "train_loss": 0.0038222873117774725,
          "validation_loss": 0.0032160645350813866,
          "train_metric": 0.0003849667846225202,
          "validation_metric": 0.00032391026616096497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.265887798999756
        },
        {
          "epoch": 569,
          "updates": 569,
          "samples_seen": 582656,
          "online_loss": 0.0038222873117774725,
          "train_loss": 0.0038095759227871895,
          "validation_loss": 0.0032057310454547405,
          "train_metric": 0.0003836865653283894,
          "validation_metric": 0.00032286951318383217,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.271465499999977
        },
        {
          "epoch": 570,
          "updates": 570,
          "samples_seen": 583680,
          "online_loss": 0.0038095759227871895,
          "train_loss": 0.0037969208788126707,
          "validation_loss": 0.003193738404661417,
          "train_metric": 0.0003824119339697063,
          "validation_metric": 0.0003216616460122168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.278378041999531
        },
        {
          "epoch": 571,
          "updates": 571,
          "samples_seen": 584704,
          "online_loss": 0.0037969211116433144,
          "train_loss": 0.003784314962103963,
          "validation_loss": 0.0031836465932428837,
          "train_metric": 0.00038114230846986175,
          "validation_metric": 0.0003206452529411763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.28424608000023
        },
        {
          "epoch": 572,
          "updates": 572,
          "samples_seen": 585728,
          "online_loss": 0.003784314962103963,
          "train_loss": 0.0037717539817094803,
          "validation_loss": 0.0031715636141598225,
          "train_metric": 0.0003798773104790598,
          "validation_metric": 0.000319428276270628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.291739016000065
        },
        {
          "epoch": 573,
          "updates": 573,
          "samples_seen": 586752,
          "online_loss": 0.0037717537488788366,
          "train_loss": 0.0037592435255646706,
          "validation_loss": 0.003161799628287554,
          "train_metric": 0.00037861726013943553,
          "validation_metric": 0.00031844485783949494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.2989413300001615
        },
        {
          "epoch": 574,
          "updates": 574,
          "samples_seen": 587776,
          "online_loss": 0.0037592435255646706,
          "train_loss": 0.003746788017451763,
          "validation_loss": 0.0031495022121816874,
          "train_metric": 0.0003773627686314285,
          "validation_metric": 0.00031720634433440864,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.304847408999194
        },
        {
          "epoch": 575,
          "updates": 575,
          "samples_seen": 588800,
          "online_loss": 0.003746788017451763,
          "train_loss": 0.0037343865260481834,
          "validation_loss": 0.0031402281019836664,
          "train_metric": 0.00037611371953971684,
          "validation_metric": 0.0003162722277920693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.311954869999681
        },
        {
          "epoch": 576,
          "updates": 576,
          "samples_seen": 589824,
          "online_loss": 0.00373438629321754,
          "train_loss": 0.0037220423109829426,
          "validation_loss": 0.0031275185756385326,
          "train_metric": 0.0003748704621102661,
          "validation_metric": 0.0003149922122247517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.31764271099928
        },
        {
          "epoch": 577,
          "updates": 577,
          "samples_seen": 590848,
          "online_loss": 0.0037220423109829426,
          "train_loss": 0.003709758399054408,
          "validation_loss": 0.0031190321315079927,
          "train_metric": 0.0003736332873813808,
          "validation_metric": 0.00031413754913955927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.323848837000696
        },
        {
          "epoch": 578,
          "updates": 578,
          "samples_seen": 591872,
          "online_loss": 0.0037097581662237644,
          "train_loss": 0.0036975585389882326,
          "validation_loss": 0.0031055740546435118,
          "train_metric": 0.0003724046400748193,
          "validation_metric": 0.00031278200913220644,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.329737920999833
        },
        {
          "epoch": 579,
          "updates": 579,
          "samples_seen": 592896,
          "online_loss": 0.0036975585389882326,
          "train_loss": 0.0036854431964457035,
          "validation_loss": 0.0030983760952949524,
          "train_metric": 0.00037118434556759894,
          "validation_metric": 0.0003120570909231901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.33546524600024
        },
        {
          "epoch": 580,
          "updates": 580,
          "samples_seen": 593920,
          "online_loss": 0.0036854431964457035,
          "train_loss": 0.003673458704724908,
          "validation_loss": 0.0030836230143904686,
          "train_metric": 0.0003699773515108973,
          "validation_metric": 0.00031057122396305203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.347769772999527
        },
        {
          "epoch": 581,
          "updates": 581,
          "samples_seen": 594944,
          "online_loss": 0.0036734589375555515,
          "train_loss": 0.003661653958261013,
          "validation_loss": 0.003078693291172385,
          "train_metric": 0.00036878837272524834,
          "validation_metric": 0.0003100747417192906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.353645753000819
        },
        {
          "epoch": 582,
          "updates": 582,
          "samples_seen": 595968,
          "online_loss": 0.0036616537254303694,
          "train_loss": 0.0036501481663435698,
          "validation_loss": 0.003061759052798152,
          "train_metric": 0.0003676296037156135,
          "validation_metric": 0.0003083691990468651,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.363720938999904
        },
        {
          "epoch": 583,
          "updates": 583,
          "samples_seen": 596992,
          "online_loss": 0.0036501483991742134,
          "train_loss": 0.003639108268544078,
          "validation_loss": 0.0030610456597059965,
          "train_metric": 0.00036651763366535306,
          "validation_metric": 0.000308297312585637,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.369547304999287
        },
        {
          "epoch": 584,
          "updates": 584,
          "samples_seen": 598016,
          "online_loss": 0.003639107570052147,
          "train_loss": 0.0036289121489971876,
          "validation_loss": 0.0030406001023948193,
          "train_metric": 0.0003654907050076872,
          "validation_metric": 0.0003062380710616708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.385884848999922
        },
        {
          "epoch": 585,
          "updates": 585,
          "samples_seen": 599040,
          "online_loss": 0.0036289123818278313,
          "train_loss": 0.003620140952989459,
          "validation_loss": 0.0030483584851026535,
          "train_metric": 0.00036460734554566443,
          "validation_metric": 0.0003070195089094341,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.403804472000047
        },
        {
          "epoch": 586,
          "updates": 586,
          "samples_seen": 600064,
          "online_loss": 0.0036201411858201027,
          "train_loss": 0.0036140726879239082,
          "validation_loss": 0.0030230949632823467,
          "train_metric": 0.0003639961942099035,
          "validation_metric": 0.0003044750774279237,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.420975697999893
        },
        {
          "epoch": 587,
          "updates": 587,
          "samples_seen": 601088,
          "online_loss": 0.0036140726879239082,
          "train_loss": 0.0036126403138041496,
          "validation_loss": 0.0030493380036205053,
          "train_metric": 0.000363851897418499,
          "validation_metric": 0.00030711814179085195,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.583203132999188
        },
        {
          "epoch": 588,
          "updates": 588,
          "samples_seen": 602112,
          "online_loss": 0.003612640080973506,
          "train_loss": 0.0036200862377882004,
          "validation_loss": 0.003020742442458868,
          "train_metric": 0.00036460181581787765,
          "validation_metric": 0.0003042380849365145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.591440313999556
        },
        {
          "epoch": 589,
          "updates": 589,
          "samples_seen": 603136,
          "online_loss": 0.003620086470618844,
          "train_loss": 0.0036415564827620983,
          "validation_loss": 0.0030892975628376007,
          "train_metric": 0.00036676423042081296,
          "validation_metric": 0.00031114276498556137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.600121512999976
        },
        {
          "epoch": 590,
          "updates": 590,
          "samples_seen": 604160,
          "online_loss": 0.0036415564827620983,
          "train_loss": 0.0036880106199532747,
          "validation_loss": 0.003069227794185281,
          "train_metric": 0.00037144290399737656,
          "validation_metric": 0.0003091213875450194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.605260890999489
        },
        {
          "epoch": 591,
          "updates": 591,
          "samples_seen": 605184,
          "online_loss": 0.0036880101542919874,
          "train_loss": 0.0037599708884954453,
          "validation_loss": 0.0032183125149458647,
          "train_metric": 0.00037869051448069513,
          "validation_metric": 0.00032413663575425744,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.614843350000228
        },
        {
          "epoch": 592,
          "updates": 592,
          "samples_seen": 606208,
          "online_loss": 0.003759971121326089,
          "train_loss": 0.003858655458316207,
          "validation_loss": 0.0032087876461446285,
          "train_metric": 0.00038862970541231334,
          "validation_metric": 0.00032317734439857304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.619871676000912
        },
        {
          "epoch": 593,
          "updates": 593,
          "samples_seen": 607232,
          "online_loss": 0.003858655458316207,
          "train_loss": 0.003911585081368685,
          "validation_loss": 0.0033726005349308252,
          "train_metric": 0.00039396053762175143,
          "validation_metric": 0.00033967598574236035,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.626392247000695
        },
        {
          "epoch": 594,
          "updates": 594,
          "samples_seen": 608256,
          "online_loss": 0.003911585547029972,
          "train_loss": 0.0038732108660042286,
          "validation_loss": 0.003219278296455741,
          "train_metric": 0.00039009563624858856,
          "validation_metric": 0.0003242339298594743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.634825783999986
        },
        {
          "epoch": 595,
          "updates": 595,
          "samples_seen": 609280,
          "online_loss": 0.003873211331665516,
          "train_loss": 0.003695037914440036,
          "validation_loss": 0.003148344811052084,
          "train_metric": 0.00037215062184259295,
          "validation_metric": 0.00031708975438959897,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.6441479889999755
        },
        {
          "epoch": 596,
          "updates": 596,
          "samples_seen": 610304,
          "online_loss": 0.003695037681609392,
          "train_loss": 0.003524166066199541,
          "validation_loss": 0.0029306162614375353,
          "train_metric": 0.0003549411194398999,
          "validation_metric": 0.0002951609203591943,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.650406233000467
        },
        {
          "epoch": 597,
          "updates": 597,
          "samples_seen": 611328,
          "online_loss": 0.0035241662990301847,
          "train_loss": 0.003487570211291313,
          "validation_loss": 0.0029065930284559727,
          "train_metric": 0.0003512552648317069,
          "validation_metric": 0.0002927413734141737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.65650723500039
        },
        {
          "epoch": 598,
          "updates": 598,
          "samples_seen": 612352,
          "online_loss": 0.003487569745630026,
          "train_loss": 0.003572525456547737,
          "validation_loss": 0.0030251173302531242,
          "train_metric": 0.0003598116454668343,
          "validation_metric": 0.0003046787460334599,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.662369890000264
        },
        {
          "epoch": 599,
          "updates": 599,
          "samples_seen": 613376,
          "online_loss": 0.0035725256893783808,
          "train_loss": 0.0036467253230512142,
          "validation_loss": 0.003028715495020151,
          "train_metric": 0.000367284839740023,
          "validation_metric": 0.0003050411178264767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.667942131000018
        },
        {
          "epoch": 600,
          "updates": 600,
          "samples_seen": 614400,
          "online_loss": 0.003646725555881858,
          "train_loss": 0.0035987279843539,
          "validation_loss": 0.003058883361518383,
          "train_metric": 0.0003624506643973291,
          "validation_metric": 0.00030807952862232924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.68150328000047
        },
        {
          "epoch": 601,
          "updates": 601,
          "samples_seen": 615424,
          "online_loss": 0.0035987279843539,
          "train_loss": 0.0034853334072977304,
          "validation_loss": 0.002902831882238388,
          "train_metric": 0.0003510299720801413,
          "validation_metric": 0.00029236258706077933,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.687611474000732
        },
        {
          "epoch": 602,
          "updates": 602,
          "samples_seen": 616448,
          "online_loss": 0.0034853334072977304,
          "train_loss": 0.0034272228367626667,
          "validation_loss": 0.002867558505386114,
          "train_metric": 0.00034517733729444444,
          "validation_metric": 0.00028880994068458676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.701890644000741
        },
        {
          "epoch": 603,
          "updates": 603,
          "samples_seen": 617472,
          "online_loss": 0.0034272228367626667,
          "train_loss": 0.0034644335974007845,
          "validation_loss": 0.0029223323799669743,
          "train_metric": 0.00034892503754235804,
          "validation_metric": 0.00029432657174766064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.715818611999566
        },
        {
          "epoch": 604,
          "updates": 604,
          "samples_seen": 618496,
          "online_loss": 0.003464433364570141,
          "train_loss": 0.0035133399069309235,
          "validation_loss": 0.0029214073438197374,
          "train_metric": 0.00035385071532800794,
          "validation_metric": 0.00029423341038636863,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.721202882999933
        },
        {
          "epoch": 605,
          "updates": 605,
          "samples_seen": 619520,
          "online_loss": 0.003513339441269636,
          "train_loss": 0.003487464040517807,
          "validation_loss": 0.0029526406433433294,
          "train_metric": 0.0003512445546220988,
          "validation_metric": 0.0002973792143166065,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.735075811999195
        },
        {
          "epoch": 606,
          "updates": 606,
          "samples_seen": 620544,
          "online_loss": 0.0034874638076871634,
          "train_loss": 0.0034154029563069344,
          "validation_loss": 0.002845588605850935,
          "train_metric": 0.00034398684510961175,
          "validation_metric": 0.00028659726376645267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.749698630999774
        },
        {
          "epoch": 607,
          "updates": 607,
          "samples_seen": 621568,
          "online_loss": 0.003415402490645647,
          "train_loss": 0.003377495566383004,
          "validation_loss": 0.002824124414473772,
          "train_metric": 0.00034016897552646697,
          "validation_metric": 0.00028443546034395695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.75738490600088
        },
        {
          "epoch": 608,
          "updates": 608,
          "samples_seen": 622592,
          "online_loss": 0.003377495566383004,
          "train_loss": 0.003398420987650752,
          "validation_loss": 0.002861836925148964,
          "train_metric": 0.00034227644209749997,
          "validation_metric": 0.00028823368484154344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.769921176000935
        },
        {
          "epoch": 609,
          "updates": 609,
          "samples_seen": 623616,
          "online_loss": 0.0033984205219894648,
          "train_loss": 0.003422517329454422,
          "validation_loss": 0.0028473185375332832,
          "train_metric": 0.0003447034687269479,
          "validation_metric": 0.00028677150839939713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.78352038400044
        },
        {
          "epoch": 610,
          "updates": 610,
          "samples_seen": 624640,
          "online_loss": 0.0034225175622850657,
          "train_loss": 0.0033990335650742054,
          "validation_loss": 0.0028709087055176497,
          "train_metric": 0.0003423382295295596,
          "validation_metric": 0.0002891473995987326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.788858686000822
        },
        {
          "epoch": 611,
          "updates": 611,
          "samples_seen": 625664,
          "online_loss": 0.0033990335650742054,
          "train_loss": 0.0033514893148094416,
          "validation_loss": 0.0027974029071629047,
          "train_metric": 0.0003375497180968523,
          "validation_metric": 0.0002817442000377923,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.803582409000228
        },
        {
          "epoch": 612,
          "updates": 612,
          "samples_seen": 626688,
          "online_loss": 0.0033514893148094416,
          "train_loss": 0.003329487983137369,
          "validation_loss": 0.002786084543913603,
          "train_metric": 0.00033533383975736797,
          "validation_metric": 0.0002806042321026325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.817889048999859
        },
        {
          "epoch": 613,
          "updates": 613,
          "samples_seen": 627712,
          "online_loss": 0.0033294884487986565,
          "train_loss": 0.003341177012771368,
          "validation_loss": 0.0028144889511168003,
          "train_metric": 0.0003365111188031733,
          "validation_metric": 0.0002834650222212076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.83196930899976
        },
        {
          "epoch": 614,
          "updates": 614,
          "samples_seen": 628736,
          "online_loss": 0.003341177012771368,
          "train_loss": 0.0033496601972728968,
          "validation_loss": 0.002791411941871047,
          "train_metric": 0.0003373654617462307,
          "validation_metric": 0.00028114079032093287,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.8377437769995595
        },
        {
          "epoch": 615,
          "updates": 615,
          "samples_seen": 629760,
          "online_loss": 0.003349659964442253,
          "train_loss": 0.003328608348965645,
          "validation_loss": 0.002805985976010561,
          "train_metric": 0.0003352453059051186,
          "validation_metric": 0.00028260861290618777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.852594641999531
        },
        {
          "epoch": 616,
          "updates": 616,
          "samples_seen": 630784,
          "online_loss": 0.003328608814626932,
          "train_loss": 0.0032965256832540035,
          "validation_loss": 0.0027537343557924032,
          "train_metric": 0.0003320139949209988,
          "validation_metric": 0.00027734602917917073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.867410377999477
        },
        {
          "epoch": 617,
          "updates": 617,
          "samples_seen": 631808,
          "online_loss": 0.003296525916084647,
          "train_loss": 0.0032821670174598694,
          "validation_loss": 0.0027444963343441486,
          "train_metric": 0.00033056779648177326,
          "validation_metric": 0.00027641563792712986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.881925162999323
        },
        {
          "epoch": 618,
          "updates": 618,
          "samples_seen": 632832,
          "online_loss": 0.003282166551798582,
          "train_loss": 0.0032865293323993683,
          "validation_loss": 0.0027628103271126747,
          "train_metric": 0.0003310072352178395,
          "validation_metric": 0.00027826018049381673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.887682864000453
        },
        {
          "epoch": 619,
          "updates": 619,
          "samples_seen": 633856,
          "online_loss": 0.0032865297980606556,
          "train_loss": 0.003286824095994234,
          "validation_loss": 0.0027379803359508514,
          "train_metric": 0.0003310368920210749,
          "validation_metric": 0.00027575931744650006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.901857271999688
        },
        {
          "epoch": 620,
          "updates": 620,
          "samples_seen": 634880,
          "online_loss": 0.003286824095994234,
          "train_loss": 0.0032695611007511616,
          "validation_loss": 0.0027479699347168207,
          "train_metric": 0.0003292982582934201,
          "validation_metric": 0.0002767654659692198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.912042497000584
        },
        {
          "epoch": 621,
          "updates": 621,
          "samples_seen": 635904,
          "online_loss": 0.0032695611007511616,
          "train_loss": 0.003246844280511141,
          "validation_loss": 0.0027106935158371925,
          "train_metric": 0.0003270102897658944,
          "validation_metric": 0.0002730111009441316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.9212407390004955
        },
        {
          "epoch": 622,
          "updates": 622,
          "samples_seen": 636928,
          "online_loss": 0.003246844280511141,
          "train_loss": 0.003235048381611705,
          "validation_loss": 0.0027030433993786573,
          "train_metric": 0.0003258221840951592,
          "validation_metric": 0.00027224060613662004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.932850797000356
        },
        {
          "epoch": 623,
          "updates": 623,
          "samples_seen": 637952,
          "online_loss": 0.003235048381611705,
          "train_loss": 0.0032340805046260357,
          "validation_loss": 0.002713555470108986,
          "train_metric": 0.00032572474447079003,
          "validation_metric": 0.00027329937438480556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 6.942423590999169
        },
        {
          "epoch": 624,
          "updates": 624,
          "samples_seen": 638976,
          "online_loss": 0.0032340805046260357,
          "train_loss": 0.0032307037618011236,
          "validation_loss": 0.002691802103072405,
          "train_metric": 0.00032538463710807264,
          "validation_metric": 0.0002711084671318531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.063911061999534
        },
        {
          "epoch": 625,
          "updates": 625,
          "samples_seen": 640000,
          "online_loss": 0.00323070352897048,
          "train_loss": 0.003217107616364956,
          "validation_loss": 0.002699314383789897,
          "train_metric": 0.0003240153018850833,
          "validation_metric": 0.0002718650212045759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.071761268000046
        },
        {
          "epoch": 626,
          "updates": 626,
          "samples_seen": 641024,
          "online_loss": 0.003217107616364956,
          "train_loss": 0.0031998688355088234,
          "validation_loss": 0.002670454094186425,
          "train_metric": 0.0003222790255676955,
          "validation_metric": 0.000268958363449201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.08078450299945
        },
        {
          "epoch": 627,
          "updates": 627,
          "samples_seen": 642048,
          "online_loss": 0.0031998686026781797,
          "train_loss": 0.003188568400219083,
          "validation_loss": 0.002663020510226488,
          "train_metric": 0.00032114089117385447,
          "validation_metric": 0.0002682096674107015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.0896653960007825
        },
        {
          "epoch": 628,
          "updates": 628,
          "samples_seen": 643072,
          "online_loss": 0.003188568400219083,
          "train_loss": 0.0031840114388614893,
          "validation_loss": 0.0026662179734557867,
          "train_metric": 0.0003206819237675518,
          "validation_metric": 0.00026853173039853573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.097854519000975
        },
        {
          "epoch": 629,
          "updates": 629,
          "samples_seen": 644096,
          "online_loss": 0.0031840112060308456,
          "train_loss": 0.0031789750792086124,
          "validation_loss": 0.002647546585649252,
          "train_metric": 0.0003201747313141823,
          "validation_metric": 0.00026665121549740434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.106463530999463
        },
        {
          "epoch": 630,
          "updates": 630,
          "samples_seen": 645120,
          "online_loss": 0.003178975312039256,
          "train_loss": 0.0031681989785283804,
          "validation_loss": 0.002653310773894191,
          "train_metric": 0.0003190894203726202,
          "validation_metric": 0.0002672317496035248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.114519602000655
        },
        {
          "epoch": 631,
          "updates": 631,
          "samples_seen": 646144,
          "online_loss": 0.0031681987456977367,
          "train_loss": 0.0031543844379484653,
          "validation_loss": 0.0026294905692338943,
          "train_metric": 0.00031769805354997516,
          "validation_metric": 0.00026483266265131533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.123036712000612
        },
        {
          "epoch": 632,
          "updates": 632,
          "samples_seen": 647168,
          "online_loss": 0.0031543842051178217,
          "train_loss": 0.00314315059222281,
          "validation_loss": 0.0026236053090542555,
          "train_metric": 0.0003165666130371392,
          "validation_metric": 0.00026423990493640304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.132075290999637
        },
        {
          "epoch": 633,
          "updates": 633,
          "samples_seen": 648192,
          "online_loss": 0.00314315059222281,
          "train_loss": 0.0031361987348645926,
          "validation_loss": 0.0026218548882752657,
          "train_metric": 0.000315866491291672,
          "validation_metric": 0.0002640636230353266,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.14093372500065
        },
        {
          "epoch": 634,
          "updates": 634,
          "samples_seen": 649216,
          "online_loss": 0.003136198502033949,
          "train_loss": 0.0031301414128392935,
          "validation_loss": 0.002607309026643634,
          "train_metric": 0.00031525641679763794,
          "validation_metric": 0.000262598623521626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.1501675990002695
        },
        {
          "epoch": 635,
          "updates": 635,
          "samples_seen": 650240,
          "online_loss": 0.0031301414128392935,
          "train_loss": 0.00312129152007401,
          "validation_loss": 0.0026120455004274845,
          "train_metric": 0.00031436508288607,
          "validation_metric": 0.00026307563530281186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.158240445000047
        },
        {
          "epoch": 636,
          "updates": 636,
          "samples_seen": 651264,
          "online_loss": 0.0031212917529046535,
          "train_loss": 0.003109904006123543,
          "validation_loss": 0.002592142205685377,
          "train_metric": 0.0003132181882392615,
          "validation_metric": 0.00026107107987627387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.1666901700009475
        },
        {
          "epoch": 637,
          "updates": 637,
          "samples_seen": 652288,
          "online_loss": 0.003109904006123543,
          "train_loss": 0.0030989202205091715,
          "validation_loss": 0.002588172210380435,
          "train_metric": 0.0003121119225397706,
          "validation_metric": 0.00026067125145345926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.174642812000457
        },
        {
          "epoch": 638,
          "updates": 638,
          "samples_seen": 653312,
          "online_loss": 0.003098920453339815,
          "train_loss": 0.0030903471633791924,
          "validation_loss": 0.0025815207045525312,
          "train_metric": 0.0003112484118901193,
          "validation_metric": 0.00026000133948400617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.183130507000897
        },
        {
          "epoch": 639,
          "updates": 639,
          "samples_seen": 654336,
          "online_loss": 0.0030903469305485487,
          "train_loss": 0.003083297749981284,
          "validation_loss": 0.0025698330719023943,
          "train_metric": 0.00031053851125761867,
          "validation_metric": 0.0002588241477496922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.1916106330008915
        },
        {
          "epoch": 640,
          "updates": 640,
          "samples_seen": 655360,
          "online_loss": 0.003083297982811928,
          "train_loss": 0.0030755074694752693,
          "validation_loss": 0.002571850549429655,
          "train_metric": 0.0003097538719885051,
          "validation_metric": 0.0002590273506939411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.199797409000894
        },
        {
          "epoch": 641,
          "updates": 641,
          "samples_seen": 656384,
          "online_loss": 0.0030755074694752693,
          "train_loss": 0.003066010307520628,
          "validation_loss": 0.0025545372627675533,
          "train_metric": 0.0003087973454967141,
          "validation_metric": 0.00025728365289978683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.208476499999961
        },
        {
          "epoch": 642,
          "updates": 642,
          "samples_seen": 657408,
          "online_loss": 0.003066010307520628,
          "train_loss": 0.003055762266740203,
          "validation_loss": 0.002552042482420802,
          "train_metric": 0.00030776511994190514,
          "validation_metric": 0.0002570323704276234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.217314875999364
        },
        {
          "epoch": 643,
          "updates": 643,
          "samples_seen": 658432,
          "online_loss": 0.003055761568248272,
          "train_loss": 0.0030463419388979673,
          "validation_loss": 0.002541285241022706,
          "train_metric": 0.0003068163641728461,
          "validation_metric": 0.00025594892213121057,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.2263974380002765
        },
        {
          "epoch": 644,
          "updates": 644,
          "samples_seen": 659456,
          "online_loss": 0.0030463417060673237,
          "train_loss": 0.0030382219702005386,
          "validation_loss": 0.0025324192829430103,
          "train_metric": 0.00030599854653701186,
          "validation_metric": 0.0002550559875089675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.235561023999253
        },
        {
          "epoch": 645,
          "updates": 645,
          "samples_seen": 660480,
          "online_loss": 0.0030382215045392513,
          "train_loss": 0.00303055951371789,
          "validation_loss": 0.0025310779456049204,
          "train_metric": 0.0003052268293686211,
          "validation_metric": 0.00025492088752798736,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.244497559999218
        },
        {
          "epoch": 646,
          "updates": 646,
          "samples_seen": 661504,
          "online_loss": 0.0030305590480566025,
          "train_loss": 0.0030222998466342688,
          "validation_loss": 0.0025171502493321896,
          "train_metric": 0.00030439498368650675,
          "validation_metric": 0.00025351817021146417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.253411824000068
        },
        {
          "epoch": 647,
          "updates": 647,
          "samples_seen": 662528,
          "online_loss": 0.003022299613803625,
          "train_loss": 0.0030131672974675894,
          "validation_loss": 0.0025158433709293604,
          "train_metric": 0.0003034752153325826,
          "validation_metric": 0.00025338653358630836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.262258016000487
        },
        {
          "epoch": 648,
          "updates": 648,
          "samples_seen": 663552,
          "online_loss": 0.0030131677631288767,
          "train_loss": 0.0030038116965442896,
          "validation_loss": 0.0025031487457454205,
          "train_metric": 0.00030253289151005447,
          "validation_metric": 0.0002521079732105136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.270729444999233
        },
        {
          "epoch": 649,
          "updates": 649,
          "samples_seen": 664576,
          "online_loss": 0.0030038116965442896,
          "train_loss": 0.002994965761899948,
          "validation_loss": 0.0024970537051558495,
          "train_metric": 0.0003016419359482825,
          "validation_metric": 0.0002514941443223506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.2793039879998105
        },
        {
          "epoch": 650,
          "updates": 650,
          "samples_seen": 665600,
          "online_loss": 0.002994965761899948,
          "train_loss": 0.00298678339459002,
          "validation_loss": 0.0024913963861763477,
          "train_metric": 0.0003008178900927305,
          "validation_metric": 0.00025092432042583823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.287783326999488
        },
        {
          "epoch": 651,
          "updates": 651,
          "samples_seen": 666624,
          "online_loss": 0.0029867831617593765,
          "train_loss": 0.0029788503888994455,
          "validation_loss": 0.0024806468281894922,
          "train_metric": 0.0003000188444275409,
          "validation_metric": 0.00024984165793284774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.296467202999338
        },
        {
          "epoch": 652,
          "updates": 652,
          "samples_seen": 667648,
          "online_loss": 0.0029788503888994455,
          "train_loss": 0.002970633329823613,
          "validation_loss": 0.002478470327332616,
          "train_metric": 0.0002991913352161646,
          "validation_metric": 0.0002496224478818476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.304894790000617
        },
        {
          "epoch": 653,
          "updates": 653,
          "samples_seen": 668672,
          "online_loss": 0.0029706330969929695,
          "train_loss": 0.0029619906563311815,
          "validation_loss": 0.002465866506099701,
          "train_metric": 0.0002983208396472037,
          "validation_metric": 0.0002483530552126467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.313926504000847
        },
        {
          "epoch": 654,
          "updates": 654,
          "samples_seen": 669696,
          "online_loss": 0.0029619906563311815,
          "train_loss": 0.002953178249299526,
          "validation_loss": 0.0024618576280772686,
          "train_metric": 0.0002974332601297647,
          "validation_metric": 0.00024794926866889,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.322484154999984
        },
        {
          "epoch": 655,
          "updates": 655,
          "samples_seen": 670720,
          "online_loss": 0.002953177783638239,
          "train_loss": 0.002944566309452057,
          "validation_loss": 0.0024525457993149757,
          "train_metric": 0.00029656587867066264,
          "validation_metric": 0.00024701145594008267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.331207805000304
        },
        {
          "epoch": 656,
          "updates": 656,
          "samples_seen": 671744,
          "online_loss": 0.0029445660766214132,
          "train_loss": 0.0029363136272877455,
          "validation_loss": 0.0024449904449284077,
          "train_metric": 0.00029573473148047924,
          "validation_metric": 0.00024625047808513045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.339514171999326
        },
        {
          "epoch": 657,
          "updates": 657,
          "samples_seen": 672768,
          "online_loss": 0.002936313161626458,
          "train_loss": 0.002928286325186491,
          "validation_loss": 0.002440272131934762,
          "train_metric": 0.00029492625617422163,
          "validation_metric": 0.00024577524163760245,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.3484608809994825
        },
        {
          "epoch": 658,
          "updates": 658,
          "samples_seen": 673792,
          "online_loss": 0.002928286325186491,
          "train_loss": 0.0029202208388596773,
          "validation_loss": 0.0024299006909132004,
          "train_metric": 0.0002941139100585133,
          "validation_metric": 0.0002447307051625103,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.357040045000758
        },
        {
          "epoch": 659,
          "updates": 659,
          "samples_seen": 674816,
          "online_loss": 0.0029202206060290337,
          "train_loss": 0.002911968156695366,
          "validation_loss": 0.0024265770334750414,
          "train_metric": 0.00029328279197216034,
          "validation_metric": 0.00024439592380076647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.365970913999263
        },
        {
          "epoch": 660,
          "updates": 660,
          "samples_seen": 675840,
          "online_loss": 0.002911968156695366,
          "train_loss": 0.002903567161411047,
          "validation_loss": 0.002415952505543828,
          "train_metric": 0.0002924366854131222,
          "validation_metric": 0.0002433258487144485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.374280129000908
        },
        {
          "epoch": 661,
          "updates": 661,
          "samples_seen": 676864,
          "online_loss": 0.002903567161411047,
          "train_loss": 0.0028951719868928194,
          "validation_loss": 0.0024109238293021917,
          "train_metric": 0.00029159116093069315,
          "validation_metric": 0.00024281944206450135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.383236931000283
        },
        {
          "epoch": 662,
          "updates": 662,
          "samples_seen": 677888,
          "online_loss": 0.002895172219723463,
          "train_loss": 0.0028869244270026684,
          "validation_loss": 0.0024028944317251444,
          "train_metric": 0.0002907604502979666,
          "validation_metric": 0.00024201073392760009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.501026183999784
        },
        {
          "epoch": 663,
          "updates": 663,
          "samples_seen": 678912,
          "online_loss": 0.0028869241941720247,
          "train_loss": 0.002878854749724269,
          "validation_loss": 0.0023953146301209927,
          "train_metric": 0.0002899477258324623,
          "validation_metric": 0.00024124732590280473,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.510348566000175
        },
        {
          "epoch": 664,
          "updates": 664,
          "samples_seen": 679936,
          "online_loss": 0.0028788545168936253,
          "train_loss": 0.002870870055630803,
          "validation_loss": 0.002390118781477213,
          "train_metric": 0.0002891434996854514,
          "validation_metric": 0.00024072402447927743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.519249135000791
        },
        {
          "epoch": 665,
          "updates": 665,
          "samples_seen": 680960,
          "online_loss": 0.0028708698228001595,
          "train_loss": 0.0028628837317228317,
          "validation_loss": 0.0023806714452803135,
          "train_metric": 0.00028833915712311864,
          "validation_metric": 0.00023977248929440975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.5280079700005444
        },
        {
          "epoch": 666,
          "updates": 666,
          "samples_seen": 681984,
          "online_loss": 0.002862883498892188,
          "train_loss": 0.0028548231348395348,
          "validation_loss": 0.0023764220532029867,
          "train_metric": 0.00028752730577252805,
          "validation_metric": 0.0002393445320194587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.536565465999956
        },
        {
          "epoch": 667,
          "updates": 667,
          "samples_seen": 683008,
          "online_loss": 0.002854822902008891,
          "train_loss": 0.002846704563125968,
          "validation_loss": 0.0023669281508773565,
          "train_metric": 0.00028670969186350703,
          "validation_metric": 0.00023838834022171795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.5451845890002005
        },
        {
          "epoch": 668,
          "updates": 668,
          "samples_seen": 684032,
          "online_loss": 0.0028467047959566116,
          "train_loss": 0.002838592976331711,
          "validation_loss": 0.0023617565166205168,
          "train_metric": 0.00028589266003109515,
          "validation_metric": 0.0002378674689680338,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.553754055999889
        },
        {
          "epoch": 669,
          "updates": 669,
          "samples_seen": 685056,
          "online_loss": 0.002838592743501067,
          "train_loss": 0.002830546349287033,
          "validation_loss": 0.002353872638195753,
          "train_metric": 0.00028508229297585785,
          "validation_metric": 0.00023707344371359795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.562805692999973
        },
        {
          "epoch": 670,
          "updates": 670,
          "samples_seen": 686080,
          "online_loss": 0.0028305465821176767,
          "train_loss": 0.0028225909918546677,
          "validation_loss": 0.002347041619941592,
          "train_metric": 0.00028428100631572306,
          "validation_metric": 0.00023638541460968554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.57170122900061
        },
        {
          "epoch": 671,
          "updates": 671,
          "samples_seen": 687104,
          "online_loss": 0.0028225905261933804,
          "train_loss": 0.0028147075790911913,
          "validation_loss": 0.002341107465326786,
          "train_metric": 0.0002834870247170329,
          "validation_metric": 0.00023578778200317174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.580581771999277
        },
        {
          "epoch": 672,
          "updates": 672,
          "samples_seen": 688128,
          "online_loss": 0.0028147068805992603,
          "train_loss": 0.00280685699544847,
          "validation_loss": 0.0023328554816544056,
          "train_metric": 0.0002826963609550148,
          "validation_metric": 0.00023495667846873403,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.589241425999717
        },
        {
          "epoch": 673,
          "updates": 673,
          "samples_seen": 689152,
          "online_loss": 0.0028068572282791138,
          "train_loss": 0.0027990019880235195,
          "validation_loss": 0.002327911090105772,
          "train_metric": 0.00028190523153170943,
          "validation_metric": 0.00023445869737770408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.5980172350009525
        },
        {
          "epoch": 674,
          "updates": 674,
          "samples_seen": 690176,
          "online_loss": 0.0027990019880235195,
          "train_loss": 0.002791136037558317,
          "validation_loss": 0.002319183899089694,
          "train_metric": 0.0002811129961628467,
          "validation_metric": 0.0002335797034902498,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.606895376000466
        },
        {
          "epoch": 675,
          "updates": 675,
          "samples_seen": 691200,
          "online_loss": 0.002791136037558317,
          "train_loss": 0.0027832617051899433,
          "validation_loss": 0.0023140828125178814,
          "train_metric": 0.00028031988767907023,
          "validation_metric": 0.00023306594812311232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.615902348999953
        },
        {
          "epoch": 676,
          "updates": 676,
          "samples_seen": 692224,
          "online_loss": 0.0027832614723592997,
          "train_loss": 0.002775403670966625,
          "validation_loss": 0.0023059470113366842,
          "train_metric": 0.00027952849632129073,
          "validation_metric": 0.00023224652977660298,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.6251341459992545
        },
        {
          "epoch": 677,
          "updates": 677,
          "samples_seen": 693248,
          "online_loss": 0.002775403670966625,
          "train_loss": 0.002767584752291441,
          "validation_loss": 0.0022999749053269625,
          "train_metric": 0.000278740975772962,
          "validation_metric": 0.00023164502636063844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.633756712000832
        },
        {
          "epoch": 678,
          "updates": 678,
          "samples_seen": 694272,
          "online_loss": 0.002767584752291441,
          "train_loss": 0.002759812632575631,
          "validation_loss": 0.0022930274717509747,
          "train_metric": 0.00027795819914899766,
          "validation_metric": 0.00023094535572454333,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.642659735000052
        },
        {
          "epoch": 679,
          "updates": 679,
          "samples_seen": 695296,
          "online_loss": 0.0027598123997449875,
          "train_loss": 0.002752084517851472,
          "validation_loss": 0.0022860663011670113,
          "train_metric": 0.0002771798463072628,
          "validation_metric": 0.0002302442298969254,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.651388162999865
        },
        {
          "epoch": 680,
          "updates": 680,
          "samples_seen": 696320,
          "online_loss": 0.0027520840521901846,
          "train_loss": 0.0027443936560302973,
          "validation_loss": 0.002280209679156542,
          "train_metric": 0.0002764053060673177,
          "validation_metric": 0.000229654397116974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.660077693000858
        },
        {
          "epoch": 681,
          "updates": 681,
          "samples_seen": 697344,
          "online_loss": 0.0027443936560302973,
          "train_loss": 0.0027367225848138332,
          "validation_loss": 0.0022725313901901245,
          "train_metric": 0.00027563265757635236,
          "validation_metric": 0.00022888103558216244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.668863920000149
        },
        {
          "epoch": 682,
          "updates": 682,
          "samples_seen": 698368,
          "online_loss": 0.0027367225848138332,
          "train_loss": 0.0027290652506053448,
          "validation_loss": 0.002267227042466402,
          "train_metric": 0.00027486143517307937,
          "validation_metric": 0.0002283467911183834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.67770949899932
        },
        {
          "epoch": 683,
          "updates": 683,
          "samples_seen": 699392,
          "online_loss": 0.0027290647849440575,
          "train_loss": 0.0027214172296226025,
          "validation_loss": 0.0022593624889850616,
          "train_metric": 0.0002740912023000419,
          "validation_metric": 0.00022755471582058817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.686362206000922
        },
        {
          "epoch": 684,
          "updates": 684,
          "samples_seen": 700416,
          "online_loss": 0.0027214176952838898,
          "train_loss": 0.002713784808292985,
          "validation_loss": 0.002254027407616377,
          "train_metric": 0.00027332251193001866,
          "validation_metric": 0.00022701735724695027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.695172417999856
        },
        {
          "epoch": 685,
          "updates": 685,
          "samples_seen": 701440,
          "online_loss": 0.0027137852739542723,
          "train_loss": 0.0027061726432293653,
          "validation_loss": 0.002246437594294548,
          "train_metric": 0.00027255582972429693,
          "validation_metric": 0.00022625297424383461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.70355723800094
        },
        {
          "epoch": 686,
          "updates": 686,
          "samples_seen": 702464,
          "online_loss": 0.002706172876060009,
          "train_loss": 0.002698575844988227,
          "validation_loss": 0.0022407013457268476,
          "train_metric": 0.0002717906900215894,
          "validation_metric": 0.00022567524865735322,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.712929581000935
        },
        {
          "epoch": 687,
          "updates": 687,
          "samples_seen": 703488,
          "online_loss": 0.002698575844988227,
          "train_loss": 0.0026910132728517056,
          "validation_loss": 0.0022336887195706367,
          "train_metric": 0.0002710290136747062,
          "validation_metric": 0.00022496892779599875,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.722130891999768
        },
        {
          "epoch": 688,
          "updates": 688,
          "samples_seen": 704512,
          "online_loss": 0.0026910132728517056,
          "train_loss": 0.0026834821328520775,
          "validation_loss": 0.002227416029199958,
          "train_metric": 0.00027027048054151237,
          "validation_metric": 0.0002243372000521049,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.731240633000198
        },
        {
          "epoch": 689,
          "updates": 689,
          "samples_seen": 705536,
          "online_loss": 0.002683481900021434,
          "train_loss": 0.0026759712491184473,
          "validation_loss": 0.002221041126176715,
          "train_metric": 0.0002695140428841114,
          "validation_metric": 0.0002236951404483989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.739760788999774
        },
        {
          "epoch": 690,
          "updates": 690,
          "samples_seen": 706560,
          "online_loss": 0.0026759712491184473,
          "train_loss": 0.0026684931945055723,
          "validation_loss": 0.0022142911329865456,
          "train_metric": 0.0002687608648557216,
          "validation_metric": 0.00022301528952084482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.749014196999269
        },
        {
          "epoch": 691,
          "updates": 691,
          "samples_seen": 707584,
          "online_loss": 0.002668493427336216,
          "train_loss": 0.002661038190126419,
          "validation_loss": 0.0022084324155002832,
          "train_metric": 0.0002680100151337683,
          "validation_metric": 0.00022242520935833454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.757875289000367
        },
        {
          "epoch": 692,
          "updates": 692,
          "samples_seen": 708608,
          "online_loss": 0.0026610384229570627,
          "train_loss": 0.0026536115910857916,
          "validation_loss": 0.00220136228017509,
          "train_metric": 0.0002672620757948607,
          "validation_metric": 0.00022171314049046487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.766879033000805
        },
        {
          "epoch": 693,
          "updates": 693,
          "samples_seen": 709632,
          "online_loss": 0.0026536118239164352,
          "train_loss": 0.0026461980305612087,
          "validation_loss": 0.002195818116888404,
          "train_metric": 0.0002665153588168323,
          "validation_metric": 0.0002211547689512372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.776239066999551
        },
        {
          "epoch": 694,
          "updates": 694,
          "samples_seen": 710656,
          "online_loss": 0.002646197797730565,
          "train_loss": 0.0026388056576251984,
          "validation_loss": 0.002188607584685087,
          "train_metric": 0.00026577082462608814,
          "validation_metric": 0.00022042857017368078,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.785422918999757
        },
        {
          "epoch": 695,
          "updates": 695,
          "samples_seen": 711680,
          "online_loss": 0.0026388056576251984,
          "train_loss": 0.0026314305141568184,
          "validation_loss": 0.0021831782069057226,
          "train_metric": 0.0002650280366651714,
          "validation_metric": 0.00021988169464748353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.7979292220006755
        },
        {
          "epoch": 696,
          "updates": 696,
          "samples_seen": 712704,
          "online_loss": 0.0026314302813261747,
          "train_loss": 0.00262407842092216,
          "validation_loss": 0.0021759881637990475,
          "train_metric": 0.0002642875479068607,
          "validation_metric": 0.00021915756224188954,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.814035441000669
        },
        {
          "epoch": 697,
          "updates": 697,
          "samples_seen": 713728,
          "online_loss": 0.00262407842092216,
          "train_loss": 0.0026167503092437983,
          "validation_loss": 0.0021705336403101683,
          "train_metric": 0.0002635495038703084,
          "validation_metric": 0.00021860819833818823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.825658920000933
        },
        {
          "epoch": 698,
          "updates": 698,
          "samples_seen": 714752,
          "online_loss": 0.0026167503092437983,
          "train_loss": 0.002609442686662078,
          "validation_loss": 0.0021634772419929504,
          "train_metric": 0.00026281349710188806,
          "validation_metric": 0.00021789749735035002,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.841846478999287
        },
        {
          "epoch": 699,
          "updates": 699,
          "samples_seen": 715776,
          "online_loss": 0.002609442686662078,
          "train_loss": 0.0026021513622254133,
          "validation_loss": 0.0021579281892627478,
          "train_metric": 0.00026207914925180376,
          "validation_metric": 0.00021733860194217414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 7.997828621999361
        },
        {
          "epoch": 700,
          "updates": 700,
          "samples_seen": 716800,
          "online_loss": 0.0026021513622254133,
          "train_loss": 0.0025948891416192055,
          "validation_loss": 0.002151066903024912,
          "train_metric": 0.0002613477408885956,
          "validation_metric": 0.0002166475896956399,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.01527444400017
        },
        {
          "epoch": 701,
          "updates": 701,
          "samples_seen": 717824,
          "online_loss": 0.002594889374449849,
          "train_loss": 0.0025876453146338463,
          "validation_loss": 0.0021454079542309046,
          "train_metric": 0.00026061810785904527,
          "validation_metric": 0.00021607766393572092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.034071006999511
        },
        {
          "epoch": 702,
          "updates": 702,
          "samples_seen": 718848,
          "online_loss": 0.002587644848972559,
          "train_loss": 0.0025804201140999794,
          "validation_loss": 0.002138744806870818,
          "train_metric": 0.000259890453889966,
          "validation_metric": 0.00021540657326113433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.053432542999872
        },
        {
          "epoch": 703,
          "updates": 703,
          "samples_seen": 719872,
          "online_loss": 0.0025804201140999794,
          "train_loss": 0.002573218196630478,
          "validation_loss": 0.0021329934243112803,
          "train_metric": 0.00025916515733115375,
          "validation_metric": 0.00021482727606780827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.075507163999646
        },
        {
          "epoch": 704,
          "updates": 704,
          "samples_seen": 720896,
          "online_loss": 0.002573218196630478,
          "train_loss": 0.002566048176959157,
          "validation_loss": 0.0021265060640871525,
          "train_metric": 0.0002584429457783699,
          "validation_metric": 0.00021417390962596983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.09451549500045
        },
        {
          "epoch": 705,
          "updates": 705,
          "samples_seen": 721920,
          "online_loss": 0.002566048176959157,
          "train_loss": 0.0025588858406990767,
          "validation_loss": 0.0021206813398748636,
          "train_metric": 0.00025772160734049976,
          "validation_metric": 0.00021358724916353822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.116508123999665
        },
        {
          "epoch": 706,
          "updates": 706,
          "samples_seen": 722944,
          "online_loss": 0.002558886306360364,
          "train_loss": 0.002551754005253315,
          "validation_loss": 0.002114326460286975,
          "train_metric": 0.00025700332480482757,
          "validation_metric": 0.0002129472268279642,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.135748233000413
        },
        {
          "epoch": 707,
          "updates": 707,
          "samples_seen": 723968,
          "online_loss": 0.002551754005253315,
          "train_loss": 0.0025446435902267694,
          "validation_loss": 0.0021084656473249197,
          "train_metric": 0.0002562871668487787,
          "validation_metric": 0.00021235694293864071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.155029472000024
        },
        {
          "epoch": 708,
          "updates": 708,
          "samples_seen": 724992,
          "online_loss": 0.002544643823057413,
          "train_loss": 0.0025375494733452797,
          "validation_loss": 0.0021022032015025616,
          "train_metric": 0.00025557269691489637,
          "validation_metric": 0.0002117262192768976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.176704257000893
        },
        {
          "epoch": 709,
          "updates": 709,
          "samples_seen": 726016,
          "online_loss": 0.0025375494733452797,
          "train_loss": 0.002530478872358799,
          "validation_loss": 0.002096335869282484,
          "train_metric": 0.0002548605261836201,
          "validation_metric": 0.0002111352951033041,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.195866789999855
        },
        {
          "epoch": 710,
          "updates": 710,
          "samples_seen": 727040,
          "online_loss": 0.0025304784066975117,
          "train_loss": 0.002523430623114109,
          "validation_loss": 0.002090145368129015,
          "train_metric": 0.0002541507128626108,
          "validation_metric": 0.00021051177463959903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.215314585999295
        },
        {
          "epoch": 711,
          "updates": 711,
          "samples_seen": 728064,
          "online_loss": 0.0025234303902834654,
          "train_loss": 0.002516401233151555,
          "validation_loss": 0.0020843043457716703,
          "train_metric": 0.0002534426748752594,
          "validation_metric": 0.00020992349891457707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.237363352000102
        },
        {
          "epoch": 712,
          "updates": 712,
          "samples_seen": 729088,
          "online_loss": 0.002516400534659624,
          "train_loss": 0.0025093951262533665,
          "validation_loss": 0.002078161807730794,
          "train_metric": 0.00025273708160966635,
          "validation_metric": 0.00020930486789438874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.256092891999288
        },
        {
          "epoch": 713,
          "updates": 713,
          "samples_seen": 730112,
          "online_loss": 0.00250939535908401,
          "train_loss": 0.0025024060159921646,
          "validation_loss": 0.0020723785273730755,
          "train_metric": 0.0002520331763662398,
          "validation_metric": 0.0002087223547277972,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.274564979999923
        },
        {
          "epoch": 714,
          "updates": 714,
          "samples_seen": 731136,
          "online_loss": 0.002502405783161521,
          "train_loss": 0.0024954439140856266,
          "validation_loss": 0.002066245535388589,
          "train_metric": 0.0002513319777790457,
          "validation_metric": 0.00020810468413401395,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.29690372100049
        },
        {
          "epoch": 715,
          "updates": 715,
          "samples_seen": 732160,
          "online_loss": 0.002495443681254983,
          "train_loss": 0.002488494850695133,
          "validation_loss": 0.0020605502650141716,
          "train_metric": 0.0002506321179680526,
          "validation_metric": 0.00020753107673954219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.318808770000032
        },
        {
          "epoch": 716,
          "updates": 716,
          "samples_seen": 733184,
          "online_loss": 0.0024884946178644896,
          "train_loss": 0.0024815723299980164,
          "validation_loss": 0.002054387703537941,
          "train_metric": 0.00024993487750180066,
          "validation_metric": 0.0002069104230031371,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.339076689000649
        },
        {
          "epoch": 717,
          "updates": 717,
          "samples_seen": 734208,
          "online_loss": 0.0024815723299980164,
          "train_loss": 0.0024746686685830355,
          "validation_loss": 0.0020488195586949587,
          "train_metric": 0.00024923955788835883,
          "validation_metric": 0.00020634959219023585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.358431353000924
        },
        {
          "epoch": 718,
          "updates": 718,
          "samples_seen": 735232,
          "online_loss": 0.002474668435752392,
          "train_loss": 0.0024677873589098454,
          "validation_loss": 0.002042568987235427,
          "train_metric": 0.00024854656658135355,
          "validation_metric": 0.00020572006178554147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.37680427899977
        },
        {
          "epoch": 719,
          "updates": 719,
          "samples_seen": 736256,
          "online_loss": 0.0024677878245711327,
          "train_loss": 0.0024609228130429983,
          "validation_loss": 0.0020371966529637575,
          "train_metric": 0.00024785511777736247,
          "validation_metric": 0.00020517899247352034,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.395360458000141
        },
        {
          "epoch": 720,
          "updates": 720,
          "samples_seen": 737280,
          "online_loss": 0.0024609225802123547,
          "train_loss": 0.0024540862068533897,
          "validation_loss": 0.0020307719241827726,
          "train_metric": 0.00024716657935641706,
          "validation_metric": 0.00020453189790714532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.417210612000417
        },
        {
          "epoch": 721,
          "updates": 721,
          "samples_seen": 738304,
          "online_loss": 0.0024540862068533897,
          "train_loss": 0.0024472717195749283,
          "validation_loss": 0.0020257234573364258,
          "train_metric": 0.00024648019461892545,
          "validation_metric": 0.00020402345398906618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.435832219000076
        },
        {
          "epoch": 722,
          "updates": 722,
          "samples_seen": 739328,
          "online_loss": 0.0024472714867442846,
          "train_loss": 0.0024404835421591997,
          "validation_loss": 0.002018971834331751,
          "train_metric": 0.00024579657474532723,
          "validation_metric": 0.00020334345754235983,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.454823546000625
        },
        {
          "epoch": 723,
          "updates": 723,
          "samples_seen": 740352,
          "online_loss": 0.0024404835421591997,
          "train_loss": 0.002433729125186801,
          "validation_loss": 0.002014464931562543,
          "train_metric": 0.0002451162727084011,
          "validation_metric": 0.0002028895542025566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.476893876000759
        },
        {
          "epoch": 724,
          "updates": 724,
          "samples_seen": 741376,
          "online_loss": 0.002433729125186801,
          "train_loss": 0.002427014522254467,
          "validation_loss": 0.0020071291364729404,
          "train_metric": 0.0002444400161039084,
          "validation_metric": 0.00020215066615492105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.495364799000527
        },
        {
          "epoch": 725,
          "updates": 725,
          "samples_seen": 742400,
          "online_loss": 0.002427014522254467,
          "train_loss": 0.0024203620851039886,
          "validation_loss": 0.0020035975612699986,
          "train_metric": 0.00024377000227104872,
          "validation_metric": 0.0002017949882429093,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.51431553799921
        },
        {
          "epoch": 726,
          "updates": 726,
          "samples_seen": 743424,
          "online_loss": 0.002420361852273345,
          "train_loss": 0.002413797425106168,
          "validation_loss": 0.0019952126313000917,
          "train_metric": 0.00024310882145073265,
          "validation_metric": 0.00020095052605029196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.533109885000158
        },
        {
          "epoch": 727,
          "updates": 727,
          "samples_seen": 744448,
          "online_loss": 0.002413797425106168,
          "train_loss": 0.0024073803797364235,
          "validation_loss": 0.0019934980664402246,
          "train_metric": 0.00024246249813586473,
          "validation_metric": 0.00020077783847227693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.55499375799991
        },
        {
          "epoch": 728,
          "updates": 728,
          "samples_seen": 745472,
          "online_loss": 0.00240738014690578,
          "train_loss": 0.00240123737603426,
          "validation_loss": 0.0019832919351756573,
          "train_metric": 0.00024184383801184595,
          "validation_metric": 0.00019974992028437555,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.564313961000153
        },
        {
          "epoch": 729,
          "updates": 729,
          "samples_seen": 746496,
          "online_loss": 0.00240123737603426,
          "train_loss": 0.0023955644574016333,
          "validation_loss": 0.0019852675031870604,
          "train_metric": 0.00024127248616423458,
          "validation_metric": 0.00019994885951746255,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.57276537799953
        },
        {
          "epoch": 730,
          "updates": 730,
          "samples_seen": 747520,
          "online_loss": 0.002395564690232277,
          "train_loss": 0.002390815643593669,
          "validation_loss": 0.001972116529941559,
          "train_metric": 0.0002407941792625934,
          "validation_metric": 0.00019862432964146137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.581425981999928
        },
        {
          "epoch": 731,
          "updates": 731,
          "samples_seen": 748544,
          "online_loss": 0.0023908154107630253,
          "train_loss": 0.0023877681232988834,
          "validation_loss": 0.0019822714384645224,
          "train_metric": 0.0002404873084742576,
          "validation_metric": 0.00019964712555520236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.589570444000856
        },
        {
          "epoch": 732,
          "updates": 732,
          "samples_seen": 749568,
          "online_loss": 0.0023877685889601707,
          "train_loss": 0.002388162538409233,
          "validation_loss": 0.0019655784126371145,
          "train_metric": 0.00024052700609900057,
          "validation_metric": 0.00019796588458120823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.598493360999782
        },
        {
          "epoch": 733,
          "updates": 733,
          "samples_seen": 750592,
          "online_loss": 0.0023881623055785894,
          "train_loss": 0.0023948373273015022,
          "validation_loss": 0.001995818456634879,
          "train_metric": 0.00024119926092680544,
          "validation_metric": 0.00020101151312701404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.606218749000618
        },
        {
          "epoch": 734,
          "updates": 734,
          "samples_seen": 751616,
          "online_loss": 0.0023948373273015022,
          "train_loss": 0.0024146016221493483,
          "validation_loss": 0.0019805103074759245,
          "train_metric": 0.00024318981741089374,
          "validation_metric": 0.00019946976681239903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.614174098000149
        },
        {
          "epoch": 735,
          "updates": 735,
          "samples_seen": 752640,
          "online_loss": 0.0024146013893187046,
          "train_loss": 0.002456055721268058,
          "validation_loss": 0.002063907217234373,
          "train_metric": 0.0002473649219609797,
          "validation_metric": 0.00020786917593795806,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.62187643300058
        },
        {
          "epoch": 736,
          "updates": 736,
          "samples_seen": 753664,
          "online_loss": 0.0024560559540987015,
          "train_loss": 0.002541623078286648,
          "validation_loss": 0.0020775741431862116,
          "train_metric": 0.00025598297361284494,
          "validation_metric": 0.00020924564159940928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.63009286399938
        },
        {
          "epoch": 737,
          "updates": 737,
          "samples_seen": 754688,
          "online_loss": 0.002541622845456004,
          "train_loss": 0.0026741998735815287,
          "validation_loss": 0.0022811945527791977,
          "train_metric": 0.0002693356364034116,
          "validation_metric": 0.00022975358297117054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.647341926000081
        },
        {
          "epoch": 738,
          "updates": 738,
          "samples_seen": 755712,
          "online_loss": 0.0026742001064121723,
          "train_loss": 0.002877634484320879,
          "validation_loss": 0.0023554887156933546,
          "train_metric": 0.00028982479125261307,
          "validation_metric": 0.00023723616322968155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.66535617199952
        },
        {
          "epoch": 739,
          "updates": 739,
          "samples_seen": 756736,
          "online_loss": 0.0028776342514902353,
          "train_loss": 0.0029937271028757095,
          "validation_loss": 0.0025871985126286745,
          "train_metric": 0.0003015171969309449,
          "validation_metric": 0.0002605731424409896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.783852029000627
        },
        {
          "epoch": 740,
          "updates": 740,
          "samples_seen": 757760,
          "online_loss": 0.0029937271028757095,
          "train_loss": 0.0029627580661326647,
          "validation_loss": 0.0024269591085612774,
          "train_metric": 0.0002983981685247272,
          "validation_metric": 0.0002444344281684607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.792761357000927
        },
        {
          "epoch": 741,
          "updates": 741,
          "samples_seen": 758784,
          "online_loss": 0.0029627582989633083,
          "train_loss": 0.002630867762491107,
          "validation_loss": 0.002231044229120016,
          "train_metric": 0.0002649713715072721,
          "validation_metric": 0.0002247026131954044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.80055524800082
        },
        {
          "epoch": 742,
          "updates": 742,
          "samples_seen": 759808,
          "online_loss": 0.0026308679953217506,
          "train_loss": 0.002347789239138365,
          "validation_loss": 0.0019152581226080656,
          "train_metric": 0.0002364607498748228,
          "validation_metric": 0.0001928977872012183,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.809332127000744
        },
        {
          "epoch": 743,
          "updates": 743,
          "samples_seen": 760832,
          "online_loss": 0.0023477894719690084,
          "train_loss": 0.002351701958104968,
          "validation_loss": 0.001919968519359827,
          "train_metric": 0.0002368547866353765,
          "validation_metric": 0.00019337217963766307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.817118667999239
        },
        {
          "epoch": 744,
          "updates": 744,
          "samples_seen": 761856,
          "online_loss": 0.0023517014924436808,
          "train_loss": 0.0025465884245932102,
          "validation_loss": 0.0021500452421605587,
          "train_metric": 0.00025648303562775254,
          "validation_metric": 0.00021654469310306013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.828037594999842
        },
        {
          "epoch": 745,
          "updates": 745,
          "samples_seen": 762880,
          "online_loss": 0.002546587958931923,
          "train_loss": 0.0026390175335109234,
          "validation_loss": 0.002156046452000737,
          "train_metric": 0.0002657921868376434,
          "validation_metric": 0.00021714906324632466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.845555919000617
        },
        {
          "epoch": 746,
          "updates": 746,
          "samples_seen": 763904,
          "online_loss": 0.002639017766341567,
          "train_loss": 0.00246888748370111,
          "validation_loss": 0.00207709614187479,
          "train_metric": 0.0002486573066562414,
          "validation_metric": 0.00020919756207149476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.863408773999254
        },
        {
          "epoch": 747,
          "updates": 747,
          "samples_seen": 764928,
          "online_loss": 0.0024688872508704662,
          "train_loss": 0.002296244027093053,
          "validation_loss": 0.0018859796691685915,
          "train_metric": 0.000231269295909442,
          "validation_metric": 0.00018994897254742682,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.881921379999767
        },
        {
          "epoch": 748,
          "updates": 748,
          "samples_seen": 765952,
          "online_loss": 0.00229624449275434,
          "train_loss": 0.00232383469119668,
          "validation_loss": 0.0019044283544644713,
          "train_metric": 0.00023404810053762048,
          "validation_metric": 0.0001918070629471913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.899930476000009
        },
        {
          "epoch": 749,
          "updates": 749,
          "samples_seen": 766976,
          "online_loss": 0.00232383469119668,
          "train_loss": 0.002446477534249425,
          "validation_loss": 0.0020592696964740753,
          "train_metric": 0.0002464002463966608,
          "validation_metric": 0.00020740210311487317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.91340444600064
        },
        {
          "epoch": 750,
          "updates": 750,
          "samples_seen": 768000,
          "online_loss": 0.002446477534249425,
          "train_loss": 0.0024518724530935287,
          "validation_loss": 0.002001617569476366,
          "train_metric": 0.000246943614911288,
          "validation_metric": 0.000201595583348535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.921380677000343
        },
        {
          "epoch": 751,
          "updates": 751,
          "samples_seen": 769024,
          "online_loss": 0.0024518724530935287,
          "train_loss": 0.0023197766859084368,
          "validation_loss": 0.0019342756131663918,
          "train_metric": 0.0002336394536541775,
          "validation_metric": 0.00019481315393932164,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.929275313999824
        },
        {
          "epoch": 752,
          "updates": 752,
          "samples_seen": 770048,
          "online_loss": 0.0023197769187390804,
          "train_loss": 0.0022563813254237175,
          "validation_loss": 0.0018630600534379482,
          "train_metric": 0.0002272544807055965,
          "validation_metric": 0.0001876405585790053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.937806822999846
        },
        {
          "epoch": 753,
          "updates": 753,
          "samples_seen": 771072,
          "online_loss": 0.002256381558254361,
          "train_loss": 0.002320318017154932,
          "validation_loss": 0.0018978975713253021,
          "train_metric": 0.00023369397968053818,
          "validation_metric": 0.0001911493018269539,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.946224102999622
        },
        {
          "epoch": 754,
          "updates": 754,
          "samples_seen": 772096,
          "online_loss": 0.0023203182499855757,
          "train_loss": 0.002367723500356078,
          "validation_loss": 0.001992973266169429,
          "train_metric": 0.00023846841941121966,
          "validation_metric": 0.00020072497136425227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.953644282999448
        },
        {
          "epoch": 755,
          "updates": 755,
          "samples_seen": 773120,
          "online_loss": 0.002367723500356078,
          "train_loss": 0.0023133475333452225,
          "validation_loss": 0.0018946671625599265,
          "train_metric": 0.00023299189342651516,
          "validation_metric": 0.00019082397921010852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.961239890999423
        },
        {
          "epoch": 756,
          "updates": 756,
          "samples_seen": 774144,
          "online_loss": 0.0023133475333452225,
          "train_loss": 0.0022382810711860657,
          "validation_loss": 0.001857775030657649,
          "train_metric": 0.0002254314604215324,
          "validation_metric": 0.00018710830772761256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.969923475000542
        },
        {
          "epoch": 757,
          "updates": 757,
          "samples_seen": 775168,
          "online_loss": 0.0022382810711860657,
          "train_loss": 0.002247883239760995,
          "validation_loss": 0.0018730437150225043,
          "train_metric": 0.00022639863891527057,
          "validation_metric": 0.00018864609592128545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.978046813999754
        },
        {
          "epoch": 758,
          "updates": 758,
          "samples_seen": 776192,
          "online_loss": 0.0022478834725916386,
          "train_loss": 0.0022956328466534615,
          "validation_loss": 0.0018810940673574805,
          "train_metric": 0.0002312076831003651,
          "validation_metric": 0.00018945691408589482,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.986039367000558
        },
        {
          "epoch": 759,
          "updates": 759,
          "samples_seen": 777216,
          "online_loss": 0.002295632613822818,
          "train_loss": 0.002282695146277547,
          "validation_loss": 0.0019148197025060654,
          "train_metric": 0.0002299046900589019,
          "validation_metric": 0.00019285362213850021,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 8.993790333999641
        },
        {
          "epoch": 760,
          "updates": 760,
          "samples_seen": 778240,
          "online_loss": 0.002282695146277547,
          "train_loss": 0.0022275166120380163,
          "validation_loss": 0.00183297588955611,
          "train_metric": 0.00022434732818510383,
          "validation_metric": 0.00018461060244590044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.002272615000038
        },
        {
          "epoch": 761,
          "updates": 761,
          "samples_seen": 779264,
          "online_loss": 0.0022275163792073727,
          "train_loss": 0.00220823148265481,
          "validation_loss": 0.0018216631142422557,
          "train_metric": 0.00022240501129999757,
          "validation_metric": 0.00018347124569118023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.011230493999392
        },
        {
          "epoch": 762,
          "updates": 762,
          "samples_seen": 780288,
          "online_loss": 0.0022082317154854536,
          "train_loss": 0.0022348864004015923,
          "validation_loss": 0.001864796970039606,
          "train_metric": 0.00022508957772515714,
          "validation_metric": 0.00018781551625579596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.022654712000076
        },
        {
          "epoch": 763,
          "updates": 763,
          "samples_seen": 781312,
          "online_loss": 0.0022348864004015923,
          "train_loss": 0.002245525363832712,
          "validation_loss": 0.0018394459038972855,
          "train_metric": 0.00022616110800299793,
          "validation_metric": 0.00018526225176174194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.042065150999406
        },
        {
          "epoch": 764,
          "updates": 764,
          "samples_seen": 782336,
          "online_loss": 0.002245525363832712,
          "train_loss": 0.0022132464218884706,
          "validation_loss": 0.001842936617322266,
          "train_metric": 0.0002229100646218285,
          "validation_metric": 0.00018561385513748974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.058873113999653
        },
        {
          "epoch": 765,
          "updates": 765,
          "samples_seen": 783360,
          "online_loss": 0.0022132464218884706,
          "train_loss": 0.002184735145419836,
          "validation_loss": 0.0018032246734946966,
          "train_metric": 0.00022003853518981487,
          "validation_metric": 0.0001816142030293122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.079697318000399
        },
        {
          "epoch": 766,
          "updates": 766,
          "samples_seen": 784384,
          "online_loss": 0.002184735145419836,
          "train_loss": 0.0021909568458795547,
          "validation_loss": 0.001801345031708479,
          "train_metric": 0.00022066512610763311,
          "validation_metric": 0.00018142488261219114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.098657233000267
        },
        {
          "epoch": 767,
          "updates": 767,
          "samples_seen": 785408,
          "online_loss": 0.0021909568458795547,
          "train_loss": 0.0022045436780899763,
          "validation_loss": 0.0018370581092312932,
          "train_metric": 0.00022203358821570873,
          "validation_metric": 0.00018502175225876272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.11484699399989
        },
        {
          "epoch": 768,
          "updates": 768,
          "samples_seen": 786432,
          "online_loss": 0.00220454391092062,
          "train_loss": 0.0021932143718004227,
          "validation_loss": 0.0017999508418142796,
          "train_metric": 0.0002208925288869068,
          "validation_metric": 0.00018128447118215263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.12423490800029
        },
        {
          "epoch": 769,
          "updates": 769,
          "samples_seen": 787456,
          "online_loss": 0.0021932143718004227,
          "train_loss": 0.0021678293123841286,
          "validation_loss": 0.0017953438218683004,
          "train_metric": 0.00021833585924468935,
          "validation_metric": 0.0001808204542612657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.133747560999836
        },
        {
          "epoch": 770,
          "updates": 770,
          "samples_seen": 788480,
          "online_loss": 0.0021678295452147722,
          "train_loss": 0.002159571973606944,
          "validation_loss": 0.001785097410902381,
          "train_metric": 0.0002175041736336425,
          "validation_metric": 0.00017978844698518515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.143571444999907
        },
        {
          "epoch": 771,
          "updates": 771,
          "samples_seen": 789504,
          "online_loss": 0.002159571973606944,
          "train_loss": 0.002167666796594858,
          "validation_loss": 0.0017783239018172026,
          "train_metric": 0.00021831944468431175,
          "validation_metric": 0.00017910626775119454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.15303538400076
        },
        {
          "epoch": 772,
          "updates": 772,
          "samples_seen": 790528,
          "online_loss": 0.002167666796594858,
          "train_loss": 0.002167004393413663,
          "validation_loss": 0.001799362013116479,
          "train_metric": 0.00021825275325682014,
          "validation_metric": 0.0001812251575756818,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.161420165999516
        },
        {
          "epoch": 773,
          "updates": 773,
          "samples_seen": 791552,
          "online_loss": 0.0021670046262443066,
          "train_loss": 0.0021510948427021503,
          "validation_loss": 0.0017659107688814402,
          "train_metric": 0.0002166504127671942,
          "validation_metric": 0.00017785608361009508,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.170225634999952
        },
        {
          "epoch": 774,
          "updates": 774,
          "samples_seen": 792576,
          "online_loss": 0.0021510948427021503,
          "train_loss": 0.0021370293106883764,
          "validation_loss": 0.0017621475271880627,
          "train_metric": 0.0002152337838197127,
          "validation_metric": 0.00017747703532222658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.179374318000555
        },
        {
          "epoch": 775,
          "updates": 775,
          "samples_seen": 793600,
          "online_loss": 0.0021370293106883764,
          "train_loss": 0.0021364078857004642,
          "validation_loss": 0.001766844536177814,
          "train_metric": 0.00021517118148040026,
          "validation_metric": 0.0001779501180863008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.187732206999499
        },
        {
          "epoch": 776,
          "updates": 776,
          "samples_seen": 794624,
          "online_loss": 0.0021364078857004642,
          "train_loss": 0.00213887938298285,
          "validation_loss": 0.0017553196521475911,
          "train_metric": 0.00021542007743846625,
          "validation_metric": 0.00017678935546427965,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.29725892100032
        },
        {
          "epoch": 777,
          "updates": 777,
          "samples_seen": 795648,
          "online_loss": 0.00213887938298285,
          "train_loss": 0.002131534041836858,
          "validation_loss": 0.001767205074429512,
          "train_metric": 0.00021468030172400177,
          "validation_metric": 0.0001779863960109651,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.306116715999451
        },
        {
          "epoch": 778,
          "updates": 778,
          "samples_seen": 796672,
          "online_loss": 0.002131534041836858,
          "train_loss": 0.002118580276146531,
          "validation_loss": 0.0017440433148294687,
          "train_metric": 0.0002133756352122873,
          "validation_metric": 0.00017565365124028176,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.318181323000317
        },
        {
          "epoch": 779,
          "updates": 779,
          "samples_seen": 797696,
          "online_loss": 0.002118580276146531,
          "train_loss": 0.0021113199181854725,
          "validation_loss": 0.0017408095300197601,
          "train_metric": 0.0002126444160239771,
          "validation_metric": 0.00017532795027364045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.335723483000038
        },
        {
          "epoch": 780,
          "updates": 780,
          "samples_seen": 798720,
          "online_loss": 0.002111319685354829,
          "train_loss": 0.002110970439389348,
          "validation_loss": 0.0017486284486949444,
          "train_metric": 0.00021260918583720922,
          "validation_metric": 0.00017611545627005398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.344156285000281
        },
        {
          "epoch": 781,
          "updates": 781,
          "samples_seen": 799744,
          "online_loss": 0.0021109702065587044,
          "train_loss": 0.0021089038345962763,
          "validation_loss": 0.0017337518511340022,
          "train_metric": 0.0002124011080013588,
          "validation_metric": 0.0001746171183185652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.353059013999882
        },
        {
          "epoch": 782,
          "updates": 782,
          "samples_seen": 800768,
          "online_loss": 0.0021089043002575636,
          "train_loss": 0.0021005196031183004,
          "validation_loss": 0.0017398967174813151,
          "train_metric": 0.000211556616704911,
          "validation_metric": 0.0001752360403770581,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.361268246999316
        },
        {
          "epoch": 783,
          "updates": 783,
          "samples_seen": 801792,
          "online_loss": 0.0021005196031183004,
          "train_loss": 0.0020910874009132385,
          "validation_loss": 0.0017237127758562565,
          "train_metric": 0.00021060669678263366,
          "validation_metric": 0.0001736060221446678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.369647773999532
        },
        {
          "epoch": 784,
          "updates": 784,
          "samples_seen": 802816,
          "online_loss": 0.0020910874009132385,
          "train_loss": 0.002086225664243102,
          "validation_loss": 0.0017193530220538378,
          "train_metric": 0.00021011702483519912,
          "validation_metric": 0.00017316693265456706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.378370225000253
        },
        {
          "epoch": 785,
          "updates": 785,
          "samples_seen": 803840,
          "online_loss": 0.002086225664243102,
          "train_loss": 0.002084341598674655,
          "validation_loss": 0.001725854817777872,
          "train_metric": 0.00020992725330870599,
          "validation_metric": 0.0001738217833917588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.386042432000977
        },
        {
          "epoch": 786,
          "updates": 786,
          "samples_seen": 804864,
          "online_loss": 0.0020843413658440113,
          "train_loss": 0.0020802246872335672,
          "validation_loss": 0.001711079035885632,
          "train_metric": 0.00020951262558810413,
          "validation_metric": 0.00017233361722901464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.394811979000224
        },
        {
          "epoch": 787,
          "updates": 787,
          "samples_seen": 805888,
          "online_loss": 0.0020802244544029236,
          "train_loss": 0.002072654664516449,
          "validation_loss": 0.0017149490304291248,
          "train_metric": 0.00020875017798971385,
          "validation_metric": 0.00017272336117457598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.40287246799926
        },
        {
          "epoch": 788,
          "updates": 788,
          "samples_seen": 806912,
          "online_loss": 0.0020726544316858053,
          "train_loss": 0.0020653584506362677,
          "validation_loss": 0.0017037229845300317,
          "train_metric": 0.000208015349926427,
          "validation_metric": 0.00017159272101707757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.411398427000677
        },
        {
          "epoch": 789,
          "updates": 789,
          "samples_seen": 807936,
          "online_loss": 0.0020653586834669113,
          "train_loss": 0.0020609635394066572,
          "validation_loss": 0.0016987515846267343,
          "train_metric": 0.00020757270976901054,
          "validation_metric": 0.00017109203326981515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.419903265999892
        },
        {
          "epoch": 790,
          "updates": 790,
          "samples_seen": 808960,
          "online_loss": 0.002060963772237301,
          "train_loss": 0.0020578522235155106,
          "validation_loss": 0.0017028446309268475,
          "train_metric": 0.00020725933427456766,
          "validation_metric": 0.00017150427447631955,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.427975484000854
        },
        {
          "epoch": 791,
          "updates": 791,
          "samples_seen": 809984,
          "online_loss": 0.0020578522235155106,
          "train_loss": 0.0020531376358121634,
          "validation_loss": 0.0016896129818633199,
          "train_metric": 0.00020678451983258128,
          "validation_metric": 0.00017017161007970572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.436493459999838
        },
        {
          "epoch": 792,
          "updates": 792,
          "samples_seen": 811008,
          "online_loss": 0.0020531376358121634,
          "train_loss": 0.00204654224216938,
          "validation_loss": 0.0016915706219151616,
          "train_metric": 0.00020612019579857588,
          "validation_metric": 0.00017036878853105009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.444544878999295
        },
        {
          "epoch": 793,
          "updates": 793,
          "samples_seen": 812032,
          "online_loss": 0.002046541776508093,
          "train_loss": 0.0020402944646775723,
          "validation_loss": 0.001682313159108162,
          "train_metric": 0.00020549098553601652,
          "validation_metric": 0.00016943641821853817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.453047332999631
        },
        {
          "epoch": 794,
          "updates": 794,
          "samples_seen": 813056,
          "online_loss": 0.0020402944646775723,
          "train_loss": 0.0020357570610940456,
          "validation_loss": 0.0016770155634731054,
          "train_metric": 0.0002050339971901849,
          "validation_metric": 0.00016890285769477487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.46213693000027
        },
        {
          "epoch": 795,
          "updates": 795,
          "samples_seen": 814080,
          "online_loss": 0.0020357570610940456,
          "train_loss": 0.0020319349132478237,
          "validation_loss": 0.0016790175577625632,
          "train_metric": 0.00020464904082473367,
          "validation_metric": 0.00016910448903217912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.470068093000009
        },
        {
          "epoch": 796,
          "updates": 796,
          "samples_seen": 815104,
          "online_loss": 0.0020319349132478237,
          "train_loss": 0.0020271972753107548,
          "validation_loss": 0.0016679448308423162,
          "train_metric": 0.00020417191262822598,
          "validation_metric": 0.00016798931756056845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.478562525999223
        },
        {
          "epoch": 797,
          "updates": 797,
          "samples_seen": 816128,
          "online_loss": 0.002027197740972042,
          "train_loss": 0.0020213474053889513,
          "validation_loss": 0.0016691996715962887,
          "train_metric": 0.00020358272013254464,
          "validation_metric": 0.00016811567184049636,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.486483702999976
        },
        {
          "epoch": 798,
          "updates": 798,
          "samples_seen": 817152,
          "online_loss": 0.002021347638219595,
          "train_loss": 0.002015611855313182,
          "validation_loss": 0.001661033253185451,
          "train_metric": 0.00020300506730563939,
          "validation_metric": 0.0001672932121437043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.495484933000625
        },
        {
          "epoch": 799,
          "updates": 799,
          "samples_seen": 818176,
          "online_loss": 0.0020156120881438255,
          "train_loss": 0.0020108481403440237,
          "validation_loss": 0.0016564502147957683,
          "train_metric": 0.00020252526155672967,
          "validation_metric": 0.00016683159628883004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.504302748000555
        },
        {
          "epoch": 800,
          "updates": 800,
          "samples_seen": 819200,
          "online_loss": 0.0020108481403440237,
          "train_loss": 0.0020066166762262583,
          "validation_loss": 0.0016566759441047907,
          "train_metric": 0.00020209906506352127,
          "validation_metric": 0.00016685432638041675,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.512395106999975
        },
        {
          "epoch": 801,
          "updates": 801,
          "samples_seen": 820224,
          "online_loss": 0.0020066166762262583,
          "train_loss": 0.0020019905641674995,
          "validation_loss": 0.0016472716815769672,
          "train_metric": 0.00020163315639365464,
          "validation_metric": 0.00016590715677011758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.520819299999857
        },
        {
          "epoch": 802,
          "updates": 802,
          "samples_seen": 821248,
          "online_loss": 0.0020019905641674995,
          "train_loss": 0.0019966762047261,
          "validation_loss": 0.001647784374654293,
          "train_metric": 0.00020109790784772485,
          "validation_metric": 0.00016595880151726305,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.52886534100071
        },
        {
          "epoch": 803,
          "updates": 803,
          "samples_seen": 822272,
          "online_loss": 0.0019966759718954563,
          "train_loss": 0.001991256373003125,
          "validation_loss": 0.0016398554434999824,
          "train_metric": 0.00020055205095559359,
          "validation_metric": 0.00016516023606527597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.537671694000892
        },
        {
          "epoch": 804,
          "updates": 804,
          "samples_seen": 823296,
          "online_loss": 0.001991256373003125,
          "train_loss": 0.0019863289780914783,
          "validation_loss": 0.0016360882436856627,
          "train_metric": 0.00020005580154247582,
          "validation_metric": 0.00016478082397952676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.546800522999547
        },
        {
          "epoch": 805,
          "updates": 805,
          "samples_seen": 824320,
          "online_loss": 0.0019863289780914783,
          "train_loss": 0.001981841865926981,
          "validation_loss": 0.0016345713520422578,
          "train_metric": 0.00019960383360739797,
          "validation_metric": 0.0001646280288696289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.555663523000476
        },
        {
          "epoch": 806,
          "updates": 806,
          "samples_seen": 825344,
          "online_loss": 0.001981841865926981,
          "train_loss": 0.001977294683456421,
          "validation_loss": 0.0016267902683466673,
          "train_metric": 0.00019914587028324604,
          "validation_metric": 0.00016384436457883567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.564481465000426
        },
        {
          "epoch": 807,
          "updates": 807,
          "samples_seen": 826368,
          "online_loss": 0.0019772944506257772,
          "train_loss": 0.001972357276827097,
          "validation_loss": 0.0016269448678940535,
          "train_metric": 0.00019864861678797752,
          "validation_metric": 0.00016385993512813002,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.572349734000454
        },
        {
          "epoch": 808,
          "updates": 808,
          "samples_seen": 827392,
          "online_loss": 0.001972357276827097,
          "train_loss": 0.001967211253941059,
          "validation_loss": 0.001619435497559607,
          "train_metric": 0.00019813029211945832,
          "validation_metric": 0.0001631036284379661,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.581210656000621
        },
        {
          "epoch": 809,
          "updates": 809,
          "samples_seen": 828416,
          "online_loss": 0.0019672110211104155,
          "train_loss": 0.0019622305408120155,
          "validation_loss": 0.0016166383866220713,
          "train_metric": 0.00019762867304962128,
          "validation_metric": 0.00016282190335914493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.589808063999953
        },
        {
          "epoch": 810,
          "updates": 810,
          "samples_seen": 829440,
          "online_loss": 0.001962230773642659,
          "train_loss": 0.001957556465640664,
          "validation_loss": 0.0016134785255417228,
          "train_metric": 0.00019715793314389884,
          "validation_metric": 0.00016250363842118531,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.59879473699948
        },
        {
          "epoch": 811,
          "updates": 811,
          "samples_seen": 830464,
          "online_loss": 0.0019575566984713078,
          "train_loss": 0.0019529986893758178,
          "validation_loss": 0.0016071602003648877,
          "train_metric": 0.00019669887842610478,
          "validation_metric": 0.00016186729772016406,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.607834831000218
        },
        {
          "epoch": 812,
          "updates": 812,
          "samples_seen": 831488,
          "online_loss": 0.0019529988057911396,
          "train_loss": 0.001948312041349709,
          "validation_loss": 0.001606492791324854,
          "train_metric": 0.00019622687250375748,
          "validation_metric": 0.000161800067871809,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.616885897999964
        },
        {
          "epoch": 813,
          "updates": 813,
          "samples_seen": 832512,
          "online_loss": 0.0019483119249343872,
          "train_loss": 0.0019434270216152072,
          "validation_loss": 0.0015992196276783943,
          "train_metric": 0.00019573484314605594,
          "validation_metric": 0.00016106753901112825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.62571734499943
        },
        {
          "epoch": 814,
          "updates": 814,
          "samples_seen": 833536,
          "online_loss": 0.001943427138030529,
          "train_loss": 0.001938509289175272,
          "validation_loss": 0.0015970608219504356,
          "train_metric": 0.00019523956871125847,
          "validation_metric": 0.000160850104293786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.634672868000052
        },
        {
          "epoch": 815,
          "updates": 815,
          "samples_seen": 834560,
          "online_loss": 0.001938509289175272,
          "train_loss": 0.0019337271805852652,
          "validation_loss": 0.0015924288891255856,
          "train_metric": 0.00019475790031719953,
          "validation_metric": 0.00016038361354731023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.743145377000474
        },
        {
          "epoch": 816,
          "updates": 816,
          "samples_seen": 835584,
          "online_loss": 0.0019337271805852652,
          "train_loss": 0.0019291032804176211,
          "validation_loss": 0.0015875728568062186,
          "train_metric": 0.00019429220992606133,
          "validation_metric": 0.00015989449457265437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.752877131999412
        },
        {
          "epoch": 817,
          "updates": 817,
          "samples_seen": 836608,
          "online_loss": 0.0019291035132482648,
          "train_loss": 0.0019245152361690998,
          "validation_loss": 0.0015857962425798178,
          "train_metric": 0.00019383011385798454,
          "validation_metric": 0.00015971560787875205,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.762034721999953
        },
        {
          "epoch": 818,
          "updates": 818,
          "samples_seen": 837632,
          "online_loss": 0.0019245152361690998,
          "train_loss": 0.0019198409281671047,
          "validation_loss": 0.0015793171478435397,
          "train_metric": 0.00019335933029651642,
          "validation_metric": 0.00015906304179225117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.771138823000001
        },
        {
          "epoch": 819,
          "updates": 819,
          "samples_seen": 838656,
          "online_loss": 0.001919840695336461,
          "train_loss": 0.0019150805892422795,
          "validation_loss": 0.0015776373911648989,
          "train_metric": 0.00019287988834548742,
          "validation_metric": 0.00015889386122580618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.78021448400068
        },
        {
          "epoch": 820,
          "updates": 820,
          "samples_seen": 839680,
          "online_loss": 0.0019150803564116359,
          "train_loss": 0.0019103168742731214,
          "validation_loss": 0.001572156441397965,
          "train_metric": 0.00019240014080423862,
          "validation_metric": 0.00015834187797736377,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.78896692499984
        },
        {
          "epoch": 821,
          "updates": 821,
          "samples_seen": 840704,
          "online_loss": 0.0019103168742731214,
          "train_loss": 0.0019056309247389436,
          "validation_loss": 0.0015686271945014596,
          "train_metric": 0.000191928178537637,
          "validation_metric": 0.00015798638924025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.797916147999786
        },
        {
          "epoch": 822,
          "updates": 822,
          "samples_seen": 841728,
          "online_loss": 0.001905630575492978,
          "train_loss": 0.001901027513667941,
          "validation_loss": 0.001565442536957562,
          "train_metric": 0.00019146452541463077,
          "validation_metric": 0.00015766563592478633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.80663809699945
        },
        {
          "epoch": 823,
          "updates": 823,
          "samples_seen": 842752,
          "online_loss": 0.001901027513667941,
          "train_loss": 0.0018964570481330156,
          "validation_loss": 0.0015600809128955007,
          "train_metric": 0.00019100421923212707,
          "validation_metric": 0.00015712562890257686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.81573093299994
        },
        {
          "epoch": 824,
          "updates": 824,
          "samples_seen": 843776,
          "online_loss": 0.0018964571645483375,
          "train_loss": 0.0018918494461104274,
          "validation_loss": 0.0015581182669848204,
          "train_metric": 0.00019054014410357922,
          "validation_metric": 0.00015692797023802996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.824202834000971
        },
        {
          "epoch": 825,
          "updates": 825,
          "samples_seen": 844800,
          "online_loss": 0.0018918492132797837,
          "train_loss": 0.0018872007494792342,
          "validation_loss": 0.0015523916808888316,
          "train_metric": 0.00019007195078302175,
          "validation_metric": 0.00015635120507795364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.833020183999906
        },
        {
          "epoch": 826,
          "updates": 826,
          "samples_seen": 845824,
          "online_loss": 0.001887200865894556,
          "train_loss": 0.0018825463484972715,
          "validation_loss": 0.0015497910790145397,
          "train_metric": 0.00018960317538585514,
          "validation_metric": 0.00015608927060384303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.841891953999948
        },
        {
          "epoch": 827,
          "updates": 827,
          "samples_seen": 846848,
          "online_loss": 0.0018825463484972715,
          "train_loss": 0.0018779307138174772,
          "validation_loss": 0.0015453535597771406,
          "train_metric": 0.0001891382853500545,
          "validation_metric": 0.0001556423376314342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.851070823000555
        },
        {
          "epoch": 828,
          "updates": 828,
          "samples_seen": 847872,
          "online_loss": 0.0018779307138174772,
          "train_loss": 0.0018733600154519081,
          "validation_loss": 0.0015413238434121013,
          "train_metric": 0.00018867795006372035,
          "validation_metric": 0.00015523649926763028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.860212387000502
        },
        {
          "epoch": 829,
          "updates": 829,
          "samples_seen": 848896,
          "online_loss": 0.0018733597826212645,
          "train_loss": 0.001868827035650611,
          "validation_loss": 0.001538446987979114,
          "train_metric": 0.0001882214128272608,
          "validation_metric": 0.0001549467269796878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.869032081000114
        },
        {
          "epoch": 830,
          "updates": 830,
          "samples_seen": 849920,
          "online_loss": 0.0018688271520659328,
          "train_loss": 0.0018642907962203026,
          "validation_loss": 0.001533318543806672,
          "train_metric": 0.000187764540896751,
          "validation_metric": 0.00015443023585248739,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.878097685999819
        },
        {
          "epoch": 831,
          "updates": 831,
          "samples_seen": 850944,
          "online_loss": 0.0018642907962203026,
          "train_loss": 0.0018597388407215476,
          "validation_loss": 0.0015309497248381376,
          "train_metric": 0.00018730606825556606,
          "validation_metric": 0.00015419162809848785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.887257784999747
        },
        {
          "epoch": 832,
          "updates": 832,
          "samples_seen": 851968,
          "online_loss": 0.0018597389571368694,
          "train_loss": 0.0018551788525655866,
          "validation_loss": 0.0015258743660524487,
          "train_metric": 0.0001868468534667045,
          "validation_metric": 0.000153680462972261,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.896881087999645
        },
        {
          "epoch": 833,
          "updates": 833,
          "samples_seen": 852992,
          "online_loss": 0.0018551789689809084,
          "train_loss": 0.0018506328342482448,
          "validation_loss": 0.001522877486422658,
          "train_metric": 0.0001863889629021287,
          "validation_metric": 0.00015337864169850945,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.905858227999488
        },
        {
          "epoch": 834,
          "updates": 834,
          "samples_seen": 854016,
          "online_loss": 0.001850632717832923,
          "train_loss": 0.0018461061408743262,
          "validation_loss": 0.0015187704702839255,
          "train_metric": 0.00018593305139802396,
          "validation_metric": 0.00015296497440431267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.91613518200029
        },
        {
          "epoch": 835,
          "updates": 835,
          "samples_seen": 855040,
          "online_loss": 0.0018461061408743262,
          "train_loss": 0.0018416109960526228,
          "validation_loss": 0.0015147425001487136,
          "train_metric": 0.00018548034131526947,
          "validation_metric": 0.0001525593106634915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.925771850999809
        },
        {
          "epoch": 836,
          "updates": 836,
          "samples_seen": 856064,
          "online_loss": 0.0018416109960526228,
          "train_loss": 0.0018371370388194919,
          "validation_loss": 0.0015116882277652621,
          "train_metric": 0.00018502972670830786,
          "validation_metric": 0.0001522516831755638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.935380704999261
        },
        {
          "epoch": 837,
          "updates": 837,
          "samples_seen": 857088,
          "online_loss": 0.00183713692240417,
          "train_loss": 0.0018326658755540848,
          "validation_loss": 0.0015069672372192144,
          "train_metric": 0.0001845793885877356,
          "validation_metric": 0.00015177621389739215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.944991427999412
        },
        {
          "epoch": 838,
          "updates": 838,
          "samples_seen": 858112,
          "online_loss": 0.0018326656427234411,
          "train_loss": 0.0018281916854903102,
          "validation_loss": 0.0015042824670672417,
          "train_metric": 0.00018412877398077399,
          "validation_metric": 0.00015150582476053387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.954957919999288
        },
        {
          "epoch": 839,
          "updates": 839,
          "samples_seen": 859136,
          "online_loss": 0.0018281915690749884,
          "train_loss": 0.0018237235490232706,
          "validation_loss": 0.0014995941892266273,
          "train_metric": 0.00018367877055425197,
          "validation_metric": 0.0001510336296632886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.964610577999338
        },
        {
          "epoch": 840,
          "updates": 840,
          "samples_seen": 860160,
          "online_loss": 0.0018237235490232706,
          "train_loss": 0.0018192582065239549,
          "validation_loss": 0.0014965733280405402,
          "train_metric": 0.00018322901451028883,
          "validation_metric": 0.0001507293782196939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.974019437999232
        },
        {
          "epoch": 841,
          "updates": 841,
          "samples_seen": 861184,
          "online_loss": 0.0018192583229392767,
          "train_loss": 0.0018148155650123954,
          "validation_loss": 0.001492482260800898,
          "train_metric": 0.000182781572220847,
          "validation_metric": 0.0001503173407400027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.983352065999497
        },
        {
          "epoch": 842,
          "updates": 842,
          "samples_seen": 862208,
          "online_loss": 0.0018148155650123954,
          "train_loss": 0.0018103846814483404,
          "validation_loss": 0.0014887700090184808,
          "train_metric": 0.00018233535229228437,
          "validation_metric": 0.00014994345838204026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 9.99293700400085
        },
        {
          "epoch": 843,
          "updates": 843,
          "samples_seen": 863232,
          "online_loss": 0.0018103847978636622,
          "train_loss": 0.0018059771973639727,
          "validation_loss": 0.0014854050241410732,
          "train_metric": 0.00018189141701441258,
          "validation_metric": 0.00014960455882828683,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.00181317500028
        },
        {
          "epoch": 844,
          "updates": 844,
          "samples_seen": 864256,
          "online_loss": 0.0018059773137792945,
          "train_loss": 0.0018015719251707196,
          "validation_loss": 0.001481144572608173,
          "train_metric": 0.00018144775822293013,
          "validation_metric": 0.00014917543740011752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.01135695999983
        },
        {
          "epoch": 845,
          "updates": 845,
          "samples_seen": 865280,
          "online_loss": 0.0018015720415860415,
          "train_loss": 0.0017971808556467295,
          "validation_loss": 0.0014781851787120104,
          "train_metric": 0.00018100548186339438,
          "validation_metric": 0.00014887739962432534,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.020341057999758
        },
        {
          "epoch": 846,
          "updates": 846,
          "samples_seen": 866304,
          "online_loss": 0.0017971807392314076,
          "train_loss": 0.0017927902517840266,
          "validation_loss": 0.001473772106692195,
          "train_metric": 0.00018056326371151954,
          "validation_metric": 0.0001484329259255901,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.030049078999582
        },
        {
          "epoch": 847,
          "updates": 847,
          "samples_seen": 867328,
          "online_loss": 0.0017927901353687048,
          "train_loss": 0.0017884147819131613,
          "validation_loss": 0.0014708017697557807,
          "train_metric": 0.0001801225880626589,
          "validation_metric": 0.00014813376765232533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.038822470000014
        },
        {
          "epoch": 848,
          "updates": 848,
          "samples_seen": 868352,
          "online_loss": 0.0017884147819131613,
          "train_loss": 0.0017840436194092035,
          "validation_loss": 0.0014666434144601226,
          "train_metric": 0.00017968234897125512,
          "validation_metric": 0.00014771496353205293,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.048126416999366
        },
        {
          "epoch": 849,
          "updates": 849,
          "samples_seen": 869376,
          "online_loss": 0.0017840437358245254,
          "train_loss": 0.0017796863103285432,
          "validation_loss": 0.0014633374521508813,
          "train_metric": 0.00017924349231179804,
          "validation_metric": 0.00014738197205588222,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.056936246999612
        },
        {
          "epoch": 850,
          "updates": 850,
          "samples_seen": 870400,
          "online_loss": 0.0017796860774978995,
          "train_loss": 0.0017753401771187782,
          "validation_loss": 0.001459623104892671,
          "train_metric": 0.00017880575614981353,
          "validation_metric": 0.00014700787141919136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.066390469000908
        },
        {
          "epoch": 851,
          "updates": 851,
          "samples_seen": 871424,
          "online_loss": 0.0017753399442881346,
          "train_loss": 0.0017710074316710234,
          "validation_loss": 0.0014559037517756224,
          "train_metric": 0.0001783693878678605,
          "validation_metric": 0.00014663329056929797,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.075392471999294
        },
        {
          "epoch": 852,
          "updates": 852,
          "samples_seen": 872448,
          "online_loss": 0.0017710074316710234,
          "train_loss": 0.0017666877247393131,
          "validation_loss": 0.0014526016311720014,
          "train_metric": 0.00017793430015444756,
          "validation_metric": 0.00014630070654675364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.084695621999344
        },
        {
          "epoch": 853,
          "updates": 853,
          "samples_seen": 873472,
          "online_loss": 0.0017666877247393131,
          "train_loss": 0.00176237802952528,
          "validation_loss": 0.0014485801802948117,
          "train_metric": 0.0001775002310751006,
          "validation_metric": 0.00014589569764211774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.198029327999393
        },
        {
          "epoch": 854,
          "updates": 854,
          "samples_seen": 874496,
          "online_loss": 0.0017623777966946363,
          "train_loss": 0.0017580774147063494,
          "validation_loss": 0.0014455000637099147,
          "train_metric": 0.00017706712242215872,
          "validation_metric": 0.00014558547991327941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.207287432000157
        },
        {
          "epoch": 855,
          "updates": 855,
          "samples_seen": 875520,
          "online_loss": 0.0017580772982910275,
          "train_loss": 0.0017537837848067284,
          "validation_loss": 0.0014413915341719985,
          "train_metric": 0.00017663465405348688,
          "validation_metric": 0.00014517168165184557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.217058452000856
        },
        {
          "epoch": 856,
          "updates": 856,
          "samples_seen": 876544,
          "online_loss": 0.0017537837848067284,
          "train_loss": 0.0017495026113465428,
          "validation_loss": 0.0014383252710103989,
          "train_metric": 0.0001762034953571856,
          "validation_metric": 0.00014486284635495394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.226528081999277
        },
        {
          "epoch": 857,
          "updates": 857,
          "samples_seen": 877568,
          "online_loss": 0.0017495026113465428,
          "train_loss": 0.0017452302854508162,
          "validation_loss": 0.0014343258226290345,
          "train_metric": 0.0001757731952238828,
          "validation_metric": 0.00014446003478951752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.235484295999413
        },
        {
          "epoch": 858,
          "updates": 858,
          "samples_seen": 878592,
          "online_loss": 0.001745230401866138,
          "train_loss": 0.0017409633146598935,
          "validation_loss": 0.0014311143895611167,
          "train_metric": 0.00017534347716718912,
          "validation_metric": 0.00014413657481782138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.244220656000834
        },
        {
          "epoch": 859,
          "updates": 859,
          "samples_seen": 879616,
          "online_loss": 0.0017409635474905372,
          "train_loss": 0.0017367119435220957,
          "validation_loss": 0.0014273443957790732,
          "train_metric": 0.00017491525795776397,
          "validation_metric": 0.0001437569153495133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.2529600630005
        },
        {
          "epoch": 860,
          "updates": 860,
          "samples_seen": 880640,
          "online_loss": 0.0017367119435220957,
          "train_loss": 0.0017324704676866531,
          "validation_loss": 0.001423908513970673,
          "train_metric": 0.00017448807193432003,
          "validation_metric": 0.00014341087080538273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.261383242999727
        },
        {
          "epoch": 861,
          "updates": 861,
          "samples_seen": 881664,
          "online_loss": 0.0017324704676866531,
          "train_loss": 0.0017282431945204735,
          "validation_loss": 0.0014203950995579362,
          "train_metric": 0.00017406234110239893,
          "validation_metric": 0.0001430569973308593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.27025016799962
        },
        {
          "epoch": 862,
          "updates": 862,
          "samples_seen": 882688,
          "online_loss": 0.0017282433109357953,
          "train_loss": 0.0017240222077816725,
          "validation_loss": 0.0014167559565976262,
          "train_metric": 0.0001736372068990022,
          "validation_metric": 0.00014269049279391766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.279389179000646
        },
        {
          "epoch": 863,
          "updates": 863,
          "samples_seen": 883712,
          "online_loss": 0.0017240222077816725,
          "train_loss": 0.001719817635603249,
          "validation_loss": 0.0014134574448689818,
          "train_metric": 0.0001732137316139415,
          "validation_metric": 0.00014235824346542358,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.288286025999696
        },
        {
          "epoch": 864,
          "updates": 864,
          "samples_seen": 884736,
          "online_loss": 0.0017198175191879272,
          "train_loss": 0.0017156164394691586,
          "validation_loss": 0.0014096726663410664,
          "train_metric": 0.0001727906201267615,
          "validation_metric": 0.00014197708514984697,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.297202826999637
        },
        {
          "epoch": 865,
          "updates": 865,
          "samples_seen": 885760,
          "online_loss": 0.0017156164394691586,
          "train_loss": 0.001711426884867251,
          "validation_loss": 0.0014065069844946265,
          "train_metric": 0.00017236867279279977,
          "validation_metric": 0.00014165823813527822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.306687930000408
        },
        {
          "epoch": 866,
          "updates": 866,
          "samples_seen": 886784,
          "online_loss": 0.0017114272341132164,
          "train_loss": 0.001707248855382204,
          "validation_loss": 0.001402666443027556,
          "train_metric": 0.00017194786050822586,
          "validation_metric": 0.00014127141912467778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.3157051649996
        },
        {
          "epoch": 867,
          "updates": 867,
          "samples_seen": 887808,
          "online_loss": 0.0017072490882128477,
          "train_loss": 0.0017030814196914434,
          "validation_loss": 0.0013995639747008681,
          "train_metric": 0.00017152812506537884,
          "validation_metric": 0.00014095894584897906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.32530818899977
        },
        {
          "epoch": 868,
          "updates": 868,
          "samples_seen": 888832,
          "online_loss": 0.0017030815361067653,
          "train_loss": 0.0016989224823191762,
          "validation_loss": 0.0013957277406007051,
          "train_metric": 0.00017110923363361508,
          "validation_metric": 0.00014057259249966592,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.33427852100067
        },
        {
          "epoch": 869,
          "updates": 869,
          "samples_seen": 889856,
          "online_loss": 0.0016989223659038544,
          "train_loss": 0.0016947718104347587,
          "validation_loss": 0.001392634934745729,
          "train_metric": 0.00017069120076484978,
          "validation_metric": 0.00014026110875420272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.344177484999818
        },
        {
          "epoch": 870,
          "updates": 870,
          "samples_seen": 890880,
          "online_loss": 0.0016947716940194368,
          "train_loss": 0.0016906347591429949,
          "validation_loss": 0.0013888515532016754,
          "train_metric": 0.00017027455032803118,
          "validation_metric": 0.00013988003775011748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.35318964399994
        },
        {
          "epoch": 871,
          "updates": 871,
          "samples_seen": 891904,
          "online_loss": 0.0016906348755583167,
          "train_loss": 0.001686505158431828,
          "validation_loss": 0.0013857289450243115,
          "train_metric": 0.00016985864203888923,
          "validation_metric": 0.00013956554175820202,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.36188465899977
        },
        {
          "epoch": 872,
          "updates": 872,
          "samples_seen": 892928,
          "online_loss": 0.001686505158431828,
          "train_loss": 0.0016823889454826713,
          "validation_loss": 0.001382018905133009,
          "train_metric": 0.00016944404342211783,
          "validation_metric": 0.000139191877678968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.370908934999534
        },
        {
          "epoch": 873,
          "updates": 873,
          "samples_seen": 893952,
          "online_loss": 0.001682389061897993,
          "train_loss": 0.0016782786697149277,
          "validation_loss": 0.0013788517098873854,
          "train_metric": 0.00016903008508961648,
          "validation_metric": 0.00013887288514524698,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.379658439000195
        },
        {
          "epoch": 874,
          "updates": 874,
          "samples_seen": 894976,
          "online_loss": 0.0016782785532996058,
          "train_loss": 0.0016741757281124592,
          "validation_loss": 0.0013752119848504663,
          "train_metric": 0.0001686168252490461,
          "validation_metric": 0.00013850632240064442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.388253565000923
        },
        {
          "epoch": 875,
          "updates": 875,
          "samples_seen": 896000,
          "online_loss": 0.0016741754952818155,
          "train_loss": 0.0016700929263606668,
          "validation_loss": 0.0013720201095566154,
          "train_metric": 0.00016820564633235335,
          "validation_metric": 0.00013818484148941934,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.397223597999982
        },
        {
          "epoch": 876,
          "updates": 876,
          "samples_seen": 897024,
          "online_loss": 0.001670092693530023,
          "train_loss": 0.0016660109395161271,
          "validation_loss": 0.0013684435980394483,
          "train_metric": 0.00016779452562332153,
          "validation_metric": 0.00013782462337985635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.406067011999767
        },
        {
          "epoch": 877,
          "updates": 877,
          "samples_seen": 898048,
          "online_loss": 0.0016660109395161271,
          "train_loss": 0.0016619430389255285,
          "validation_loss": 0.0013652228517457843,
          "train_metric": 0.00016738481645006686,
          "validation_metric": 0.00013750026118941605,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.4146738620002
        },
        {
          "epoch": 878,
          "updates": 878,
          "samples_seen": 899072,
          "online_loss": 0.0016619430389255285,
          "train_loss": 0.0016578859649598598,
          "validation_loss": 0.0013617109507322311,
          "train_metric": 0.0001669761841185391,
          "validation_metric": 0.00013714654778596014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.42336042200077
        },
        {
          "epoch": 879,
          "updates": 879,
          "samples_seen": 900096,
          "online_loss": 0.001657885848544538,
          "train_loss": 0.00165383773855865,
          "validation_loss": 0.001358474139124155,
          "train_metric": 0.00016656848310958594,
          "validation_metric": 0.0001368205266771838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.43205951299933
        },
        {
          "epoch": 880,
          "updates": 880,
          "samples_seen": 901120,
          "online_loss": 0.00165383773855865,
          "train_loss": 0.0016497966134920716,
          "validation_loss": 0.0013549994910135865,
          "train_metric": 0.00016616148059256375,
          "validation_metric": 0.00013647059677168727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.440975511000943
        },
        {
          "epoch": 881,
          "updates": 881,
          "samples_seen": 902144,
          "online_loss": 0.0016497967299073935,
          "train_loss": 0.0016457675956189632,
          "validation_loss": 0.0013517647748813033,
          "train_metric": 0.00016575567133259028,
          "validation_metric": 0.00013614477938972414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.451124982000692
        },
        {
          "epoch": 882,
          "updates": 882,
          "samples_seen": 903168,
          "online_loss": 0.0016457675956189632,
          "train_loss": 0.0016417509177699685,
          "validation_loss": 0.001348318182863295,
          "train_metric": 0.00016535115719307214,
          "validation_metric": 0.0001357976725557819,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.459812215000056
        },
        {
          "epoch": 883,
          "updates": 883,
          "samples_seen": 904192,
          "online_loss": 0.0016417510341852903,
          "train_loss": 0.0016377437859773636,
          "validation_loss": 0.0013450946426019073,
          "train_metric": 0.00016494755982421339,
          "validation_metric": 0.00013547299022320658,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.468357805999403
        },
        {
          "epoch": 884,
          "updates": 884,
          "samples_seen": 905216,
          "online_loss": 0.0016377437859773636,
          "train_loss": 0.001633743871934712,
          "validation_loss": 0.001341652823612094,
          "train_metric": 0.00016454470460303128,
          "validation_metric": 0.0001351263781543821,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.47715387900098
        },
        {
          "epoch": 885,
          "updates": 885,
          "samples_seen": 906240,
          "online_loss": 0.001633743871934712,
          "train_loss": 0.0016297547845169902,
          "validation_loss": 0.0013384755002334714,
          "train_metric": 0.00016414291167166084,
          "validation_metric": 0.00013480633788276464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.48562810600015
        },
        {
          "epoch": 886,
          "updates": 886,
          "samples_seen": 907264,
          "online_loss": 0.0016297545516863465,
          "train_loss": 0.001625776058062911,
          "validation_loss": 0.0013350138906389475,
          "train_metric": 0.00016374222468584776,
          "validation_metric": 0.00013445771764963865,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.49428198500027
        },
        {
          "epoch": 887,
          "updates": 887,
          "samples_seen": 908288,
          "online_loss": 0.001625776058062911,
          "train_loss": 0.0016218061791732907,
          "validation_loss": 0.0013319002464413643,
          "train_metric": 0.00016334239626303315,
          "validation_metric": 0.00013414410932455212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.503685678999318
        },
        {
          "epoch": 888,
          "updates": 888,
          "samples_seen": 909312,
          "online_loss": 0.0016218062955886126,
          "train_loss": 0.0016178474761545658,
          "validation_loss": 0.001328391837887466,
          "train_metric": 0.00016294368833769113,
          "validation_metric": 0.00013379075971897691,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.512683537999692
        },
        {
          "epoch": 889,
          "updates": 889,
          "samples_seen": 910336,
          "online_loss": 0.0016178474761545658,
          "train_loss": 0.0016138999490067363,
          "validation_loss": 0.0013253859942778945,
          "train_metric": 0.0001625461009098217,
          "validation_metric": 0.0001334880362264812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.522102363000158
        },
        {
          "epoch": 890,
          "updates": 890,
          "samples_seen": 911360,
          "online_loss": 0.001613900065422058,
          "train_loss": 0.0016099592903628945,
          "validation_loss": 0.0013217792147770524,
          "train_metric": 0.00016214919742196798,
          "validation_metric": 0.00013312474766280502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.533669709000606
        },
        {
          "epoch": 891,
          "updates": 891,
          "samples_seen": 912384,
          "online_loss": 0.0016099591739475727,
          "train_loss": 0.001606030622497201,
          "validation_loss": 0.0013189367018640041,
          "train_metric": 0.00016175350174307823,
          "validation_metric": 0.0001328384823864326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.641358482000214
        },
        {
          "epoch": 892,
          "updates": 892,
          "samples_seen": 913408,
          "online_loss": 0.0016060305060818791,
          "train_loss": 0.001602114294655621,
          "validation_loss": 0.0013151579769328237,
          "train_metric": 0.00016135908663272858,
          "validation_metric": 0.0001324579061474651,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.650247790999856
        },
        {
          "epoch": 893,
          "updates": 893,
          "samples_seen": 914432,
          "online_loss": 0.001602114294655621,
          "train_loss": 0.0015982070472091436,
          "validation_loss": 0.0013126011472195387,
          "train_metric": 0.00016096554463729262,
          "validation_metric": 0.00013220036635175347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.659067436999976
        },
        {
          "epoch": 894,
          "updates": 894,
          "samples_seen": 915456,
          "online_loss": 0.0015982070472091436,
          "train_loss": 0.001594316796399653,
          "validation_loss": 0.0013084844686090946,
          "train_metric": 0.00016057371976785362,
          "validation_metric": 0.00013178576773498207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.667937440999594
        },
        {
          "epoch": 895,
          "updates": 895,
          "samples_seen": 916480,
          "online_loss": 0.0015943166799843311,
          "train_loss": 0.001590440864674747,
          "validation_loss": 0.0013064327649772167,
          "train_metric": 0.0001601833791937679,
          "validation_metric": 0.0001315791014349088,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.677028725000127
        },
        {
          "epoch": 896,
          "updates": 896,
          "samples_seen": 917504,
          "online_loss": 0.001590440864674747,
          "train_loss": 0.0015865863533690572,
          "validation_loss": 0.0013017074670642614,
          "train_metric": 0.00015979517775122076,
          "validation_metric": 0.00013110321015119553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.68538105799962
        },
        {
          "epoch": 897,
          "updates": 897,
          "samples_seen": 918528,
          "online_loss": 0.0015865865861997008,
          "train_loss": 0.001582761644385755,
          "validation_loss": 0.0013005752116441727,
          "train_metric": 0.00015940997400321066,
          "validation_metric": 0.00013098918134346604,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.694077505999303
        },
        {
          "epoch": 898,
          "updates": 898,
          "samples_seen": 919552,
          "online_loss": 0.0015827617608010769,
          "train_loss": 0.0015789930475875735,
          "validation_loss": 0.0012947291834279895,
          "train_metric": 0.00015903040184639394,
          "validation_metric": 0.00013040036719758064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.702367891999529
        },
        {
          "epoch": 899,
          "updates": 899,
          "samples_seen": 920576,
          "online_loss": 0.0015789931640028954,
          "train_loss": 0.0015753076877444983,
          "validation_loss": 0.0012953514233231544,
          "train_metric": 0.00015865921159274876,
          "validation_metric": 0.0001304630422964692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.710029031999511
        },
        {
          "epoch": 900,
          "updates": 900,
          "samples_seen": 921600,
          "online_loss": 0.0015753075713291764,
          "train_loss": 0.001571781001985073,
          "validation_loss": 0.0012874527601525187,
          "train_metric": 0.00015830402844585478,
          "validation_metric": 0.00012966751819476485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.718407771000784
        },
        {
          "epoch": 901,
          "updates": 901,
          "samples_seen": 922624,
          "online_loss": 0.0015717808855697513,
          "train_loss": 0.001568546169437468,
          "validation_loss": 0.0012916026171296835,
          "train_metric": 0.00015797822561580688,
          "validation_metric": 0.00013008550740778446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.72681985500094
        },
        {
          "epoch": 902,
          "updates": 902,
          "samples_seen": 923648,
          "online_loss": 0.0015685462858527899,
          "train_loss": 0.0015658902702853084,
          "validation_loss": 0.0012800313998013735,
          "train_metric": 0.0001577107177581638,
          "validation_metric": 0.00012892008817289025,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.735312749000514
        },
        {
          "epoch": 903,
          "updates": 903,
          "samples_seen": 924672,
          "online_loss": 0.0015658901538699865,
          "train_loss": 0.0015643277438357472,
          "validation_loss": 0.0012918151915073395,
          "train_metric": 0.00015755335334688425,
          "validation_metric": 0.00013010688417125493,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.743543956999929
        },
        {
          "epoch": 904,
          "updates": 904,
          "samples_seen": 925696,
          "online_loss": 0.0015643276274204254,
          "train_loss": 0.0015650559216737747,
          "validation_loss": 0.0012743002735078335,
          "train_metric": 0.00015762668044772,
          "validation_metric": 0.00012834285735152662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.752210323000327
        },
        {
          "epoch": 905,
          "updates": 905,
          "samples_seen": 926720,
          "online_loss": 0.001565055805258453,
          "train_loss": 0.0015701003139838576,
          "validation_loss": 0.001304215402342379,
          "train_metric": 0.0001581347460160032,
          "validation_metric": 0.00013135581684764475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.760585370000626
        },
        {
          "epoch": 906,
          "updates": 906,
          "samples_seen": 927744,
          "online_loss": 0.0015701004303991795,
          "train_loss": 0.0015846578171476722,
          "validation_loss": 0.0012804983416572213,
          "train_metric": 0.0001596009242348373,
          "validation_metric": 0.00012896707630716264,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.768378279999524
        },
        {
          "epoch": 907,
          "updates": 907,
          "samples_seen": 928768,
          "online_loss": 0.001584657933562994,
          "train_loss": 0.0016160124214366078,
          "validation_loss": 0.0013582374667748809,
          "train_metric": 0.00016275886446237564,
          "validation_metric": 0.00013679670519195497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.776298872000552
        },
        {
          "epoch": 908,
          "updates": 908,
          "samples_seen": 929792,
          "online_loss": 0.0016160124214366078,
          "train_loss": 0.0016861301846802235,
          "validation_loss": 0.0013453528517857194,
          "train_metric": 0.00016982085071504116,
          "validation_metric": 0.0001354990090476349,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.78403028899993
        },
        {
          "epoch": 909,
          "updates": 909,
          "samples_seen": 930816,
          "online_loss": 0.0016861299518495798,
          "train_loss": 0.0018091574311256409,
          "validation_loss": 0.0015512469690293074,
          "train_metric": 0.0001822116901166737,
          "validation_metric": 0.00015623591025359929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.791912230999515
        },
        {
          "epoch": 910,
          "updates": 910,
          "samples_seen": 931840,
          "online_loss": 0.0018091569654643536,
          "train_loss": 0.00205240398645401,
          "validation_loss": 0.0016189885791391134,
          "train_metric": 0.00020671059610322118,
          "validation_metric": 0.0001630585902603343,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.799600815000304
        },
        {
          "epoch": 911,
          "updates": 911,
          "samples_seen": 932864,
          "online_loss": 0.0020524037536233664,
          "train_loss": 0.002308767754584551,
          "validation_loss": 0.00202060304582119,
          "train_metric": 0.00023253064136952162,
          "validation_metric": 0.00020350771956145763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.80717785699926
        },
        {
          "epoch": 912,
          "updates": 912,
          "samples_seen": 933888,
          "online_loss": 0.002308767754584551,
          "train_loss": 0.0026032349560409784,
          "validation_loss": 0.002056852448731661,
          "train_metric": 0.00026218825951218605,
          "validation_metric": 0.00020715863502118737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.814874854000664
        },
        {
          "epoch": 913,
          "updates": 913,
          "samples_seen": 934912,
          "online_loss": 0.0026032349560409784,
          "train_loss": 0.0023572780191898346,
          "validation_loss": 0.002068163827061653,
          "train_metric": 0.00023741641780361533,
          "validation_metric": 0.0002082979044644162,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.822793126999386
        },
        {
          "epoch": 914,
          "updates": 914,
          "samples_seen": 935936,
          "online_loss": 0.0023572780191898346,
          "train_loss": 0.001893886597827077,
          "validation_loss": 0.0014922746922820807,
          "train_metric": 0.00019074534066021442,
          "validation_metric": 0.00015029642963781953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.830956808000337
        },
        {
          "epoch": 915,
          "updates": 915,
          "samples_seen": 936960,
          "online_loss": 0.001893886481411755,
          "train_loss": 0.0015331266913563013,
          "validation_loss": 0.0012540500611066818,
          "train_metric": 0.00015441089635714889,
          "validation_metric": 0.00012630331912077963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.839558555999247
        },
        {
          "epoch": 916,
          "updates": 916,
          "samples_seen": 937984,
          "online_loss": 0.0015331265749409795,
          "train_loss": 0.0016726211179047823,
          "validation_loss": 0.0014063350390642881,
          "train_metric": 0.0001684602611931041,
          "validation_metric": 0.0001416409359080717,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.84773938500075
        },
        {
          "epoch": 917,
          "updates": 917,
          "samples_seen": 939008,
          "online_loss": 0.0016726211179047823,
          "train_loss": 0.002003488363698125,
          "validation_loss": 0.0015838323161005974,
          "train_metric": 0.00020178403065074235,
          "validation_metric": 0.00015951778914313763,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.855414363000818
        },
        {
          "epoch": 918,
          "updates": 918,
          "samples_seen": 940032,
          "online_loss": 0.0020034885965287685,
          "train_loss": 0.0019459802424535155,
          "validation_loss": 0.0016709620831534266,
          "train_metric": 0.00019599200459197164,
          "validation_metric": 0.00016829320520628244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.862949742000637
        },
        {
          "epoch": 919,
          "updates": 919,
          "samples_seen": 941056,
          "online_loss": 0.0019459801260381937,
          "train_loss": 0.0016467347741127014,
          "validation_loss": 0.0013112983433529735,
          "train_metric": 0.00016585308185312897,
          "validation_metric": 0.00013206916628405452,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.870832663000328
        },
        {
          "epoch": 920,
          "updates": 920,
          "samples_seen": 942080,
          "online_loss": 0.0016467348905280232,
          "train_loss": 0.0015064971521496773,
          "validation_loss": 0.0012199742486700416,
          "train_metric": 0.00015172886196523905,
          "validation_metric": 0.00012287132267374545,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.879207269999824
        },
        {
          "epoch": 921,
          "updates": 921,
          "samples_seen": 943104,
          "online_loss": 0.0015064970357343554,
          "train_loss": 0.0016795904375612736,
          "validation_loss": 0.0014161438448354602,
          "train_metric": 0.0001691621873760596,
          "validation_metric": 0.00014262882177717984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.88710320700011
        },
        {
          "epoch": 922,
          "updates": 922,
          "samples_seen": 944128,
          "online_loss": 0.0016795903211459517,
          "train_loss": 0.0018257424235343933,
          "validation_loss": 0.0014421619707718492,
          "train_metric": 0.00018388207536190748,
          "validation_metric": 0.000145249257911928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.894841098999677
        },
        {
          "epoch": 923,
          "updates": 923,
          "samples_seen": 945152,
          "online_loss": 0.0018257424235343933,
          "train_loss": 0.0016649032477289438,
          "validation_loss": 0.0014089307514950633,
          "train_metric": 0.00016768297064118087,
          "validation_metric": 0.00014190234651323408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.902383394000935
        },
        {
          "epoch": 924,
          "updates": 924,
          "samples_seen": 946176,
          "online_loss": 0.0016649033641442657,
          "train_loss": 0.001496428856626153,
          "validation_loss": 0.001212538336403668,
          "train_metric": 0.00015071482630446553,
          "validation_metric": 0.00012212240835651755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.910942689000876
        },
        {
          "epoch": 925,
          "updates": 925,
          "samples_seen": 947200,
          "online_loss": 0.001496428856626153,
          "train_loss": 0.0015513852704316378,
          "validation_loss": 0.001244565937668085,
          "train_metric": 0.00015624983643647283,
          "validation_metric": 0.00012534813140518963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.918833759000336
        },
        {
          "epoch": 926,
          "updates": 926,
          "samples_seen": 948224,
          "online_loss": 0.0015513852704316378,
          "train_loss": 0.0016683123540133238,
          "validation_loss": 0.0014257454313337803,
          "train_metric": 0.00016802630852907896,
          "validation_metric": 0.00014359585475176573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.926714798999456
        },
        {
          "epoch": 927,
          "updates": 927,
          "samples_seen": 949248,
          "online_loss": 0.0016683124704286456,
          "train_loss": 0.001630913931876421,
          "validation_loss": 0.0012975174468010664,
          "train_metric": 0.00016425969079136848,
          "validation_metric": 0.0001306812046095729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 10.934458642999743
        },
        {
          "epoch": 928,
          "updates": 928,
          "samples_seen": 950272,
          "online_loss": 0.0016309141647070646,
          "train_loss": 0.0014971905620768666,
          "validation_loss": 0.001248089480213821,
          "train_metric": 0.0001507915585534647,
          "validation_metric": 0.0001257029944099486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.040571883999291
        },
        {
          "epoch": 929,
          "updates": 929,
          "samples_seen": 951296,
          "online_loss": 0.0014971906784921885,
          "train_loss": 0.0014961186097934842,
          "validation_loss": 0.0012492758687585592,
          "train_metric": 0.00015068359789438546,
          "validation_metric": 0.00012582249473780394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.048760267000034
        },
        {
          "epoch": 930,
          "updates": 930,
          "samples_seen": 952320,
          "online_loss": 0.001496118726208806,
          "train_loss": 0.001585305784828961,
          "validation_loss": 0.0012677788035944104,
          "train_metric": 0.0001596661750227213,
          "validation_metric": 0.00012768604210577905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.056403128999591
        },
        {
          "epoch": 931,
          "updates": 931,
          "samples_seen": 953344,
          "online_loss": 0.001585305668413639,
          "train_loss": 0.001575316651724279,
          "validation_loss": 0.0013393133413046598,
          "train_metric": 0.0001586600992595777,
          "validation_metric": 0.00013489072443917394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.064302807999411
        },
        {
          "epoch": 932,
          "updates": 932,
          "samples_seen": 954368,
          "online_loss": 0.001575316535308957,
          "train_loss": 0.001489046961069107,
          "validation_loss": 0.0012048115022480488,
          "train_metric": 0.0001499713835073635,
          "validation_metric": 0.00012134420103393495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.073032534000959
        },
        {
          "epoch": 933,
          "updates": 933,
          "samples_seen": 955392,
          "online_loss": 0.0014890471938997507,
          "train_loss": 0.0014662720495834947,
          "validation_loss": 0.001196807250380516,
          "train_metric": 0.0001476775505580008,
          "validation_metric": 0.00012053803948219866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.081849901000169
        },
        {
          "epoch": 934,
          "updates": 934,
          "samples_seen": 956416,
          "online_loss": 0.0014662720495834947,
          "train_loss": 0.001519662793725729,
          "validation_loss": 0.0012800735421478748,
          "train_metric": 0.00015305487613659352,
          "validation_metric": 0.00012892430822830647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.089583208000477
        },
        {
          "epoch": 935,
          "updates": 935,
          "samples_seen": 957440,
          "online_loss": 0.001519662793725729,
          "train_loss": 0.0015360453398898244,
          "validation_loss": 0.0012287227436900139,
          "train_metric": 0.00015470487414859235,
          "validation_metric": 0.0001237524556927383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.09698161400047
        },
        {
          "epoch": 936,
          "updates": 936,
          "samples_seen": 958464,
          "online_loss": 0.0015360452234745026,
          "train_loss": 0.0014793627196922898,
          "validation_loss": 0.0012345107970759273,
          "train_metric": 0.0001489959831815213,
          "validation_metric": 0.0001243354199687019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.104698122999253
        },
        {
          "epoch": 937,
          "updates": 937,
          "samples_seen": 959488,
          "online_loss": 0.001479362603276968,
          "train_loss": 0.0014483373379334807,
          "validation_loss": 0.001191608957014978,
          "train_metric": 0.00014587123587261885,
          "validation_metric": 0.00012001449795207009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.113112548999197
        },
        {
          "epoch": 938,
          "updates": 938,
          "samples_seen": 960512,
          "online_loss": 0.001448337221518159,
          "train_loss": 0.0014768902910873294,
          "validation_loss": 0.0011930993059650064,
          "train_metric": 0.00014874697080813348,
          "validation_metric": 0.0001201646082336083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.120804136000515
        },
        {
          "epoch": 939,
          "updates": 939,
          "samples_seen": 961536,
          "online_loss": 0.0014768902910873294,
          "train_loss": 0.0014953003264963627,
          "validation_loss": 0.0012541187461465597,
          "train_metric": 0.00015060120495036244,
          "validation_metric": 0.00012631024583242834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.12853433700002
        },
        {
          "epoch": 940,
          "updates": 940,
          "samples_seen": 962560,
          "online_loss": 0.0014953003264963627,
          "train_loss": 0.00146642851177603,
          "validation_loss": 0.0011825530091300607,
          "train_metric": 0.0001476932957302779,
          "validation_metric": 0.0001191024057334289,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.137016590000712
        },
        {
          "epoch": 941,
          "updates": 941,
          "samples_seen": 963584,
          "online_loss": 0.0014664283953607082,
          "train_loss": 0.001435967511497438,
          "validation_loss": 0.0011792215518653393,
          "train_metric": 0.00014462538820225745,
          "validation_metric": 0.00011876688222400844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.14580623800066
        },
        {
          "epoch": 942,
          "updates": 942,
          "samples_seen": 964608,
          "online_loss": 0.001435967511497438,
          "train_loss": 0.0014460897073149681,
          "validation_loss": 0.001198092824779451,
          "train_metric": 0.00014564486627932638,
          "validation_metric": 0.00012066750787198544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.153949547000593
        },
        {
          "epoch": 943,
          "updates": 943,
          "samples_seen": 965632,
          "online_loss": 0.0014460897073149681,
          "train_loss": 0.0014644067268818617,
          "validation_loss": 0.0011771214194595814,
          "train_metric": 0.0001474896853324026,
          "validation_metric": 0.00011855534103233367,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.162438043000293
        },
        {
          "epoch": 944,
          "updates": 944,
          "samples_seen": 966656,
          "online_loss": 0.0014644067268818617,
          "train_loss": 0.0014501010300591588,
          "validation_loss": 0.0012060162844136357,
          "train_metric": 0.0001460488565498963,
          "validation_metric": 0.00012146553490310907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.170626180999534
        },
        {
          "epoch": 945,
          "updates": 945,
          "samples_seen": 967680,
          "online_loss": 0.0014501011464744806,
          "train_loss": 0.0014256418216973543,
          "validation_loss": 0.0011630039662122726,
          "train_metric": 0.00014358543558046222,
          "validation_metric": 0.00011713348794728518,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.179076262000308
        },
        {
          "epoch": 946,
          "updates": 946,
          "samples_seen": 968704,
          "online_loss": 0.0014256418216973543,
          "train_loss": 0.001424581976607442,
          "validation_loss": 0.0011609930079430342,
          "train_metric": 0.00014347871183417737,
          "validation_metric": 0.000116930968943052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.18776222300039
        },
        {
          "epoch": 947,
          "updates": 947,
          "samples_seen": 969728,
          "online_loss": 0.0014245822094380856,
          "train_loss": 0.001436986611224711,
          "validation_loss": 0.001195825869217515,
          "train_metric": 0.00014472803741227835,
          "validation_metric": 0.00012043920287396759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.195483921999767
        },
        {
          "epoch": 948,
          "updates": 948,
          "samples_seen": 970752,
          "online_loss": 0.0014369864948093891,
          "train_loss": 0.0014339719200506806,
          "validation_loss": 0.0011608143104240298,
          "train_metric": 0.00014442439714912325,
          "validation_metric": 0.0001169129682239145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.20418053100002
        },
        {
          "epoch": 949,
          "updates": 949,
          "samples_seen": 971776,
          "online_loss": 0.0014339721528813243,
          "train_loss": 0.0014161554863676429,
          "validation_loss": 0.0011720253387466073,
          "train_metric": 0.00014263000048231333,
          "validation_metric": 0.00011804209498222917,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.211952487999952
        },
        {
          "epoch": 950,
          "updates": 950,
          "samples_seen": 972800,
          "online_loss": 0.0014161556027829647,
          "train_loss": 0.0014083451824262738,
          "validation_loss": 0.001159912208095193,
          "train_metric": 0.00014184338215272874,
          "validation_metric": 0.00011682210606522858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.22040448900043
        },
        {
          "epoch": 951,
          "updates": 951,
          "samples_seen": 973824,
          "online_loss": 0.0014083452988415956,
          "train_loss": 0.0014145722379907966,
          "validation_loss": 0.0011501496192067862,
          "train_metric": 0.00014247052604332566,
          "validation_metric": 0.0001158388695330359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.229294904000199
        },
        {
          "epoch": 952,
          "updates": 952,
          "samples_seen": 974848,
          "online_loss": 0.0014145722379907966,
          "train_loss": 0.0014163544401526451,
          "validation_loss": 0.0011786012910306454,
          "train_metric": 0.0001426500384695828,
          "validation_metric": 0.00011870442540384829,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.237547656999595
        },
        {
          "epoch": 953,
          "updates": 953,
          "samples_seen": 975872,
          "online_loss": 0.0014163544401526451,
          "train_loss": 0.0014061995316296816,
          "validation_loss": 0.0011461438843980432,
          "train_metric": 0.0001416272425558418,
          "validation_metric": 0.00011543542495928705,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.245926691000932
        },
        {
          "epoch": 954,
          "updates": 954,
          "samples_seen": 976896,
          "online_loss": 0.0014061995316296816,
          "train_loss": 0.0013960172655060887,
          "validation_loss": 0.0011500073596835136,
          "train_metric": 0.00014060173998586833,
          "validation_metric": 0.00011582453589653596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.25383873499959
        },
        {
          "epoch": 955,
          "updates": 955,
          "samples_seen": 977920,
          "online_loss": 0.001396017149090767,
          "train_loss": 0.0013958836207166314,
          "validation_loss": 0.0011546759633347392,
          "train_metric": 0.00014058829401619732,
          "validation_metric": 0.00011629473738139495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.261685815999954
        },
        {
          "epoch": 956,
          "updates": 956,
          "samples_seen": 978944,
          "online_loss": 0.001395883853547275,
          "train_loss": 0.0013987916754558682,
          "validation_loss": 0.0011391243897378445,
          "train_metric": 0.00014088116586208344,
          "validation_metric": 0.00011472843470983207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.270384637000461
        },
        {
          "epoch": 957,
          "updates": 957,
          "samples_seen": 979968,
          "online_loss": 0.00139879179187119,
          "train_loss": 0.0013944506645202637,
          "validation_loss": 0.0011581520084291697,
          "train_metric": 0.00014044396812096238,
          "validation_metric": 0.00011664483463391662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.278475523000452
        },
        {
          "epoch": 958,
          "updates": 958,
          "samples_seen": 980992,
          "online_loss": 0.0013944506645202637,
          "train_loss": 0.0013857090380042791,
          "validation_loss": 0.0011349869892001152,
          "train_metric": 0.0001395635335939005,
          "validation_metric": 0.00011431172606535256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.286961244999475
        },
        {
          "epoch": 959,
          "updates": 959,
          "samples_seen": 982016,
          "online_loss": 0.0013857090380042791,
          "train_loss": 0.001381204929202795,
          "validation_loss": 0.0011334771988913417,
          "train_metric": 0.0001391098921885714,
          "validation_metric": 0.00011415967310313135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.295803172000888
        },
        {
          "epoch": 960,
          "updates": 960,
          "samples_seen": 983040,
          "online_loss": 0.0013812050456181169,
          "train_loss": 0.0013818624429404736,
          "validation_loss": 0.0011433156905695796,
          "train_metric": 0.00013917613250669092,
          "validation_metric": 0.00011515056394273415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.303641806000087
        },
        {
          "epoch": 961,
          "updates": 961,
          "samples_seen": 984064,
          "online_loss": 0.0013818623265251517,
          "train_loss": 0.0013810014352202415,
          "validation_loss": 0.001126002287492156,
          "train_metric": 0.00013908938854001462,
          "validation_metric": 0.00011340681521687657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.312300393999976
        },
        {
          "epoch": 962,
          "updates": 962,
          "samples_seen": 985088,
          "online_loss": 0.0013810015516355634,
          "train_loss": 0.0013753932435065508,
          "validation_loss": 0.0011371755972504616,
          "train_metric": 0.00013852455595042557,
          "validation_metric": 0.00011453216575318947,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.320276848000503
        },
        {
          "epoch": 963,
          "updates": 963,
          "samples_seen": 986112,
          "online_loss": 0.001375393127091229,
          "train_loss": 0.0013694040244445205,
          "validation_loss": 0.0011228864314034581,
          "train_metric": 0.0001379213499603793,
          "validation_metric": 0.00011309301044093445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.328922966000391
        },
        {
          "epoch": 964,
          "updates": 964,
          "samples_seen": 987136,
          "online_loss": 0.0013694040244445205,
          "train_loss": 0.001367036486044526,
          "validation_loss": 0.001119443099014461,
          "train_metric": 0.0001376829168293625,
          "validation_metric": 0.00011274620919721201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.337953797999944
        },
        {
          "epoch": 965,
          "updates": 965,
          "samples_seen": 988160,
          "online_loss": 0.001367036486044526,
          "train_loss": 0.0013664973666891456,
          "validation_loss": 0.0011295084841549397,
          "train_metric": 0.00013762860908173025,
          "validation_metric": 0.00011375996109563857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.345818440000585
        },
        {
          "epoch": 966,
          "updates": 966,
          "samples_seen": 989184,
          "online_loss": 0.0013664973666891456,
          "train_loss": 0.0013638795353472233,
          "validation_loss": 0.0011135418899357319,
          "train_metric": 0.00013736492837779224,
          "validation_metric": 0.0001121518507716246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.354316881999694
        },
        {
          "epoch": 967,
          "updates": 967,
          "samples_seen": 990208,
          "online_loss": 0.0013638791861012578,
          "train_loss": 0.0013588751899078488,
          "validation_loss": 0.0011204156326130033,
          "train_metric": 0.00013686093734577298,
          "validation_metric": 0.00011284415813861415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.465243194999857
        },
        {
          "epoch": 968,
          "updates": 968,
          "samples_seen": 991232,
          "online_loss": 0.0013588751899078488,
          "train_loss": 0.0013545778347179294,
          "validation_loss": 0.0011118854163214564,
          "train_metric": 0.00013642811973113567,
          "validation_metric": 0.0001119850276154466,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.474484959999245
        },
        {
          "epoch": 969,
          "updates": 969,
          "samples_seen": 992256,
          "online_loss": 0.0013545777183026075,
          "train_loss": 0.0013525267131626606,
          "validation_loss": 0.0011072667548432946,
          "train_metric": 0.00013622154074255377,
          "validation_metric": 0.0001115198465413414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.483414149000964
        },
        {
          "epoch": 970,
          "updates": 970,
          "samples_seen": 993280,
          "online_loss": 0.001352526480332017,
          "train_loss": 0.0013509646523743868,
          "validation_loss": 0.0011149964993819594,
          "train_metric": 0.00013606420543510467,
          "validation_metric": 0.00011229835945414379,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.491427320999719
        },
        {
          "epoch": 971,
          "updates": 971,
          "samples_seen": 994304,
          "online_loss": 0.0013509646523743868,
          "train_loss": 0.0013478893088176847,
          "validation_loss": 0.0011012298054993153,
          "train_metric": 0.00013575446791946888,
          "validation_metric": 0.00011091183114331216,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.500198174000616
        },
        {
          "epoch": 972,
          "updates": 972,
          "samples_seen": 995328,
          "online_loss": 0.0013478893088176847,
          "train_loss": 0.0013436743756756186,
          "validation_loss": 0.0011062935227528214,
          "train_metric": 0.0001353299740003422,
          "validation_metric": 0.00011142183939227834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.50804745700043
        },
        {
          "epoch": 973,
          "updates": 973,
          "samples_seen": 996352,
          "online_loss": 0.0013436744920909405,
          "train_loss": 0.001340166898444295,
          "validation_loss": 0.0011008227011188865,
          "train_metric": 0.00013497669715434313,
          "validation_metric": 0.00011087083112215623,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.516550219999772
        },
        {
          "epoch": 974,
          "updates": 974,
          "samples_seen": 997376,
          "online_loss": 0.001340166898444295,
          "train_loss": 0.0013379314914345741,
          "validation_loss": 0.001095831859856844,
          "train_metric": 0.00013475156447384506,
          "validation_metric": 0.00011036817886633798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.525695378999444
        },
        {
          "epoch": 975,
          "updates": 975,
          "samples_seen": 998400,
          "online_loss": 0.0013379314914345741,
          "train_loss": 0.001335812732577324,
          "validation_loss": 0.0011014771880581975,
          "train_metric": 0.00013453816063702106,
          "validation_metric": 0.00011093674402218312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.534154322000177
        },
        {
          "epoch": 976,
          "updates": 976,
          "samples_seen": 999424,
          "online_loss": 0.001335812616162002,
          "train_loss": 0.001332768821157515,
          "validation_loss": 0.0010903736110776663,
          "train_metric": 0.0001342316099908203,
          "validation_metric": 0.00010981845116475597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.542983462000848
        },
        {
          "epoch": 977,
          "updates": 977,
          "samples_seen": 1000448,
          "online_loss": 0.0013327689375728369,
          "train_loss": 0.001329119666479528,
          "validation_loss": 0.0010939796920865774,
          "train_metric": 0.00013386407226789743,
          "validation_metric": 0.00011018163058906794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.551279160999911
        },
        {
          "epoch": 978,
          "updates": 978,
          "samples_seen": 1001472,
          "online_loss": 0.001329119666479528,
          "train_loss": 0.001325905555859208,
          "validation_loss": 0.0010890039848163724,
          "train_metric": 0.00013354035036172718,
          "validation_metric": 0.00010968049173243344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.561243193000337
        },
        {
          "epoch": 979,
          "updates": 979,
          "samples_seen": 1002496,
          "online_loss": 0.0013259053230285645,
          "train_loss": 0.0013234346406534314,
          "validation_loss": 0.0010843216441571712,
          "train_metric": 0.00013329148350749165,
          "validation_metric": 0.00010920890781562775,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.572585976000482
        },
        {
          "epoch": 980,
          "updates": 980,
          "samples_seen": 1003520,
          "online_loss": 0.0013234346406534314,
          "train_loss": 0.001321063027717173,
          "validation_loss": 0.0010882093338295817,
          "train_metric": 0.00013305264292284846,
          "validation_metric": 0.0001096004489227198,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.581722630000513
        },
        {
          "epoch": 981,
          "updates": 981,
          "samples_seen": 1004544,
          "online_loss": 0.001321063144132495,
          "train_loss": 0.0013181789545342326,
          "validation_loss": 0.0010791063541546464,
          "train_metric": 0.0001327621575910598,
          "validation_metric": 0.00010868364915950224,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.591982366000593
        },
        {
          "epoch": 982,
          "updates": 982,
          "samples_seen": 1005568,
          "online_loss": 0.0013181788381189108,
          "train_loss": 0.0013148997677490115,
          "validation_loss": 0.0010817686561495066,
          "train_metric": 0.0001324318873230368,
          "validation_metric": 0.00010895179002545774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.60077697700035
        },
        {
          "epoch": 983,
          "updates": 983,
          "samples_seen": 1006592,
          "online_loss": 0.0013148996513336897,
          "train_loss": 0.00131180917378515,
          "validation_loss": 0.0010767907369881868,
          "train_metric": 0.00013212062185630202,
          "validation_metric": 0.0001084504256141372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.610075336999216
        },
        {
          "epoch": 984,
          "updates": 984,
          "samples_seen": 1007616,
          "online_loss": 0.0013118092902004719,
          "train_loss": 0.0013091569999232888,
          "validation_loss": 0.0010733981616795063,
          "train_metric": 0.0001318534923484549,
          "validation_metric": 0.00010810873209265992,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.618768731999808
        },
        {
          "epoch": 985,
          "updates": 985,
          "samples_seen": 1008640,
          "online_loss": 0.001309156883507967,
          "train_loss": 0.0013066557003185153,
          "validation_loss": 0.0010755416005849838,
          "train_metric": 0.00013160158414393663,
          "validation_metric": 0.00010832460975507274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.627693910999369
        },
        {
          "epoch": 986,
          "updates": 986,
          "samples_seen": 1009664,
          "online_loss": 0.0013066558167338371,
          "train_loss": 0.001303914119489491,
          "validation_loss": 0.001067904755473137,
          "train_metric": 0.00013132544700056314,
          "validation_metric": 0.00010755546827567741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.636263525999311
        },
        {
          "epoch": 987,
          "updates": 987,
          "samples_seen": 1010688,
          "online_loss": 0.001303914119489491,
          "train_loss": 0.001300897216424346,
          "validation_loss": 0.001070016180165112,
          "train_metric": 0.00013102158845867962,
          "validation_metric": 0.00010776810813695192,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.645001803000923
        },
        {
          "epoch": 988,
          "updates": 988,
          "samples_seen": 1011712,
          "online_loss": 0.0013008971000090241,
          "train_loss": 0.0012978946324437857,
          "validation_loss": 0.0010649020550772548,
          "train_metric": 0.0001307191705564037,
          "validation_metric": 0.00010725303582148626,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.65358078600002
        },
        {
          "epoch": 989,
          "updates": 989,
          "samples_seen": 1012736,
          "online_loss": 0.0012978946324437857,
          "train_loss": 0.0012951288372278214,
          "validation_loss": 0.0010626617586240172,
          "train_metric": 0.00013044064689893275,
          "validation_metric": 0.00010702740837587044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.663025990999813
        },
        {
          "epoch": 990,
          "updates": 990,
          "samples_seen": 1013760,
          "online_loss": 0.0012951287208124995,
          "train_loss": 0.0012925288174301386,
          "validation_loss": 0.0010629512835294008,
          "train_metric": 0.00013017875608056784,
          "validation_metric": 0.00010705656313803047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.670972371000971
        },
        {
          "epoch": 991,
          "updates": 991,
          "samples_seen": 1014784,
          "online_loss": 0.0012925287010148168,
          "train_loss": 0.0012898765271529555,
          "validation_loss": 0.0010568253928795457,
          "train_metric": 0.00012991165567655116,
          "validation_metric": 0.00010643958376022056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.68432718099939
        },
        {
          "epoch": 992,
          "updates": 992,
          "samples_seen": 1015808,
          "online_loss": 0.0012898766435682774,
          "train_loss": 0.0012870538048446178,
          "validation_loss": 0.0010586653370410204,
          "train_metric": 0.00012962734035681933,
          "validation_metric": 0.0001066248951246962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.698607653999716
        },
        {
          "epoch": 993,
          "updates": 993,
          "samples_seen": 1016832,
          "online_loss": 0.0012870538048446178,
          "train_loss": 0.001284151105210185,
          "validation_loss": 0.0010534168686717749,
          "train_metric": 0.00012933500693179667,
          "validation_metric": 0.00010609628952806816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.713460622999264
        },
        {
          "epoch": 994,
          "updates": 994,
          "samples_seen": 1017856,
          "online_loss": 0.001284151105210185,
          "train_loss": 0.0012813382782042027,
          "validation_loss": 0.0010522380471229553,
          "train_metric": 0.0001290516956942156,
          "validation_metric": 0.00010597756045171991,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.728414306999184
        },
        {
          "epoch": 995,
          "updates": 995,
          "samples_seen": 1018880,
          "online_loss": 0.0012813382782042027,
          "train_loss": 0.0012786551378667355,
          "validation_loss": 0.0010508819250389934,
          "train_metric": 0.00012878146662842482,
          "validation_metric": 0.00010584098345134407,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.741432671000439
        },
        {
          "epoch": 996,
          "updates": 996,
          "samples_seen": 1019904,
          "online_loss": 0.0012786551378667355,
          "train_loss": 0.001276023918762803,
          "validation_loss": 0.0010463682701811194,
          "train_metric": 0.0001285164471482858,
          "validation_metric": 0.00010538638161960989,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.75139204900006
        },
        {
          "epoch": 997,
          "updates": 997,
          "samples_seen": 1020928,
          "online_loss": 0.0012760236859321594,
          "train_loss": 0.0012733276234939694,
          "validation_loss": 0.0010473734000697732,
          "train_metric": 0.00012824487930629402,
          "validation_metric": 0.00010548761929385364,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.75978408499941
        },
        {
          "epoch": 998,
          "updates": 998,
          "samples_seen": 1021952,
          "online_loss": 0.0012733275070786476,
          "train_loss": 0.001270551118068397,
          "validation_loss": 0.0010420934995636344,
          "train_metric": 0.00012796526425518095,
          "validation_metric": 0.00010495582682779059,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.768512949000069
        },
        {
          "epoch": 999,
          "updates": 999,
          "samples_seen": 1022976,
          "online_loss": 0.001270551118068397,
          "train_loss": 0.0012677606428042054,
          "validation_loss": 0.0010417874436825514,
          "train_metric": 0.00012768420856446028,
          "validation_metric": 0.00010492499859537929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.777992213999823
        },
        {
          "epoch": 1000,
          "updates": 1000,
          "samples_seen": 1024000,
          "online_loss": 0.0012677606428042054,
          "train_loss": 0.001265028491616249,
          "validation_loss": 0.0010390349198132753,
          "train_metric": 0.00012740903184749186,
          "validation_metric": 0.00010464779916219413,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.78679011299937
        },
        {
          "epoch": 1001,
          "updates": 1001,
          "samples_seen": 1025024,
          "online_loss": 0.001265028491616249,
          "train_loss": 0.0012623693328350782,
          "validation_loss": 0.001036064582876861,
          "train_metric": 0.00012714120384771377,
          "validation_metric": 0.00010434861906105652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.795669028999328
        },
        {
          "epoch": 1002,
          "updates": 1002,
          "samples_seen": 1026048,
          "online_loss": 0.0012623693328350782,
          "train_loss": 0.0012597284512594342,
          "validation_loss": 0.0010358226718381047,
          "train_metric": 0.0001268752384930849,
          "validation_metric": 0.00010432425187900662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.804671580999639
        },
        {
          "epoch": 1003,
          "updates": 1003,
          "samples_seen": 1027072,
          "online_loss": 0.0012597284512594342,
          "train_loss": 0.0012570539256557822,
          "validation_loss": 0.0010312055237591267,
          "train_metric": 0.0001266058679902926,
          "validation_metric": 0.00010385923815192655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.81376301400087
        },
        {
          "epoch": 1004,
          "updates": 1004,
          "samples_seen": 1028096,
          "online_loss": 0.001257054042071104,
          "train_loss": 0.0012543417979031801,
          "validation_loss": 0.001031304243952036,
          "train_metric": 0.00012633272854145616,
          "validation_metric": 0.00010386918438598514,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.822688580999966
        },
        {
          "epoch": 1005,
          "updates": 1005,
          "samples_seen": 1029120,
          "online_loss": 0.0012543417979031801,
          "train_loss": 0.001251628273166716,
          "validation_loss": 0.001027515041641891,
          "train_metric": 0.00012605939991772175,
          "validation_metric": 0.00010348756040912122,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.928031657000247
        },
        {
          "epoch": 1006,
          "updates": 1006,
          "samples_seen": 1030144,
          "online_loss": 0.0012516281567513943,
          "train_loss": 0.0012489412911236286,
          "validation_loss": 0.0010258740512654185,
          "train_metric": 0.00012578880705405027,
          "validation_metric": 0.00010332227975595742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.936571395999636
        },
        {
          "epoch": 1007,
          "updates": 1007,
          "samples_seen": 1031168,
          "online_loss": 0.0012489412911236286,
          "train_loss": 0.001246293424628675,
          "validation_loss": 0.0010241593699902296,
          "train_metric": 0.00012552211410366,
          "validation_metric": 0.0001031495921779424,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.945495654999831
        },
        {
          "epoch": 1008,
          "updates": 1008,
          "samples_seen": 1032192,
          "online_loss": 0.0012462935410439968,
          "train_loss": 0.0012436641845852137,
          "validation_loss": 0.0010207800660282373,
          "train_metric": 0.00012525729835033417,
          "validation_metric": 0.0001028092228807509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.9548223769998
        },
        {
          "epoch": 1009,
          "updates": 1009,
          "samples_seen": 1033216,
          "online_loss": 0.0012436641845852137,
          "train_loss": 0.0012410274939611554,
          "validation_loss": 0.0010204104473814368,
          "train_metric": 0.00012499172589741647,
          "validation_metric": 0.00010277199908159673,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.964201987999331
        },
        {
          "epoch": 1010,
          "updates": 1010,
          "samples_seen": 1034240,
          "online_loss": 0.0012410273775458336,
          "train_loss": 0.0012383752036839724,
          "validation_loss": 0.0010164628038182855,
          "train_metric": 0.00012472461094148457,
          "validation_metric": 0.00010237441165372729,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.973140931000671
        },
        {
          "epoch": 1011,
          "updates": 1011,
          "samples_seen": 1035264,
          "online_loss": 0.0012383753200992942,
          "train_loss": 0.0012357168598100543,
          "validation_loss": 0.00101576279848814,
          "train_metric": 0.00012445687025319785,
          "validation_metric": 0.00010230390762444586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.981682247000208
        },
        {
          "epoch": 1012,
          "updates": 1012,
          "samples_seen": 1036288,
          "online_loss": 0.0012357169762253761,
          "train_loss": 0.001233067479915917,
          "validation_loss": 0.0010128405410796404,
          "train_metric": 0.00012419004633557051,
          "validation_metric": 0.00010200960241490975,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.99051693399997
        },
        {
          "epoch": 1013,
          "updates": 1013,
          "samples_seen": 1037312,
          "online_loss": 0.0012330672470852733,
          "train_loss": 0.0012304394040256739,
          "validation_loss": 0.0010108123533427715,
          "train_metric": 0.00012392536154948175,
          "validation_metric": 0.00010180532262893394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 11.999278658999174
        },
        {
          "epoch": 1014,
          "updates": 1014,
          "samples_seen": 1038336,
          "online_loss": 0.0012304394040256739,
          "train_loss": 0.0012278346111997962,
          "validation_loss": 0.0010092720622196794,
          "train_metric": 0.00012366301962174475,
          "validation_metric": 0.00010165019193664193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.00802187500085
        },
        {
          "epoch": 1015,
          "updates": 1015,
          "samples_seen": 1039360,
          "online_loss": 0.001227834727615118,
          "train_loss": 0.0012252309825271368,
          "validation_loss": 0.001006085192784667,
          "train_metric": 0.00012340079410932958,
          "validation_metric": 0.00010132923489436507,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.016945549999946
        },
        {
          "epoch": 1016,
          "updates": 1016,
          "samples_seen": 1040384,
          "online_loss": 0.0012252309825271368,
          "train_loss": 0.0012226299149915576,
          "validation_loss": 0.0010053287260234356,
          "train_metric": 0.0001231388159794733,
          "validation_metric": 0.00010125302651431412,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.026012964999609
        },
        {
          "epoch": 1017,
          "updates": 1017,
          "samples_seen": 1041408,
          "online_loss": 0.0012226300314068794,
          "train_loss": 0.0012200213968753815,
          "validation_loss": 0.0010019246255978942,
          "train_metric": 0.00012287609570194036,
          "validation_metric": 0.00010091017611557618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.034880178000094
        },
        {
          "epoch": 1018,
          "updates": 1018,
          "samples_seen": 1042432,
          "online_loss": 0.0012200213968753815,
          "train_loss": 0.0012174148578196764,
          "validation_loss": 0.0010008311364799738,
          "train_metric": 0.00012261357915122062,
          "validation_metric": 0.0001008000472211279,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.043676604000211
        },
        {
          "epoch": 1019,
          "updates": 1019,
          "samples_seen": 1043456,
          "online_loss": 0.0012174148578196764,
          "train_loss": 0.001214815303683281,
          "validation_loss": 0.0009981323964893818,
          "train_metric": 0.00012235174654051661,
          "validation_metric": 0.00010052822472061962,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.05245722700056
        },
        {
          "epoch": 1020,
          "updates": 1020,
          "samples_seen": 1044480,
          "online_loss": 0.0012148155365139246,
          "train_loss": 0.001212230883538723,
          "validation_loss": 0.0009961803443729877,
          "train_metric": 0.00012209145643282682,
          "validation_metric": 0.00010033162834588438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.0620160819999
        },
        {
          "epoch": 1021,
          "updates": 1021,
          "samples_seen": 1045504,
          "online_loss": 0.0012122305342927575,
          "train_loss": 0.0012096550781279802,
          "validation_loss": 0.0009944362100213766,
          "train_metric": 0.00012183203944005072,
          "validation_metric": 0.00010015596490120515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.071783674000471
        },
        {
          "epoch": 1022,
          "updates": 1022,
          "samples_seen": 1046528,
          "online_loss": 0.001209655194543302,
          "train_loss": 0.0012070882366970181,
          "validation_loss": 0.0009916323469951749,
          "train_metric": 0.00012157351011410356,
          "validation_metric": 9.987357771024108e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.080342176999693
        },
        {
          "epoch": 1023,
          "updates": 1023,
          "samples_seen": 1047552,
          "online_loss": 0.0012070882366970181,
          "train_loss": 0.001204524072818458,
          "validation_loss": 0.0009904782054945827,
          "train_metric": 0.00012131526455050334,
          "validation_metric": 9.975733701139688e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.089053534000413
        },
        {
          "epoch": 1024,
          "updates": 1024,
          "samples_seen": 1048576,
          "online_loss": 0.00120452418923378,
          "train_loss": 0.0012019607238471508,
          "validation_loss": 0.0009874297538772225,
          "train_metric": 0.00012105709174647927,
          "validation_metric": 9.945030615199357e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.098120894999738
        },
        {
          "epoch": 1025,
          "updates": 1025,
          "samples_seen": 1049600,
          "online_loss": 0.0012019607238471508,
          "train_loss": 0.0011994065716862679,
          "validation_loss": 0.0009862231090664864,
          "train_metric": 0.00012079985026502982,
          "validation_metric": 9.932877583196387e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.10692380799992
        },
        {
          "epoch": 1026,
          "updates": 1026,
          "samples_seen": 1050624,
          "online_loss": 0.0011994065716862679,
          "train_loss": 0.0011968502076342702,
          "validation_loss": 0.0009834753582254052,
          "train_metric": 0.0001205423686769791,
          "validation_metric": 9.905204933602363e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.115710465999655
        },
        {
          "epoch": 1027,
          "updates": 1027,
          "samples_seen": 1051648,
          "online_loss": 0.0011968502076342702,
          "train_loss": 0.001194304320961237,
          "validation_loss": 0.0009817874524742365,
          "train_metric": 0.00012028596393065527,
          "validation_metric": 9.888203931041062e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.124551102000623
        },
        {
          "epoch": 1028,
          "updates": 1028,
          "samples_seen": 1052672,
          "online_loss": 0.001194304320961237,
          "train_loss": 0.001191762275993824,
          "validation_loss": 0.0009796775411814451,
          "train_metric": 0.00012002992298221216,
          "validation_metric": 9.866952314041555e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.133454074000838
        },
        {
          "epoch": 1029,
          "updates": 1029,
          "samples_seen": 1053696,
          "online_loss": 0.0011917620431631804,
          "train_loss": 0.001189233036711812,
          "validation_loss": 0.0009773848578333855,
          "train_metric": 0.00011977519898209721,
          "validation_metric": 9.843862062552944e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.142411872999219
        },
        {
          "epoch": 1030,
          "updates": 1030,
          "samples_seen": 1054720,
          "online_loss": 0.001189233036711812,
          "train_loss": 0.0011867088032886386,
          "validation_loss": 0.0009757978259585798,
          "train_metric": 0.00011952096247114241,
          "validation_metric": 9.827878966461867e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.151786414000526
        },
        {
          "epoch": 1031,
          "updates": 1031,
          "samples_seen": 1055744,
          "online_loss": 0.0011867088032886386,
          "train_loss": 0.0011841884115710855,
          "validation_loss": 0.0009731499012559652,
          "train_metric": 0.00011926710430998355,
          "validation_metric": 9.801209671422839e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.160835556000166
        },
        {
          "epoch": 1032,
          "updates": 1032,
          "samples_seen": 1056768,
          "online_loss": 0.0011841881787404418,
          "train_loss": 0.0011816739570349455,
          "validation_loss": 0.0009717944194562733,
          "train_metric": 0.00011901387188117951,
          "validation_metric": 9.787558519747108e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.169210158999704
        },
        {
          "epoch": 1033,
          "updates": 1033,
          "samples_seen": 1057792,
          "online_loss": 0.0011816739570349455,
          "train_loss": 0.0011791624128818512,
          "validation_loss": 0.0009690987062640488,
          "train_metric": 0.00011876091593876481,
          "validation_metric": 9.760408283909783e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.178069383000548
        },
        {
          "epoch": 1034,
          "updates": 1034,
          "samples_seen": 1058816,
          "online_loss": 0.0011791622964665294,
          "train_loss": 0.0011766570387408137,
          "validation_loss": 0.0009676303016021848,
          "train_metric": 0.00011850857845274732,
          "validation_metric": 9.745619172463194e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.186236293999173
        },
        {
          "epoch": 1035,
          "updates": 1035,
          "samples_seen": 1059840,
          "online_loss": 0.0011766570387408137,
          "train_loss": 0.0011741560883820057,
          "validation_loss": 0.0009651974542066455,
          "train_metric": 0.0001182566920761019,
          "validation_metric": 9.721114474814385e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.195114657000886
        },
        {
          "epoch": 1036,
          "updates": 1036,
          "samples_seen": 1060864,
          "online_loss": 0.0011741560883820057,
          "train_loss": 0.0011716630542650819,
          "validation_loss": 0.0009634369052946568,
          "train_metric": 0.00011800561333075166,
          "validation_metric": 9.703384421300143e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.20343647700065
        },
        {
          "epoch": 1037,
          "updates": 1037,
          "samples_seen": 1061888,
          "online_loss": 0.0011716631706804037,
          "train_loss": 0.0011691746767610312,
          "validation_loss": 0.0009613477159291506,
          "train_metric": 0.00011775500024668872,
          "validation_metric": 9.682342351879925e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.21218703699924
        },
        {
          "epoch": 1038,
          "updates": 1038,
          "samples_seen": 1062912,
          "online_loss": 0.0011691746767610312,
          "train_loss": 0.001166693982668221,
          "validation_loss": 0.0009592529386281967,
          "train_metric": 0.00011750515841413289,
          "validation_metric": 9.661244257586077e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.220819272000881
        },
        {
          "epoch": 1039,
          "updates": 1039,
          "samples_seen": 1063936,
          "online_loss": 0.0011666940990835428,
          "train_loss": 0.0011642195750027895,
          "validation_loss": 0.0009574941359460354,
          "train_metric": 0.00011725594231393188,
          "validation_metric": 9.643531666370109e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.229633409999224
        },
        {
          "epoch": 1040,
          "updates": 1040,
          "samples_seen": 1064960,
          "online_loss": 0.0011642196914181113,
          "train_loss": 0.001161746447905898,
          "validation_loss": 0.0009551551192998886,
          "train_metric": 0.00011700684990501031,
          "validation_metric": 9.619972115615383e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.238254489999235
        },
        {
          "epoch": 1041,
          "updates": 1041,
          "samples_seen": 1065984,
          "online_loss": 0.001161746447905898,
          "train_loss": 0.0011592847295105457,
          "validation_loss": 0.0009535739081911743,
          "train_metric": 0.00011675892164930701,
          "validation_metric": 9.60404722718522e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.246956083999976
        },
        {
          "epoch": 1042,
          "updates": 1042,
          "samples_seen": 1067008,
          "online_loss": 0.0011592847295105457,
          "train_loss": 0.001156823942437768,
          "validation_loss": 0.0009511305252090096,
          "train_metric": 0.00011651108070509508,
          "validation_metric": 9.579437755746767e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.256197696000527
        },
        {
          "epoch": 1043,
          "updates": 1043,
          "samples_seen": 1068032,
          "online_loss": 0.0011568241752684116,
          "train_loss": 0.0011543702566996217,
          "validation_loss": 0.0009496064740233123,
          "train_metric": 0.00011626396008068696,
          "validation_metric": 9.564088395563886e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.366178931000832
        },
        {
          "epoch": 1044,
          "updates": 1044,
          "samples_seen": 1069056,
          "online_loss": 0.001154370023868978,
          "train_loss": 0.0011519179679453373,
          "validation_loss": 0.0009471899247728288,
          "train_metric": 0.00011601696314755827,
          "validation_metric": 9.539750317344442e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.375572905000809
        },
        {
          "epoch": 1045,
          "updates": 1045,
          "samples_seen": 1070080,
          "online_loss": 0.0011519178515300155,
          "train_loss": 0.0011494773207232356,
          "validation_loss": 0.0009455920080654323,
          "train_metric": 0.00011577115947147831,
          "validation_metric": 9.523655899101868e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.384244760000001
        },
        {
          "epoch": 1046,
          "updates": 1046,
          "samples_seen": 1071104,
          "online_loss": 0.0011494773207232356,
          "train_loss": 0.001147037255577743,
          "validation_loss": 0.0009433047962374985,
          "train_metric": 0.00011552538489922881,
          "validation_metric": 9.500620944891125e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.393112966999979
        },
        {
          "epoch": 1047,
          "updates": 1047,
          "samples_seen": 1072128,
          "online_loss": 0.0011470373719930649,
          "train_loss": 0.0011446046410128474,
          "validation_loss": 0.0009415807435289025,
          "train_metric": 0.00011528038885444403,
          "validation_metric": 9.483256872044876e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.401775360000102
        },
        {
          "epoch": 1048,
          "updates": 1048,
          "samples_seen": 1073152,
          "online_loss": 0.0011446047574281693,
          "train_loss": 0.0011421791277825832,
          "validation_loss": 0.0009394562803208828,
          "train_metric": 0.00011503610585350543,
          "validation_metric": 9.461859008297324e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.410537199999453
        },
        {
          "epoch": 1049,
          "updates": 1049,
          "samples_seen": 1074176,
          "online_loss": 0.0011421791277825832,
          "train_loss": 0.001139759086072445,
          "validation_loss": 0.0009375779191032052,
          "train_metric": 0.00011479234672151506,
          "validation_metric": 9.442940063308924e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.419384222999724
        },
        {
          "epoch": 1050,
          "updates": 1050,
          "samples_seen": 1075200,
          "online_loss": 0.001139758969657123,
          "train_loss": 0.0011373406741768122,
          "validation_loss": 0.0009356143418699503,
          "train_metric": 0.00011454878404038027,
          "validation_metric": 9.423165465705097e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.428910148999421
        },
        {
          "epoch": 1051,
          "updates": 1051,
          "samples_seen": 1076224,
          "online_loss": 0.0011373406741768122,
          "train_loss": 0.0011349328560754657,
          "validation_loss": 0.0009336092043668032,
          "train_metric": 0.00011430626182118431,
          "validation_metric": 9.402970317751169e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.437694565000129
        },
        {
          "epoch": 1052,
          "updates": 1052,
          "samples_seen": 1077248,
          "online_loss": 0.0011349327396601439,
          "train_loss": 0.001132525852881372,
          "validation_loss": 0.0009317834628745914,
          "train_metric": 0.0001140638705692254,
          "validation_metric": 9.384582517668605e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.446338887999445
        },
        {
          "epoch": 1053,
          "updates": 1053,
          "samples_seen": 1078272,
          "online_loss": 0.001132525852881372,
          "train_loss": 0.0011301260674372315,
          "validation_loss": 0.0009296748321503401,
          "train_metric": 0.000113822148705367,
          "validation_metric": 9.363343269797042e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.455230686000505
        },
        {
          "epoch": 1054,
          "updates": 1054,
          "samples_seen": 1079296,
          "online_loss": 0.0011301259510219097,
          "train_loss": 0.0011277325684204698,
          "validation_loss": 0.0009279553196392953,
          "train_metric": 0.0001135810962296091,
          "validation_metric": 9.346026490675285e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.463966028000868
        },
        {
          "epoch": 1055,
          "updates": 1055,
          "samples_seen": 1080320,
          "online_loss": 0.001127732452005148,
          "train_loss": 0.0011253461707383394,
          "validation_loss": 0.000925784872379154,
          "train_metric": 0.00011334074224578217,
          "validation_metric": 9.324167331214994e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.473095712999566
        },
        {
          "epoch": 1056,
          "updates": 1056,
          "samples_seen": 1081344,
          "online_loss": 0.0011253462871536613,
          "train_loss": 0.001122957793995738,
          "validation_loss": 0.0009241333464160562,
          "train_metric": 0.00011310019908705726,
          "validation_metric": 9.307533036917448e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.481976353999926
        },
        {
          "epoch": 1057,
          "updates": 1057,
          "samples_seen": 1082368,
          "online_loss": 0.001122957793995738,
          "train_loss": 0.0011205823393538594,
          "validation_loss": 0.0009219177300110459,
          "train_metric": 0.0001128609583247453,
          "validation_metric": 9.285217674914747e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.490827655999965
        },
        {
          "epoch": 1058,
          "updates": 1058,
          "samples_seen": 1083392,
          "online_loss": 0.0011205823393538594,
          "train_loss": 0.0011182083981111646,
          "validation_loss": 0.0009203271474689245,
          "train_metric": 0.000112621855805628,
          "validation_metric": 9.269198926631361e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.499530883000261
        },
        {
          "epoch": 1059,
          "updates": 1059,
          "samples_seen": 1084416,
          "online_loss": 0.0011182083981111646,
          "train_loss": 0.0011158434208482504,
          "validation_loss": 0.0009180731722153723,
          "train_metric": 0.00011238367733312771,
          "validation_metric": 9.246497211279348e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.508422464999967
        },
        {
          "epoch": 1060,
          "updates": 1060,
          "samples_seen": 1085440,
          "online_loss": 0.001115843653678894,
          "train_loss": 0.0011134796077385545,
          "validation_loss": 0.0009165367227979004,
          "train_metric": 0.00011214558617211878,
          "validation_metric": 9.231021977029741e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.516987585000606
        },
        {
          "epoch": 1061,
          "updates": 1061,
          "samples_seen": 1086464,
          "online_loss": 0.0011134794913232327,
          "train_loss": 0.001111120916903019,
          "validation_loss": 0.0009142388589680195,
          "train_metric": 0.00011190803343197331,
          "validation_metric": 9.207878611050546e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.526073634000568
        },
        {
          "epoch": 1062,
          "updates": 1062,
          "samples_seen": 1087488,
          "online_loss": 0.0011111211497336626,
          "train_loss": 0.0011087721213698387,
          "validation_loss": 0.0009127812227234244,
          "train_metric": 0.00011167145567014813,
          "validation_metric": 9.193197911372408e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.534706668000581
        },
        {
          "epoch": 1063,
          "updates": 1063,
          "samples_seen": 1088512,
          "online_loss": 0.0011087723542004824,
          "train_loss": 0.001106425654143095,
          "validation_loss": 0.0009104133932851255,
          "train_metric": 0.00011143513984279707,
          "validation_metric": 9.16935023269616e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.543441161000374
        },
        {
          "epoch": 1064,
          "updates": 1064,
          "samples_seen": 1089536,
          "online_loss": 0.0011064257705584168,
          "train_loss": 0.001104082795791328,
          "validation_loss": 0.0009090548846870661,
          "train_metric": 0.0001111991805373691,
          "validation_metric": 9.155667794402689e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.551949099000922
        },
        {
          "epoch": 1065,
          "updates": 1065,
          "samples_seen": 1090560,
          "online_loss": 0.001104082795791328,
          "train_loss": 0.0011017480865120888,
          "validation_loss": 0.0009065789636224508,
          "train_metric": 0.00011096402886323631,
          "validation_metric": 9.130730904871598e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.560937059000935
        },
        {
          "epoch": 1066,
          "updates": 1066,
          "samples_seen": 1091584,
          "online_loss": 0.001101747970096767,
          "train_loss": 0.0010994187323376536,
          "validation_loss": 0.0009053780813701451,
          "train_metric": 0.00011072942288592458,
          "validation_metric": 9.118636080529541e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.56960719900053
        },
        {
          "epoch": 1067,
          "updates": 1067,
          "samples_seen": 1092608,
          "online_loss": 0.0010994187323376536,
          "train_loss": 0.0010970985749736428,
          "validation_loss": 0.000902727129869163,
          "train_metric": 0.00011049575550714508,
          "validation_metric": 9.091937681660056e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.578517803000068
        },
        {
          "epoch": 1068,
          "updates": 1068,
          "samples_seen": 1093632,
          "online_loss": 0.0010970985749736428,
          "train_loss": 0.0010947799310088158,
          "validation_loss": 0.0009017709526233375,
          "train_metric": 0.00011026224092347547,
          "validation_metric": 9.082306496566162e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.587153726999532
        },
        {
          "epoch": 1069,
          "updates": 1069,
          "samples_seen": 1094656,
          "online_loss": 0.0010947796981781721,
          "train_loss": 0.0010924662929028273,
          "validation_loss": 0.0008988180779851973,
          "train_metric": 0.00011002919927705079,
          "validation_metric": 9.052566747413948e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.595687441000337
        },
        {
          "epoch": 1070,
          "updates": 1070,
          "samples_seen": 1095680,
          "online_loss": 0.0010924662929028273,
          "train_loss": 0.0010901677887886763,
          "validation_loss": 0.0008982761646620929,
          "train_metric": 0.00010979770740959793,
          "validation_metric": 9.047108324011788e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.6042411060007
        },
        {
          "epoch": 1071,
          "updates": 1071,
          "samples_seen": 1096704,
          "online_loss": 0.001090167905203998,
          "train_loss": 0.0010878756875172257,
          "validation_loss": 0.000894824683200568,
          "train_metric": 0.00010956686310237274,
          "validation_metric": 9.012346708914265e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.613363529000708
        },
        {
          "epoch": 1072,
          "updates": 1072,
          "samples_seen": 1097728,
          "online_loss": 0.0010878756875172257,
          "train_loss": 0.0010855969740077853,
          "validation_loss": 0.0008949671173468232,
          "train_metric": 0.00010933736484730616,
          "validation_metric": 9.013780072564259e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.621416248000969
        },
        {
          "epoch": 1073,
          "updates": 1073,
          "samples_seen": 1098752,
          "online_loss": 0.0010855970904231071,
          "train_loss": 0.0010833381675183773,
          "validation_loss": 0.0008906747680157423,
          "train_metric": 0.00010910985292866826,
          "validation_metric": 8.970549242803827e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.630354507000447
        },
        {
          "epoch": 1074,
          "updates": 1074,
          "samples_seen": 1099776,
          "online_loss": 0.0010833381675183773,
          "train_loss": 0.001081107184290886,
          "validation_loss": 0.0008919985266402364,
          "train_metric": 0.00010888514952966943,
          "validation_metric": 8.98388316272758e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.638920536999649
        },
        {
          "epoch": 1075,
          "updates": 1075,
          "samples_seen": 1100800,
          "online_loss": 0.001081107067875564,
          "train_loss": 0.0010789197403937578,
          "validation_loss": 0.0008862606482580304,
          "train_metric": 0.00010866484080906957,
          "validation_metric": 8.926093141781166e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.647700887999235
        },
        {
          "epoch": 1076,
          "updates": 1076,
          "samples_seen": 1101824,
          "online_loss": 0.0010789197403937578,
          "train_loss": 0.0010768062202259898,
          "validation_loss": 0.0008896754588931799,
          "train_metric": 0.00010845197539310902,
          "validation_metric": 8.960483683040366e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.65586846900078
        },
        {
          "epoch": 1077,
          "updates": 1077,
          "samples_seen": 1102848,
          "online_loss": 0.0010768062202259898,
          "train_loss": 0.0010748285567387938,
          "validation_loss": 0.0008814450120553374,
          "train_metric": 0.00010825280332937837,
          "validation_metric": 8.87759160832502e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.664931045999765
        },
        {
          "epoch": 1078,
          "updates": 1078,
          "samples_seen": 1103872,
          "online_loss": 0.0010748286731541157,
          "train_loss": 0.0010730719659477472,
          "validation_loss": 0.0008887207950465381,
          "train_metric": 0.00010807588114403188,
          "validation_metric": 8.950869960244745e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.672722880999572
        },
        {
          "epoch": 1079,
          "updates": 1079,
          "samples_seen": 1104896,
          "online_loss": 0.0010730718495324254,
          "train_loss": 0.0010717416880652308,
          "validation_loss": 0.0008761831559240818,
          "train_metric": 0.00010794190166052431,
          "validation_metric": 8.824597171042114e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.681281240000317
        },
        {
          "epoch": 1080,
          "updates": 1080,
          "samples_seen": 1105920,
          "online_loss": 0.0010717416880652308,
          "train_loss": 0.0010711421491578221,
          "validation_loss": 0.0008909597527235746,
          "train_metric": 0.00010788151121232659,
          "validation_metric": 8.973420335678384e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.782650380999257
        },
        {
          "epoch": 1081,
          "updates": 1081,
          "samples_seen": 1106944,
          "online_loss": 0.0010711421491578221,
          "train_loss": 0.0010720080463215709,
          "validation_loss": 0.0008712231065146625,
          "train_metric": 0.00010796874266816303,
          "validation_metric": 8.774640446063131e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.791111480999461
        },
        {
          "epoch": 1082,
          "updates": 1082,
          "samples_seen": 1107968,
          "online_loss": 0.0010720080463215709,
          "train_loss": 0.0010753917740657926,
          "validation_loss": 0.0009015493560582399,
          "train_metric": 0.00010830954124685377,
          "validation_metric": 9.080074960365891e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.799084975000369
        },
        {
          "epoch": 1083,
          "updates": 1083,
          "samples_seen": 1108992,
          "online_loss": 0.0010753918904811144,
          "train_loss": 0.001084139454178512,
          "validation_loss": 0.0008711380651220679,
          "train_metric": 0.00010919055057456717,
          "validation_metric": 8.773784793447703e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.808373914000185
        },
        {
          "epoch": 1084,
          "updates": 1084,
          "samples_seen": 1110016,
          "online_loss": 0.001084139570593834,
          "train_loss": 0.0011015385389328003,
          "validation_loss": 0.000936353113502264,
          "train_metric": 0.00011094293586211279,
          "validation_metric": 9.430605132365599e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.816750848000083
        },
        {
          "epoch": 1085,
          "updates": 1085,
          "samples_seen": 1111040,
          "online_loss": 0.0011015385389328003,
          "train_loss": 0.0011389929568395019,
          "validation_loss": 0.0008965483866631985,
          "train_metric": 0.00011471520701888949,
          "validation_metric": 9.029707143781707e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.825165346999711
        },
        {
          "epoch": 1086,
          "updates": 1086,
          "samples_seen": 1112064,
          "online_loss": 0.0011389931896701455,
          "train_loss": 0.001203211722895503,
          "validation_loss": 0.0010443504434078932,
          "train_metric": 0.00012118308950448409,
          "validation_metric": 0.00010518314957153052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.833234947000165
        },
        {
          "epoch": 1087,
          "updates": 1087,
          "samples_seen": 1113088,
          "online_loss": 0.0012032118393108249,
          "train_loss": 0.0013363341568037868,
          "validation_loss": 0.0010216908995062113,
          "train_metric": 0.00013459067849908024,
          "validation_metric": 0.00010290095815435052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.841038443999423
        },
        {
          "epoch": 1088,
          "updates": 1088,
          "samples_seen": 1114112,
          "online_loss": 0.0013363341568037868,
          "train_loss": 0.0015117654111236334,
          "validation_loss": 0.001338376197963953,
          "train_metric": 0.00015225948300212622,
          "validation_metric": 0.00013479635526891798,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.84856154299996
        },
        {
          "epoch": 1089,
          "updates": 1089,
          "samples_seen": 1115136,
          "online_loss": 0.0015117654111236334,
          "train_loss": 0.0018295913469046354,
          "validation_loss": 0.001373708713799715,
          "train_metric": 0.0001842697529355064,
          "validation_metric": 0.00013835490972269326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.856268969000666
        },
        {
          "epoch": 1090,
          "updates": 1090,
          "samples_seen": 1116160,
          "online_loss": 0.0018295912304893136,
          "train_loss": 0.0019524048548191786,
          "validation_loss": 0.001744953915476799,
          "train_metric": 0.0001966390700545162,
          "validation_metric": 0.00017574538651388139,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.86381061999964
        },
        {
          "epoch": 1091,
          "updates": 1091,
          "samples_seen": 1117184,
          "online_loss": 0.001952404621988535,
          "train_loss": 0.002009509364143014,
          "validation_loss": 0.0015095383860170841,
          "train_metric": 0.00020239043806213886,
          "validation_metric": 0.00015203517978079617,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.87164837400087
        },
        {
          "epoch": 1092,
          "updates": 1092,
          "samples_seen": 1118208,
          "online_loss": 0.002009509364143014,
          "train_loss": 0.0015293204924091697,
          "validation_loss": 0.001357313129119575,
          "train_metric": 0.00015402755525428802,
          "validation_metric": 0.00013670360203832388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.87922414499917
        },
        {
          "epoch": 1093,
          "updates": 1093,
          "samples_seen": 1119232,
          "online_loss": 0.0015293207252398133,
          "train_loss": 0.001126894960179925,
          "validation_loss": 0.0008778056362643838,
          "train_metric": 0.00011349672422511503,
          "validation_metric": 8.840938244247809e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.886568392999834
        },
        {
          "epoch": 1094,
          "updates": 1094,
          "samples_seen": 1120256,
          "online_loss": 0.001126894960179925,
          "train_loss": 0.001076551852747798,
          "validation_loss": 0.0008508304017595947,
          "train_metric": 0.0001084263640223071,
          "validation_metric": 8.569252531742677e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.895095201000004
        },
        {
          "epoch": 1095,
          "updates": 1095,
          "samples_seen": 1121280,
          "online_loss": 0.001076551852747798,
          "train_loss": 0.0013323313323780894,
          "validation_loss": 0.0011566225439310074,
          "train_metric": 0.0001341875467915088,
          "validation_metric": 0.00011649078805930912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.90312768700096
        },
        {
          "epoch": 1096,
          "updates": 1096,
          "samples_seen": 1122304,
          "online_loss": 0.0013323314487934113,
          "train_loss": 0.0015548731898888946,
          "validation_loss": 0.0011719041503965855,
          "train_metric": 0.00015660113422200084,
          "validation_metric": 0.0001180299004772678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.911102481999478
        },
        {
          "epoch": 1097,
          "updates": 1097,
          "samples_seen": 1123328,
          "online_loss": 0.0015548730734735727,
          "train_loss": 0.0013737098779529333,
          "validation_loss": 0.0011924117570742965,
          "train_metric": 0.0001383550261380151,
          "validation_metric": 0.00012009534111712128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.91894562699963
        },
        {
          "epoch": 1098,
          "updates": 1098,
          "samples_seen": 1124352,
          "online_loss": 0.0013737099943682551,
          "train_loss": 0.001111841993406415,
          "validation_loss": 0.0008684188942424953,
          "train_metric": 0.00011198064748896286,
          "validation_metric": 8.746396633796394e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.92684708699926
        },
        {
          "epoch": 1099,
          "updates": 1099,
          "samples_seen": 1125376,
          "online_loss": 0.0011118418769910932,
          "train_loss": 0.0010428089881315827,
          "validation_loss": 0.0008320619235746562,
          "train_metric": 0.00010502792429178953,
          "validation_metric": 8.380223880521953e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.935464927000794
        },
        {
          "epoch": 1100,
          "updates": 1100,
          "samples_seen": 1126400,
          "online_loss": 0.0010428091045469046,
          "train_loss": 0.0011977347312495112,
          "validation_loss": 0.0010310886427760124,
          "train_metric": 0.00012063145550200716,
          "validation_metric": 0.00010384748020442203,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.943408365999858
        },
        {
          "epoch": 1101,
          "updates": 1101,
          "samples_seen": 1127424,
          "online_loss": 0.001197734847664833,
          "train_loss": 0.001324369921348989,
          "validation_loss": 0.0010035393061116338,
          "train_metric": 0.00013338567805476487,
          "validation_metric": 0.00010107278649229556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.950845212999411
        },
        {
          "epoch": 1102,
          "updates": 1102,
          "samples_seen": 1128448,
          "online_loss": 0.0013243698049336672,
          "train_loss": 0.0012039697030559182,
          "validation_loss": 0.0010453399736434221,
          "train_metric": 0.00012125940702389926,
          "validation_metric": 0.0001052828156389296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.958929254000395
        },
        {
          "epoch": 1103,
          "updates": 1103,
          "samples_seen": 1129472,
          "online_loss": 0.0012039697030559182,
          "train_loss": 0.001048239995725453,
          "validation_loss": 0.0008360219653695822,
          "train_metric": 0.00010557488712947816,
          "validation_metric": 8.4201070421841e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.966759109000122
        },
        {
          "epoch": 1104,
          "updates": 1104,
          "samples_seen": 1130496,
          "online_loss": 0.001048239995725453,
          "train_loss": 0.0010410563554614782,
          "validation_loss": 0.0008349645650014281,
          "train_metric": 0.00010485140956006944,
          "validation_metric": 8.409456495428458e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.974365967000267
        },
        {
          "epoch": 1105,
          "updates": 1105,
          "samples_seen": 1131520,
          "online_loss": 0.0010410565882921219,
          "train_loss": 0.0011462531983852386,
          "validation_loss": 0.0009979354217648506,
          "train_metric": 0.00011544642620719969,
          "validation_metric": 0.00010050839773612097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.981690942000569
        },
        {
          "epoch": 1106,
          "updates": 1106,
          "samples_seen": 1132544,
          "online_loss": 0.0011462530819699168,
          "train_loss": 0.0011892211623489857,
          "validation_loss": 0.000919572776183486,
          "train_metric": 0.0001197740130010061,
          "validation_metric": 9.261599916499108e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.989415838999776
        },
        {
          "epoch": 1107,
          "updates": 1107,
          "samples_seen": 1133568,
          "online_loss": 0.0011892212787643075,
          "train_loss": 0.0010918313637375832,
          "validation_loss": 0.0009431283106096089,
          "train_metric": 0.00010996522905770689,
          "validation_metric": 9.498841973254457e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 12.997066878000624
        },
        {
          "epoch": 1108,
          "updates": 1108,
          "samples_seen": 1134592,
          "online_loss": 0.0010918311309069395,
          "train_loss": 0.0010149043519049883,
          "validation_loss": 0.0008299705805256963,
          "train_metric": 0.00010221744742011651,
          "validation_metric": 8.359159983228892e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.005493483000464
        },
        {
          "epoch": 1109,
          "updates": 1109,
          "samples_seen": 1135616,
          "online_loss": 0.0010149041190743446,
          "train_loss": 0.0010449718683958054,
          "validation_loss": 0.0008330082055181265,
          "train_metric": 0.00010524573735892773,
          "validation_metric": 8.38975392980501e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.013747662000242
        },
        {
          "epoch": 1110,
          "updates": 1110,
          "samples_seen": 1136640,
          "online_loss": 0.0010449718683958054,
          "train_loss": 0.001105652772821486,
          "validation_loss": 0.000957106938585639,
          "train_metric": 0.00011135729437228292,
          "validation_metric": 9.639629570301622e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.02167740600089
        },
        {
          "epoch": 1111,
          "updates": 1111,
          "samples_seen": 1137664,
          "online_loss": 0.0011056528892368078,
          "train_loss": 0.0011015948839485645,
          "validation_loss": 0.0008615844417363405,
          "train_metric": 0.00011094858928117901,
          "validation_metric": 8.677563164383173e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.029797844999848
        },
        {
          "epoch": 1112,
          "updates": 1112,
          "samples_seen": 1138688,
          "online_loss": 0.0011015947675332427,
          "train_loss": 0.0010338685242459178,
          "validation_loss": 0.0008748964173719287,
          "train_metric": 0.0001041274590534158,
          "validation_metric": 8.811635780148208e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.037237816000925
        },
        {
          "epoch": 1113,
          "updates": 1113,
          "samples_seen": 1139712,
          "online_loss": 0.0010338685242459178,
          "train_loss": 0.0010036991443485022,
          "validation_loss": 0.000828885065857321,
          "train_metric": 0.0001010889100143686,
          "validation_metric": 8.348225674126297e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.045921287000965
        },
        {
          "epoch": 1114,
          "updates": 1114,
          "samples_seen": 1140736,
          "online_loss": 0.0010036991443485022,
          "train_loss": 0.0010351459495723248,
          "validation_loss": 0.0008258582092821598,
          "train_metric": 0.00010425609070807695,
          "validation_metric": 8.317740866914392e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.054368764000174
        },
        {
          "epoch": 1115,
          "updates": 1115,
          "samples_seen": 1141760,
          "online_loss": 0.001035145833157003,
          "train_loss": 0.001063586212694645,
          "validation_loss": 0.0009097175789065659,
          "train_metric": 0.00010712051152950153,
          "validation_metric": 9.162341302726418e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.062488501000189
        },
        {
          "epoch": 1116,
          "updates": 1116,
          "samples_seen": 1142784,
          "online_loss": 0.0010635863291099668,
          "train_loss": 0.0010458760662004352,
          "validation_loss": 0.0008273490821011364,
          "train_metric": 0.00010533681779634207,
          "validation_metric": 8.332756988238543e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.070010322000599
        },
        {
          "epoch": 1117,
          "updates": 1117,
          "samples_seen": 1143808,
          "online_loss": 0.001045876182615757,
          "train_loss": 0.0010047301184386015,
          "validation_loss": 0.0008387473062612116,
          "train_metric": 0.0001011927452054806,
          "validation_metric": 8.447554864687845e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.078175325999837
        },
        {
          "epoch": 1118,
          "updates": 1118,
          "samples_seen": 1144832,
          "online_loss": 0.0010047301184386015,
          "train_loss": 0.0009967013029381633,
          "validation_loss": 0.0008266155491583049,
          "train_metric": 0.00010038410982815549,
          "validation_metric": 8.325369708472863e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.086351071000536
        },
        {
          "epoch": 1119,
          "updates": 1119,
          "samples_seen": 1145856,
          "online_loss": 0.0009967013029381633,
          "train_loss": 0.0010200283722952008,
          "validation_loss": 0.0008143244194798172,
          "train_metric": 0.00010273352381773293,
          "validation_metric": 8.201577293220907e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.189709932999904
        },
        {
          "epoch": 1120,
          "updates": 1120,
          "samples_seen": 1146880,
          "online_loss": 0.0010200283722952008,
          "train_loss": 0.0010295746615156531,
          "validation_loss": 0.0008755639428272843,
          "train_metric": 0.00010369499796070158,
          "validation_metric": 8.818359492579475e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.1976355099996
        },
        {
          "epoch": 1121,
          "updates": 1121,
          "samples_seen": 1147904,
          "online_loss": 0.0010295746615156531,
          "train_loss": 0.001010801293887198,
          "validation_loss": 0.0008108838810585439,
          "train_metric": 0.00010180420213146135,
          "validation_metric": 8.166925545083359e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.205842180999753
        },
        {
          "epoch": 1122,
          "updates": 1122,
          "samples_seen": 1148928,
          "online_loss": 0.0010108014103025198,
          "train_loss": 0.000988099374808371,
          "validation_loss": 0.0008203101460821927,
          "train_metric": 9.951773972716182e-05,
          "validation_metric": 8.261863695224747e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.214179370999773
        },
        {
          "epoch": 1123,
          "updates": 1123,
          "samples_seen": 1149952,
          "online_loss": 0.000988099374808371,
          "train_loss": 0.0009887319756671786,
          "validation_loss": 0.0008241281611844897,
          "train_metric": 9.958144801203161e-05,
          "validation_metric": 8.300317858811468e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.221696691000034
        },
        {
          "epoch": 1124,
          "updates": 1124,
          "samples_seen": 1150976,
          "online_loss": 0.000988731742836535,
          "train_loss": 0.0010029773693531752,
          "validation_loss": 0.0008056863443925977,
          "train_metric": 0.00010101620136993006,
          "validation_metric": 8.114578668028116e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.230554951000158
        },
        {
          "epoch": 1125,
          "updates": 1125,
          "samples_seen": 1152000,
          "online_loss": 0.0010029772529378533,
          "train_loss": 0.0010045627132058144,
          "validation_loss": 0.000850571901537478,
          "train_metric": 0.00010117587225977331,
          "validation_metric": 8.566649194108322e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.23830825100049
        },
        {
          "epoch": 1126,
          "updates": 1126,
          "samples_seen": 1153024,
          "online_loss": 0.0010045627132058144,
          "train_loss": 0.0009905511979013681,
          "validation_loss": 0.0008014547638595104,
          "train_metric": 9.976468572858721e-05,
          "validation_metric": 8.071958291111514e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.2475354649996
        },
        {
          "epoch": 1127,
          "updates": 1127,
          "samples_seen": 1154048,
          "online_loss": 0.0009905511979013681,
          "train_loss": 0.0009774493519216776,
          "validation_loss": 0.0008105084416456521,
          "train_metric": 9.84451180556789e-05,
          "validation_metric": 8.163143502315506e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.255643586000588
        },
        {
          "epoch": 1128,
          "updates": 1128,
          "samples_seen": 1155072,
          "online_loss": 0.0009774492355063558,
          "train_loss": 0.0009784901048988104,
          "validation_loss": 0.00081681739538908,
          "train_metric": 9.854994277702644e-05,
          "validation_metric": 8.226686622947454e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.263225149999926
        },
        {
          "epoch": 1129,
          "updates": 1129,
          "samples_seen": 1156096,
          "online_loss": 0.0009784899884834886,
          "train_loss": 0.000986187718808651,
          "validation_loss": 0.0007982179522514343,
          "train_metric": 9.932521788869053e-05,
          "validation_metric": 8.039358363021165e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.271805041000334
        },
        {
          "epoch": 1130,
          "updates": 1130,
          "samples_seen": 1157120,
          "online_loss": 0.000986187718808651,
          "train_loss": 0.000985836493782699,
          "validation_loss": 0.000832376885227859,
          "train_metric": 9.928984945872799e-05,
          "validation_metric": 8.383395470445976e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.280513912999595
        },
        {
          "epoch": 1131,
          "updates": 1131,
          "samples_seen": 1158144,
          "online_loss": 0.000985836610198021,
          "train_loss": 0.0009765205322764814,
          "validation_loss": 0.0007948302663862705,
          "train_metric": 9.835156379267573e-05,
          "validation_metric": 8.005238487385213e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.289125122999394
        },
        {
          "epoch": 1132,
          "updates": 1132,
          "samples_seen": 1159168,
          "online_loss": 0.0009765204158611596,
          "train_loss": 0.0009680301300249994,
          "validation_loss": 0.0008030849858187139,
          "train_metric": 9.74964423221536e-05,
          "validation_metric": 8.088378672255203e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.297195316999932
        },
        {
          "epoch": 1133,
          "updates": 1133,
          "samples_seen": 1160192,
          "online_loss": 0.0009680301882326603,
          "train_loss": 0.0009678088827058673,
          "validation_loss": 0.0008064364083111286,
          "train_metric": 9.747415606398135e-05,
          "validation_metric": 8.122131112031639e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.304569780000747
        },
        {
          "epoch": 1134,
          "updates": 1134,
          "samples_seen": 1161216,
          "online_loss": 0.0009678089991211891,
          "train_loss": 0.0009716422646306455,
          "validation_loss": 0.0007901015342213213,
          "train_metric": 9.7860254754778e-05,
          "validation_metric": 7.957612979225814e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.31337642300059
        },
        {
          "epoch": 1135,
          "updates": 1135,
          "samples_seen": 1162240,
          "online_loss": 0.0009716422064229846,
          "train_loss": 0.0009711059974506497,
          "validation_loss": 0.0008157130796462297,
          "train_metric": 9.780623076949269e-05,
          "validation_metric": 8.215563866542652e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.321018161999746
        },
        {
          "epoch": 1136,
          "updates": 1136,
          "samples_seen": 1163264,
          "online_loss": 0.0009711058810353279,
          "train_loss": 0.0009651139844208956,
          "validation_loss": 0.000786810414865613,
          "train_metric": 9.72027555690147e-05,
          "validation_metric": 7.924465899122879e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.330006423000668
        },
        {
          "epoch": 1137,
          "updates": 1137,
          "samples_seen": 1164288,
          "online_loss": 0.0009651139262132347,
          "train_loss": 0.0009590078261680901,
          "validation_loss": 0.0007947981357574463,
          "train_metric": 9.658775525167584e-05,
          "validation_metric": 8.004914707271382e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.3376388150009
        },
        {
          "epoch": 1138,
          "updates": 1138,
          "samples_seen": 1165312,
          "online_loss": 0.0009590077679604292,
          "train_loss": 0.000957476906478405,
          "validation_loss": 0.0007941590156406164,
          "train_metric": 9.643356315791607e-05,
          "validation_metric": 7.998479122761637e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.345589714999733
        },
        {
          "epoch": 1139,
          "updates": 1139,
          "samples_seen": 1166336,
          "online_loss": 0.0009574768482707441,
          "train_loss": 0.0009589383262209594,
          "validation_loss": 0.0007824557251296937,
          "train_metric": 9.6580755780451e-05,
          "validation_metric": 7.880607154220343e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.353764372999649
        },
        {
          "epoch": 1140,
          "updates": 1140,
          "samples_seen": 1167360,
          "online_loss": 0.0009589383262209594,
          "train_loss": 0.0009586719679646194,
          "validation_loss": 0.0008017686777748168,
          "train_metric": 9.655393660068512e-05,
          "validation_metric": 8.075121877482161e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.361556851999921
        },
        {
          "epoch": 1141,
          "updates": 1141,
          "samples_seen": 1168384,
          "online_loss": 0.0009586720261722803,
          "train_loss": 0.0009550469112582505,
          "validation_loss": 0.0007798500591889024,
          "train_metric": 9.618882177164778e-05,
          "validation_metric": 7.854364957893267e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.369921180999881
        },
        {
          "epoch": 1142,
          "updates": 1142,
          "samples_seen": 1169408,
          "online_loss": 0.0009550468530505896,
          "train_loss": 0.0009503577603027225,
          "validation_loss": 0.0007887116516940296,
          "train_metric": 9.571654663886875e-05,
          "validation_metric": 7.943614764371887e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.378022357999726
        },
        {
          "epoch": 1143,
          "updates": 1143,
          "samples_seen": 1170432,
          "online_loss": 0.0009503577020950615,
          "train_loss": 0.0009478045394644141,
          "validation_loss": 0.0007837685407139361,
          "train_metric": 9.545939974486828e-05,
          "validation_metric": 7.893829751992598e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.385718466999606
        },
        {
          "epoch": 1144,
          "updates": 1144,
          "samples_seen": 1171456,
          "online_loss": 0.0009478045394644141,
          "train_loss": 0.0009475745027884841,
          "validation_loss": 0.0007766972412355244,
          "train_metric": 9.543623309582472e-05,
          "validation_metric": 7.822610496077687e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.394542013000319
        },
        {
          "epoch": 1145,
          "updates": 1145,
          "samples_seen": 1172480,
          "online_loss": 0.0009475744445808232,
          "train_loss": 0.0009473048849031329,
          "validation_loss": 0.0007903052028268576,
          "train_metric": 9.540909377392381e-05,
          "validation_metric": 7.959664799273014e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.402366514999812
        },
        {
          "epoch": 1146,
          "updates": 1146,
          "samples_seen": 1173504,
          "online_loss": 0.0009473051759414375,
          "train_loss": 0.0009452418889850378,
          "validation_loss": 0.0007733027450740337,
          "train_metric": 9.520129970042035e-05,
          "validation_metric": 7.788422226440161e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.411619496999265
        },
        {
          "epoch": 1147,
          "updates": 1147,
          "samples_seen": 1174528,
          "online_loss": 0.0009452420054003596,
          "train_loss": 0.0009418183472007513,
          "validation_loss": 0.0007830926333554089,
          "train_metric": 9.485650662099943e-05,
          "validation_metric": 7.887022366048768e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.419744336999429
        },
        {
          "epoch": 1148,
          "updates": 1148,
          "samples_seen": 1175552,
          "online_loss": 0.0009418184054084122,
          "train_loss": 0.0009389036567881703,
          "validation_loss": 0.0007748313946649432,
          "train_metric": 9.456293628318235e-05,
          "validation_metric": 7.803818880347535e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.427776332999201
        },
        {
          "epoch": 1149,
          "updates": 1149,
          "samples_seen": 1176576,
          "online_loss": 0.0009389036567881703,
          "train_loss": 0.0009374339133501053,
          "validation_loss": 0.0007720233406871557,
          "train_metric": 9.44149142014794e-05,
          "validation_metric": 7.775535777909681e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.436285907999263
        },
        {
          "epoch": 1150,
          "updates": 1150,
          "samples_seen": 1177600,
          "online_loss": 0.0009374339133501053,
          "train_loss": 0.0009367068414576352,
          "validation_loss": 0.0007792481919750571,
          "train_metric": 9.434167441213503e-05,
          "validation_metric": 7.84830263000913e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.444758126999659
        },
        {
          "epoch": 1151,
          "updates": 1151,
          "samples_seen": 1178624,
          "online_loss": 0.0009367067250423133,
          "train_loss": 0.0009354506619274616,
          "validation_loss": 0.0007666010060347617,
          "train_metric": 9.421516733709723e-05,
          "validation_metric": 7.720924622844905e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.453102211000441
        },
        {
          "epoch": 1152,
          "updates": 1152,
          "samples_seen": 1179648,
          "online_loss": 0.0009354505455121398,
          "train_loss": 0.0009331515175290406,
          "validation_loss": 0.0007760699954815209,
          "train_metric": 9.398360998602584e-05,
          "validation_metric": 7.816292782081291e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.461135982999622
        },
        {
          "epoch": 1153,
          "updates": 1153,
          "samples_seen": 1180672,
          "online_loss": 0.0009331515175290406,
          "train_loss": 0.0009304629638791084,
          "validation_loss": 0.0007657574606128037,
          "train_metric": 9.371282067149878e-05,
          "validation_metric": 7.712429942330346e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.469645457000297
        },
        {
          "epoch": 1154,
          "updates": 1154,
          "samples_seen": 1181696,
          "online_loss": 0.0009304629056714475,
          "train_loss": 0.0009282809915021062,
          "validation_loss": 0.0007670159102417529,
          "train_metric": 9.349305764771998e-05,
          "validation_metric": 7.72510320530273e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.477592822999213
        },
        {
          "epoch": 1155,
          "updates": 1155,
          "samples_seen": 1182720,
          "online_loss": 0.0009282810497097671,
          "train_loss": 0.000926853041164577,
          "validation_loss": 0.0007683293661102653,
          "train_metric": 9.33492265176028e-05,
          "validation_metric": 7.738331623841077e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.485270686000149
        },
        {
          "epoch": 1156,
          "updates": 1156,
          "samples_seen": 1183744,
          "online_loss": 0.0009268529247492552,
          "train_loss": 0.0009256708435714245,
          "validation_loss": 0.0007606527069583535,
          "train_metric": 9.323017729911953e-05,
          "validation_metric": 7.661014387849718e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.494114114000695
        },
        {
          "epoch": 1157,
          "updates": 1157,
          "samples_seen": 1184768,
          "online_loss": 0.0009256707271561027,
          "train_loss": 0.0009241122752428055,
          "validation_loss": 0.000768161378800869,
          "train_metric": 9.307322034146637e-05,
          "validation_metric": 7.736639236100018e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.598467518999314
        },
        {
          "epoch": 1158,
          "updates": 1158,
          "samples_seen": 1185792,
          "online_loss": 0.0009241123916581273,
          "train_loss": 0.0009220403735525906,
          "validation_loss": 0.000757944246288389,
          "train_metric": 9.286453860113397e-05,
          "validation_metric": 7.633737550349906e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.607163966000371
        },
        {
          "epoch": 1159,
          "updates": 1159,
          "samples_seen": 1186816,
          "online_loss": 0.000922040140721947,
          "train_loss": 0.0009198079933412373,
          "validation_loss": 0.000762049516197294,
          "train_metric": 9.263969695894048e-05,
          "validation_metric": 7.675083179492503e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.615323873000307
        },
        {
          "epoch": 1160,
          "updates": 1160,
          "samples_seen": 1187840,
          "online_loss": 0.0009198079351335764,
          "train_loss": 0.0009178515756502748,
          "validation_loss": 0.0007582450052723289,
          "train_metric": 9.244264219887555e-05,
          "validation_metric": 7.636765803908929e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.623009126000397
        },
        {
          "epoch": 1161,
          "updates": 1161,
          "samples_seen": 1188864,
          "online_loss": 0.0009178514592349529,
          "train_loss": 0.0009162825299426913,
          "validation_loss": 0.0007552955066785216,
          "train_metric": 9.228462295141071e-05,
          "validation_metric": 7.607060251757503e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.631657125999482
        },
        {
          "epoch": 1162,
          "updates": 1162,
          "samples_seen": 1189888,
          "online_loss": 0.0009162825299426913,
          "train_loss": 0.000914855336304754,
          "validation_loss": 0.0007590075256302953,
          "train_metric": 9.214087185682729e-05,
          "validation_metric": 7.644446304766461e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.640029755999421
        },
        {
          "epoch": 1163,
          "updates": 1163,
          "samples_seen": 1190912,
          "online_loss": 0.000914855336304754,
          "train_loss": 0.0009132677223533392,
          "validation_loss": 0.0007512252195738256,
          "train_metric": 9.198098996421322e-05,
          "validation_metric": 7.566065323771909e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.648660253000344
        },
        {
          "epoch": 1164,
          "updates": 1164,
          "samples_seen": 1191936,
          "online_loss": 0.0009132676641456783,
          "train_loss": 0.0009114167187362909,
          "validation_loss": 0.0007562718237750232,
          "train_metric": 9.1794558102265e-05,
          "validation_metric": 7.61689298087731e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.657245413000055
        },
        {
          "epoch": 1165,
          "updates": 1165,
          "samples_seen": 1192960,
          "online_loss": 0.00091141666052863,
          "train_loss": 0.0009094400447793305,
          "validation_loss": 0.0007496878388337791,
          "train_metric": 9.15954660740681e-05,
          "validation_metric": 7.550582085968927e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.666117585000393
        },
        {
          "epoch": 1166,
          "updates": 1166,
          "samples_seen": 1193984,
          "online_loss": 0.0009094400447793305,
          "train_loss": 0.0009075519628822803,
          "validation_loss": 0.0007507505360990763,
          "train_metric": 9.140530892182142e-05,
          "validation_metric": 7.561284292023629e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.674296925000817
        },
        {
          "epoch": 1167,
          "updates": 1167,
          "samples_seen": 1195008,
          "online_loss": 0.0009075520210899413,
          "train_loss": 0.0009058574214577675,
          "validation_loss": 0.000749684520997107,
          "train_metric": 9.123464406002313e-05,
          "validation_metric": 7.550547161372378e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.682731180000701
        },
        {
          "epoch": 1168,
          "updates": 1168,
          "samples_seen": 1196032,
          "online_loss": 0.0009058574214577675,
          "train_loss": 0.0009042976889759302,
          "validation_loss": 0.00074580661021173,
          "train_metric": 9.107755613513291e-05,
          "validation_metric": 7.511490548495203e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.691364890999466
        },
        {
          "epoch": 1169,
          "updates": 1169,
          "samples_seen": 1197056,
          "online_loss": 0.000904297805391252,
          "train_loss": 0.0009027188061736524,
          "validation_loss": 0.00074897485319525,
          "train_metric": 9.091853280551732e-05,
          "validation_metric": 7.543399988207966e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.69918874799987
        },
        {
          "epoch": 1170,
          "updates": 1170,
          "samples_seen": 1198080,
          "online_loss": 0.0009027186315506697,
          "train_loss": 0.0009010254871100187,
          "validation_loss": 0.0007426809170283377,
          "train_metric": 9.074798435904086e-05,
          "validation_metric": 7.480010390281677e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.707571110999197
        },
        {
          "epoch": 1171,
          "updates": 1171,
          "samples_seen": 1199104,
          "online_loss": 0.0009010254289023578,
          "train_loss": 0.0008992255898192525,
          "validation_loss": 0.0007457533502019942,
          "train_metric": 9.056669659912586e-05,
          "validation_metric": 7.5109550380148e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.715350878999743
        },
        {
          "epoch": 1172,
          "updates": 1172,
          "samples_seen": 1200128,
          "online_loss": 0.0008992255316115916,
          "train_loss": 0.0008974037482403219,
          "validation_loss": 0.0007411409169435501,
          "train_metric": 9.0383218775969e-05,
          "validation_metric": 7.464500231435522e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.72423188600078
        },
        {
          "epoch": 1173,
          "updates": 1173,
          "samples_seen": 1201152,
          "online_loss": 0.0008974035736173391,
          "train_loss": 0.0008956508827395737,
          "validation_loss": 0.0007411072147078812,
          "train_metric": 9.020668221637607e-05,
          "validation_metric": 7.464159716619179e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.73276735500076
        },
        {
          "epoch": 1174,
          "updates": 1174,
          "samples_seen": 1202176,
          "online_loss": 0.0008956509991548955,
          "train_loss": 0.0008939887629821897,
          "validation_loss": 0.000740302843041718,
          "train_metric": 9.003927698358893e-05,
          "validation_metric": 7.456059393007308e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.74173833399982
        },
        {
          "epoch": 1175,
          "updates": 1175,
          "samples_seen": 1203200,
          "online_loss": 0.0008939887629821897,
          "train_loss": 0.0008923785062506795,
          "validation_loss": 0.0007368740043602884,
          "train_metric": 8.98771031643264e-05,
          "validation_metric": 7.421524787787348e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.750254063999819
        },
        {
          "epoch": 1176,
          "updates": 1176,
          "samples_seen": 1204224,
          "online_loss": 0.0008923785062506795,
          "train_loss": 0.0008907568990252912,
          "validation_loss": 0.0007387757068499923,
          "train_metric": 8.971377246780321e-05,
          "validation_metric": 7.440678018610924e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.758024773000216
        },
        {
          "epoch": 1177,
          "updates": 1177,
          "samples_seen": 1205248,
          "online_loss": 0.0008907568990252912,
          "train_loss": 0.0008890852914191782,
          "validation_loss": 0.0007338178111240268,
          "train_metric": 8.9545406808611e-05,
          "validation_metric": 7.390744576696306e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.766303270000208
        },
        {
          "epoch": 1178,
          "updates": 1178,
          "samples_seen": 1206272,
          "online_loss": 0.0008890852332115173,
          "train_loss": 0.0008873663027770817,
          "validation_loss": 0.0007357556605711579,
          "train_metric": 8.937227539718151e-05,
          "validation_metric": 7.41026087780483e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.774700326999664
        },
        {
          "epoch": 1179,
          "updates": 1179,
          "samples_seen": 1207296,
          "online_loss": 0.0008873663027770817,
          "train_loss": 0.0008856340427882969,
          "validation_loss": 0.0007318346761167049,
          "train_metric": 8.919783431338146e-05,
          "validation_metric": 7.370769890258089e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.783430384000894
        },
        {
          "epoch": 1180,
          "updates": 1180,
          "samples_seen": 1208320,
          "online_loss": 0.000885633984580636,
          "train_loss": 0.0008839222136884928,
          "validation_loss": 0.000731891777832061,
          "train_metric": 8.902540866984054e-05,
          "validation_metric": 7.37134541850537e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.791707046999363
        },
        {
          "epoch": 1181,
          "updates": 1181,
          "samples_seen": 1209344,
          "online_loss": 0.0008839222718961537,
          "train_loss": 0.0008822567178867757,
          "validation_loss": 0.0007304386235773563,
          "train_metric": 8.885767601896077e-05,
          "validation_metric": 7.35670982976444e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.800258315000974
        },
        {
          "epoch": 1182,
          "updates": 1182,
          "samples_seen": 1210368,
          "online_loss": 0.0008822567760944366,
          "train_loss": 0.0008806241676211357,
          "validation_loss": 0.0007281602011062205,
          "train_metric": 8.869323210092261e-05,
          "validation_metric": 7.333761459449306e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.809702174999984
        },
        {
          "epoch": 1183,
          "updates": 1183,
          "samples_seen": 1211392,
          "online_loss": 0.0008806241094134748,
          "train_loss": 0.0008789977873675525,
          "validation_loss": 0.0007287607877515256,
          "train_metric": 8.852945029502735e-05,
          "validation_metric": 7.3398114182055e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.817889841000579
        },
        {
          "epoch": 1184,
          "updates": 1184,
          "samples_seen": 1212416,
          "online_loss": 0.0008789976709522307,
          "train_loss": 0.0008773649460636079,
          "validation_loss": 0.0007250539492815733,
          "train_metric": 8.836498454911634e-05,
          "validation_metric": 7.302477752091363e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.826712968999345
        },
        {
          "epoch": 1185,
          "updates": 1185,
          "samples_seen": 1213440,
          "online_loss": 0.0008773649460636079,
          "train_loss": 0.0008757094037719071,
          "validation_loss": 0.0007262564613483846,
          "train_metric": 8.81982414284721e-05,
          "validation_metric": 7.314589311135933e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.834533489000023
        },
        {
          "epoch": 1186,
          "updates": 1186,
          "samples_seen": 1214464,
          "online_loss": 0.0008757094037719071,
          "train_loss": 0.0008740425109863281,
          "validation_loss": 0.0007226028246805072,
          "train_metric": 8.803035598248243e-05,
          "validation_metric": 7.2777911555022e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.84345222400043
        },
        {
          "epoch": 1187,
          "updates": 1187,
          "samples_seen": 1215488,
          "online_loss": 0.0008740424527786672,
          "train_loss": 0.0008723711362108588,
          "validation_loss": 0.0007230645278468728,
          "train_metric": 8.786201942712069e-05,
          "validation_metric": 7.282441220013425e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.850845945000401
        },
        {
          "epoch": 1188,
          "updates": 1188,
          "samples_seen": 1216512,
          "online_loss": 0.0008723710780031979,
          "train_loss": 0.0008707069791853428,
          "validation_loss": 0.0007206521113403141,
          "train_metric": 8.769442501943558e-05,
          "validation_metric": 7.258144614752382e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.859217332999833
        },
        {
          "epoch": 1189,
          "updates": 1189,
          "samples_seen": 1217536,
          "online_loss": 0.0008707069791853428,
          "train_loss": 0.0008690670365467668,
          "validation_loss": 0.0007196744554676116,
          "train_metric": 8.7529253505636e-05,
          "validation_metric": 7.248298061313108e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.868563882999297
        },
        {
          "epoch": 1190,
          "updates": 1190,
          "samples_seen": 1218560,
          "online_loss": 0.0008690670365467668,
          "train_loss": 0.0008674364071339369,
          "validation_loss": 0.0007187615847215056,
          "train_metric": 8.736502059036866e-05,
          "validation_metric": 7.239103433676064e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.877984058000948
        },
        {
          "epoch": 1191,
          "updates": 1191,
          "samples_seen": 1219584,
          "online_loss": 0.0008674364071339369,
          "train_loss": 0.0008658167789690197,
          "validation_loss": 0.000716446025762707,
          "train_metric": 8.720188634470105e-05,
          "validation_metric": 7.21578107913956e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.88682353799959
        },
        {
          "epoch": 1192,
          "updates": 1192,
          "samples_seen": 1220608,
          "online_loss": 0.0008658167207613587,
          "train_loss": 0.0008642011089250445,
          "validation_loss": 0.0007165821734815836,
          "train_metric": 8.703915955265984e-05,
          "validation_metric": 7.217153324745595e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.89471912800036
        },
        {
          "epoch": 1193,
          "updates": 1193,
          "samples_seen": 1221632,
          "online_loss": 0.0008642011089250445,
          "train_loss": 0.000862582353875041,
          "validation_loss": 0.0007135794730857015,
          "train_metric": 8.687612717039883e-05,
          "validation_metric": 7.186910806922242e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.903239129000212
        },
        {
          "epoch": 1194,
          "updates": 1194,
          "samples_seen": 1222656,
          "online_loss": 0.000862582353875041,
          "train_loss": 0.0008609554497525096,
          "validation_loss": 0.0007140054367482662,
          "train_metric": 8.671227988088503e-05,
          "validation_metric": 7.191201439127326e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 13.911312170999736
        },
        {
          "epoch": 1195,
          "updates": 1195,
          "samples_seen": 1223680,
          "online_loss": 0.0008609554497525096,
          "train_loss": 0.0008593271486461163,
          "validation_loss": 0.0007110992446541786,
          "train_metric": 8.654829434817657e-05,
          "validation_metric": 7.161930261645466e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.019905024000764
        },
        {
          "epoch": 1196,
          "updates": 1196,
          "samples_seen": 1224704,
          "online_loss": 0.0008593272068537772,
          "train_loss": 0.0008577007101848722,
          "validation_loss": 0.0007111331215128303,
          "train_metric": 8.638446888653561e-05,
          "validation_metric": 7.162272231653333e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.028428722000172
        },
        {
          "epoch": 1197,
          "updates": 1197,
          "samples_seen": 1225728,
          "online_loss": 0.0008577007101848722,
          "train_loss": 0.0008560775313526392,
          "validation_loss": 0.0007088452111929655,
          "train_metric": 8.622099994681776e-05,
          "validation_metric": 7.139228546293452e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.037019180999778
        },
        {
          "epoch": 1198,
          "updates": 1198,
          "samples_seen": 1226752,
          "online_loss": 0.0008560775313526392,
          "train_loss": 0.0008544618613086641,
          "validation_loss": 0.0007081069052219391,
          "train_metric": 8.605828043073416e-05,
          "validation_metric": 7.13179397280328e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.045981691999259
        },
        {
          "epoch": 1199,
          "updates": 1199,
          "samples_seen": 1227776,
          "online_loss": 0.0008544619777239859,
          "train_loss": 0.0008528507314622402,
          "validation_loss": 0.0007066528196446598,
          "train_metric": 8.589600474806502e-05,
          "validation_metric": 7.11714819772169e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.054810334999274
        },
        {
          "epoch": 1200,
          "updates": 1200,
          "samples_seen": 1228800,
          "online_loss": 0.0008528507314622402,
          "train_loss": 0.0008512507192790508,
          "validation_loss": 0.0007051177090033889,
          "train_metric": 8.573484956286848e-05,
          "validation_metric": 7.10168678779155e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.063778961000935
        },
        {
          "epoch": 1201,
          "updates": 1201,
          "samples_seen": 1229824,
          "online_loss": 0.0008512506610713899,
          "train_loss": 0.0008496532682329416,
          "validation_loss": 0.0007044006488285959,
          "train_metric": 8.557396358810365e-05,
          "validation_metric": 7.094464672263712e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.072681358000409
        },
        {
          "epoch": 1202,
          "updates": 1202,
          "samples_seen": 1230848,
          "online_loss": 0.0008496532682329416,
          "train_loss": 0.0008480592514388263,
          "validation_loss": 0.0007022925419732928,
          "train_metric": 8.54134195833467e-05,
          "validation_metric": 7.073233427945524e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.081741141000748
        },
        {
          "epoch": 1203,
          "updates": 1203,
          "samples_seen": 1231872,
          "online_loss": 0.0008480591932311654,
          "train_loss": 0.0008464653510600328,
          "validation_loss": 0.000702027406077832,
          "train_metric": 8.525288285454735e-05,
          "validation_metric": 7.070562423905358e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.090292169999884
        },
        {
          "epoch": 1204,
          "updates": 1204,
          "samples_seen": 1232896,
          "online_loss": 0.0008464654674753547,
          "train_loss": 0.0008448748849332333,
          "validation_loss": 0.0006996378069743514,
          "train_metric": 8.509270992362872e-05,
          "validation_metric": 7.046495011309162e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.098947602000408
        },
        {
          "epoch": 1205,
          "updates": 1205,
          "samples_seen": 1233920,
          "online_loss": 0.0008448748849332333,
          "train_loss": 0.0008432852919213474,
          "validation_loss": 0.0006994996219873428,
          "train_metric": 8.493260975228623e-05,
          "validation_metric": 7.045103120617568e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.107665171999543
        },
        {
          "epoch": 1206,
          "updates": 1206,
          "samples_seen": 1234944,
          "online_loss": 0.0008432852337136865,
          "train_loss": 0.0008416971541009843,
          "validation_loss": 0.0006971225957386196,
          "train_metric": 8.477266965201125e-05,
          "validation_metric": 7.021164492471144e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.116415937000056
        },
        {
          "epoch": 1207,
          "updates": 1207,
          "samples_seen": 1235968,
          "online_loss": 0.0008416972705163062,
          "train_loss": 0.0008401103550568223,
          "validation_loss": 0.0006968709640204906,
          "train_metric": 8.461284596705809e-05,
          "validation_metric": 7.018628821242601e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.125065868000092
        },
        {
          "epoch": 1208,
          "updates": 1208,
          "samples_seen": 1236992,
          "online_loss": 0.0008401103550568223,
          "train_loss": 0.0008385276887565851,
          "validation_loss": 0.0006946988287381828,
          "train_metric": 8.445344428764656e-05,
          "validation_metric": 6.996752199484035e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.134013802000482
        },
        {
          "epoch": 1209,
          "updates": 1209,
          "samples_seen": 1238016,
          "online_loss": 0.0008385276887565851,
          "train_loss": 0.0008369477582164109,
          "validation_loss": 0.000694186775945127,
          "train_metric": 8.429431909462437e-05,
          "validation_metric": 6.991595000727102e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.14278432500032
        },
        {
          "epoch": 1210,
          "updates": 1210,
          "samples_seen": 1239040,
          "online_loss": 0.00083694770000875,
          "train_loss": 0.0008353690500371158,
          "validation_loss": 0.0006923007313162088,
          "train_metric": 8.4135310316924e-05,
          "validation_metric": 6.972599658183753e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.151381518999187
        },
        {
          "epoch": 1211,
          "updates": 1211,
          "samples_seen": 1240064,
          "online_loss": 0.0008353689918294549,
          "train_loss": 0.0008337947656400502,
          "validation_loss": 0.0006915007252246141,
          "train_metric": 8.397675264859572e-05,
          "validation_metric": 6.964542262721807e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.16491950300042
        },
        {
          "epoch": 1212,
          "updates": 1212,
          "samples_seen": 1241088,
          "online_loss": 0.0008337948238477111,
          "train_loss": 0.0008322257781401277,
          "validation_loss": 0.0006898943684063852,
          "train_metric": 8.381872612517327e-05,
          "validation_metric": 6.94836417096667e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.19021161800083
        },
        {
          "epoch": 1213,
          "updates": 1213,
          "samples_seen": 1242112,
          "online_loss": 0.0008322256035171449,
          "train_loss": 0.0008306577801704407,
          "validation_loss": 0.0006888163625262678,
          "train_metric": 8.366080146515742e-05,
          "validation_metric": 6.937506259419024e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.212629272000413
        },
        {
          "epoch": 1214,
          "updates": 1214,
          "samples_seen": 1243136,
          "online_loss": 0.0008306577801704407,
          "train_loss": 0.0008290950208902359,
          "validation_loss": 0.000687481660861522,
          "train_metric": 8.35034079500474e-05,
          "validation_metric": 6.924063200131059e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.231412384000578
        },
        {
          "epoch": 1215,
          "updates": 1215,
          "samples_seen": 1244160,
          "online_loss": 0.000829094962682575,
          "train_loss": 0.0008275338914245367,
          "validation_loss": 0.000686168554238975,
          "train_metric": 8.334618178196251e-05,
          "validation_metric": 6.910837691975757e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.250120037999295
        },
        {
          "epoch": 1216,
          "updates": 1216,
          "samples_seen": 1245184,
          "online_loss": 0.0008275339496321976,
          "train_loss": 0.0008259774185717106,
          "validation_loss": 0.0006850624922662973,
          "train_metric": 8.318941399920732e-05,
          "validation_metric": 6.899698200868443e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.273178560000815
        },
        {
          "epoch": 1217,
          "updates": 1217,
          "samples_seen": 1246208,
          "online_loss": 0.0008259773021563888,
          "train_loss": 0.0008244237396866083,
          "validation_loss": 0.0006835515960119665,
          "train_metric": 8.303293725475669e-05,
          "validation_metric": 6.884480535518378e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.292029516000184
        },
        {
          "epoch": 1218,
          "updates": 1218,
          "samples_seen": 1247232,
          "online_loss": 0.0008244236814789474,
          "train_loss": 0.0008228719816543162,
          "validation_loss": 0.0006826332537457347,
          "train_metric": 8.287664968520403e-05,
          "validation_metric": 6.87523206579499e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.309229948999928
        },
        {
          "epoch": 1219,
          "updates": 1219,
          "samples_seen": 1248256,
          "online_loss": 0.0008228720398619771,
          "train_loss": 0.0008213235414586961,
          "validation_loss": 0.0006809556507505476,
          "train_metric": 8.272071136161685e-05,
          "validation_metric": 6.858335109427571e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.331057875000624
        },
        {
          "epoch": 1220,
          "updates": 1220,
          "samples_seen": 1249280,
          "online_loss": 0.000821323657874018,
          "train_loss": 0.0008197785937227309,
          "validation_loss": 0.000680206750985235,
          "train_metric": 8.256508590420708e-05,
          "validation_metric": 6.85079357936047e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.347119731999555
        },
        {
          "epoch": 1221,
          "updates": 1221,
          "samples_seen": 1250304,
          "online_loss": 0.00081977853551507,
          "train_loss": 0.0008182352758012712,
          "validation_loss": 0.0006783780409023166,
          "train_metric": 8.240966417361051e-05,
          "validation_metric": 6.832375947851688e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.355686982000407
        },
        {
          "epoch": 1222,
          "updates": 1222,
          "samples_seen": 1251328,
          "online_loss": 0.0008182353340089321,
          "train_loss": 0.0008166951593011618,
          "validation_loss": 0.0006778050446882844,
          "train_metric": 8.225454803323373e-05,
          "validation_metric": 6.826603930676356e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.364711526000974
        },
        {
          "epoch": 1223,
          "updates": 1223,
          "samples_seen": 1252352,
          "online_loss": 0.0008166952175088227,
          "train_loss": 0.0008151595829986036,
          "validation_loss": 0.0006758064264431596,
          "train_metric": 8.20998684503138e-05,
          "validation_metric": 6.806474993936718e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.37346472399986
        },
        {
          "epoch": 1224,
          "updates": 1224,
          "samples_seen": 1253376,
          "online_loss": 0.0008151595829986036,
          "train_loss": 0.0008136298856697977,
          "validation_loss": 0.0006754336063750088,
          "train_metric": 8.194582915166393e-05,
          "validation_metric": 6.802720599807799e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.38219206299982
        },
        {
          "epoch": 1225,
          "updates": 1225,
          "samples_seen": 1254400,
          "online_loss": 0.0008136300020851195,
          "train_loss": 0.0008121002465486526,
          "validation_loss": 0.0006732072215527296,
          "train_metric": 8.179176802514121e-05,
          "validation_metric": 6.780296826036647e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.390745044000141
        },
        {
          "epoch": 1226,
          "updates": 1226,
          "samples_seen": 1255424,
          "online_loss": 0.0008121002465486526,
          "train_loss": 0.0008105739834718406,
          "validation_loss": 0.000673112750519067,
          "train_metric": 8.163804886862636e-05,
          "validation_metric": 6.779345858376473e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.399584982000306
        },
        {
          "epoch": 1227,
          "updates": 1227,
          "samples_seen": 1256448,
          "online_loss": 0.0008105740416795015,
          "train_loss": 0.0008090533083304763,
          "validation_loss": 0.0006705645355395973,
          "train_metric": 8.14848899608478e-05,
          "validation_metric": 6.753679917892441e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.408301913999821
        },
        {
          "epoch": 1228,
          "updates": 1228,
          "samples_seen": 1257472,
          "online_loss": 0.0008090532501228154,
          "train_loss": 0.0008075378718785942,
          "validation_loss": 0.0006708856672048569,
          "train_metric": 8.133225492201746e-05,
          "validation_metric": 6.756914808647707e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.416002724999998
        },
        {
          "epoch": 1229,
          "updates": 1229,
          "samples_seen": 1258496,
          "online_loss": 0.0008075378718785942,
          "train_loss": 0.000806029886007309,
          "validation_loss": 0.0006678439094685018,
          "train_metric": 8.118036930682138e-05,
          "validation_metric": 6.726279389113188e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.424760829000661
        },
        {
          "epoch": 1230,
          "updates": 1230,
          "samples_seen": 1259520,
          "online_loss": 0.0008060299442149699,
          "train_loss": 0.0008045344147831202,
          "validation_loss": 0.0006688187131658196,
          "train_metric": 8.10297642601654e-05,
          "validation_metric": 6.736097566317767e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.432565627999793
        },
        {
          "epoch": 1231,
          "updates": 1231,
          "samples_seen": 1260544,
          "online_loss": 0.0008045344729907811,
          "train_loss": 0.0008030534954741597,
          "validation_loss": 0.0006649846909567714,
          "train_metric": 8.08805925771594e-05,
          "validation_metric": 6.697481876472011e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.441384265999659
        },
        {
          "epoch": 1232,
          "updates": 1232,
          "samples_seen": 1261568,
          "online_loss": 0.0008030534954741597,
          "train_loss": 0.0008015880594030023,
          "validation_loss": 0.0006670023431070149,
          "train_metric": 8.073300705291331e-05,
          "validation_metric": 6.717803626088426e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.549917848000405
        },
        {
          "epoch": 1233,
          "updates": 1233,
          "samples_seen": 1262592,
          "online_loss": 0.0008015878847800195,
          "train_loss": 0.0008001554524526,
          "validation_loss": 0.0006618993356823921,
          "train_metric": 8.058871753746644e-05,
          "validation_metric": 6.666407716693357e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.558978180999475
        },
        {
          "epoch": 1234,
          "updates": 1234,
          "samples_seen": 1263616,
          "online_loss": 0.0008001554524526,
          "train_loss": 0.0007987763965502381,
          "validation_loss": 0.0006656686891801655,
          "train_metric": 8.044984133448452e-05,
          "validation_metric": 6.704371480736881e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.567301869999937
        },
        {
          "epoch": 1235,
          "updates": 1235,
          "samples_seen": 1264640,
          "online_loss": 0.0007987763965502381,
          "train_loss": 0.0007974784239195287,
          "validation_loss": 0.0006584858056157827,
          "train_metric": 8.031910692807287e-05,
          "validation_metric": 6.632028089370579e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.576156060000358
        },
        {
          "epoch": 1236,
          "updates": 1236,
          "samples_seen": 1265664,
          "online_loss": 0.0007974784239195287,
          "train_loss": 0.0007963198586367071,
          "validation_loss": 0.0006652817828580737,
          "train_metric": 8.020240784389898e-05,
          "validation_metric": 6.700474477838725e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.584275761000754
        },
        {
          "epoch": 1237,
          "updates": 1237,
          "samples_seen": 1266688,
          "online_loss": 0.0007963197422213852,
          "train_loss": 0.0007954066968522966,
          "validation_loss": 0.000654644041787833,
          "train_metric": 8.01104397396557e-05,
          "validation_metric": 6.593334546778351e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.592792130999442
        },
        {
          "epoch": 1238,
          "updates": 1238,
          "samples_seen": 1267712,
          "online_loss": 0.0007954066386446357,
          "train_loss": 0.0007948851562105119,
          "validation_loss": 0.000666903390083462,
          "train_metric": 8.005792187759653e-05,
          "validation_metric": 6.716806819895282e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.600342784000532
        },
        {
          "epoch": 1239,
          "updates": 1239,
          "samples_seen": 1268736,
          "online_loss": 0.0007948852144181728,
          "train_loss": 0.0007951295701786876,
          "validation_loss": 0.0006505743367597461,
          "train_metric": 8.008254371816292e-05,
          "validation_metric": 6.55234616715461e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.608679352000763
        },
        {
          "epoch": 1240,
          "updates": 1240,
          "samples_seen": 1269760,
          "online_loss": 0.0007951295119710267,
          "train_loss": 0.0007965927943587303,
          "validation_loss": 0.0006731958710588515,
          "train_metric": 8.022990368772298e-05,
          "validation_metric": 6.780181865906343e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.616382927999439
        },
        {
          "epoch": 1241,
          "updates": 1241,
          "samples_seen": 1270784,
          "online_loss": 0.0007965927943587303,
          "train_loss": 0.0008005439303815365,
          "validation_loss": 0.0006479498115368187,
          "train_metric": 8.062785491347313e-05,
          "validation_metric": 6.525913340738043e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.624851557000511
        },
        {
          "epoch": 1242,
          "updates": 1242,
          "samples_seen": 1271808,
          "online_loss": 0.0008005439885891974,
          "train_loss": 0.0008083009161055088,
          "validation_loss": 0.0006914298865012825,
          "train_metric": 8.140910358633846e-05,
          "validation_metric": 6.963828491279855e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.632960904000356
        },
        {
          "epoch": 1243,
          "updates": 1243,
          "samples_seen": 1272832,
          "online_loss": 0.0008083009161055088,
          "train_loss": 0.0008245599456131458,
          "validation_loss": 0.0006542927585542202,
          "train_metric": 8.304664515890181e-05,
          "validation_metric": 6.589796976186335e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.640664594000555
        },
        {
          "epoch": 1244,
          "updates": 1244,
          "samples_seen": 1273856,
          "online_loss": 0.0008245598874054849,
          "train_loss": 0.000852334254886955,
          "validation_loss": 0.0007425511139445007,
          "train_metric": 8.584396709920838e-05,
          "validation_metric": 7.478702900698408e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.648458009000024
        },
        {
          "epoch": 1245,
          "updates": 1245,
          "samples_seen": 1274880,
          "online_loss": 0.000852334254886955,
          "train_loss": 0.0009092267137020826,
          "validation_loss": 0.0006980984471738338,
          "train_metric": 9.157398017123342e-05,
          "validation_metric": 7.030992128420621e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.656288909000068
        },
        {
          "epoch": 1246,
          "updates": 1246,
          "samples_seen": 1275904,
          "online_loss": 0.0009092266554944217,
          "train_loss": 0.0009937923168763518,
          "validation_loss": 0.000883033499121666,
          "train_metric": 0.00010009113611886278,
          "validation_metric": 8.893587801139802e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.664165210000647
        },
        {
          "epoch": 1247,
          "updates": 1247,
          "samples_seen": 1276928,
          "online_loss": 0.0009937923168763518,
          "train_loss": 0.0011651143431663513,
          "validation_loss": 0.000861270644236356,
          "train_metric": 0.00011734602594515309,
          "validation_metric": 8.67440176079981e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.672340618999442
        },
        {
          "epoch": 1248,
          "updates": 1248,
          "samples_seen": 1277952,
          "online_loss": 0.0011651142267510295,
          "train_loss": 0.0013385550118982792,
          "validation_loss": 0.0011995772365480661,
          "train_metric": 0.00013481434143614024,
          "validation_metric": 0.00012081702152499929,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.680177363999974
        },
        {
          "epoch": 1249,
          "updates": 1249,
          "samples_seen": 1278976,
          "online_loss": 0.0013385548954829574,
          "train_loss": 0.001645239070057869,
          "validation_loss": 0.0011957708047702909,
          "train_metric": 0.00016570245497860014,
          "validation_metric": 0.0001204336658702232,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.68790207299935
        },
        {
          "epoch": 1250,
          "updates": 1250,
          "samples_seen": 1280000,
          "online_loss": 0.001645239070057869,
          "train_loss": 0.0016293313819915056,
          "validation_loss": 0.0014647949719801545,
          "train_metric": 0.00016410028911195695,
          "validation_metric": 0.0001475287863286212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.69546278100097
        },
        {
          "epoch": 1251,
          "updates": 1251,
          "samples_seen": 1281024,
          "online_loss": 0.0016293313819915056,
          "train_loss": 0.0015467330813407898,
          "validation_loss": 0.001125172944739461,
          "train_metric": 0.00015578132297378033,
          "validation_metric": 0.00011332330905133858,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.702976196000236
        },
        {
          "epoch": 1252,
          "updates": 1252,
          "samples_seen": 1282048,
          "online_loss": 0.0015467333141714334,
          "train_loss": 0.001083094161003828,
          "validation_loss": 0.0009649295825511217,
          "train_metric": 0.00010908527474384755,
          "validation_metric": 9.718418004922569e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.710689842999273
        },
        {
          "epoch": 1253,
          "updates": 1253,
          "samples_seen": 1283072,
          "online_loss": 0.0010830942774191499,
          "train_loss": 0.0007996511412784457,
          "validation_loss": 0.0006290386081673205,
          "train_metric": 8.053793135331944e-05,
          "validation_metric": 6.335446232696995e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.71915645000081
        },
        {
          "epoch": 1254,
          "updates": 1254,
          "samples_seen": 1284096,
          "online_loss": 0.000799650966655463,
          "train_loss": 0.0008434942574240267,
          "validation_loss": 0.0006471911910921335,
          "train_metric": 8.495365182170644e-05,
          "validation_metric": 6.518272857647389e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.727481067999179
        },
        {
          "epoch": 1255,
          "updates": 1255,
          "samples_seen": 1285120,
          "online_loss": 0.0008434941992163658,
          "train_loss": 0.0010751434601843357,
          "validation_loss": 0.0009390387567691505,
          "train_metric": 0.00010828449740074575,
          "validation_metric": 9.457653504796326e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.735214484000608
        },
        {
          "epoch": 1256,
          "updates": 1256,
          "samples_seen": 1286144,
          "online_loss": 0.0010751434601843357,
          "train_loss": 0.0012543316697701812,
          "validation_loss": 0.0009171679266728461,
          "train_metric": 0.00012633168080355972,
          "validation_metric": 9.237379708793014e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.742877573000442
        },
        {
          "epoch": 1257,
          "updates": 1257,
          "samples_seen": 1287168,
          "online_loss": 0.0012543315533548594,
          "train_loss": 0.0010904670925810933,
          "validation_loss": 0.0009556670556776226,
          "train_metric": 0.00010982784442603588,
          "validation_metric": 9.625129314372316e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.759254275000785
        },
        {
          "epoch": 1258,
          "updates": 1258,
          "samples_seen": 1288192,
          "online_loss": 0.0010904670925810933,
          "train_loss": 0.000866179121658206,
          "validation_loss": 0.0006631013238802552,
          "train_metric": 8.723836799617857e-05,
          "validation_metric": 6.678514910163358e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.776845172000321
        },
        {
          "epoch": 1259,
          "updates": 1259,
          "samples_seen": 1289216,
          "online_loss": 0.0008661790052428842,
          "train_loss": 0.0007690906641073525,
          "validation_loss": 0.0006210622959770262,
          "train_metric": 7.745999755570665e-05,
          "validation_metric": 6.255111657083035e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.792197013999612
        },
        {
          "epoch": 1260,
          "updates": 1260,
          "samples_seen": 1290240,
          "online_loss": 0.0007690907223150134,
          "train_loss": 0.0008734007133170962,
          "validation_loss": 0.000756793946493417,
          "train_metric": 8.796572365099564e-05,
          "validation_metric": 7.62215131544508e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.800154247999671
        },
        {
          "epoch": 1261,
          "updates": 1261,
          "samples_seen": 1291264,
          "online_loss": 0.0008734006551094353,
          "train_loss": 0.0010157788638025522,
          "validation_loss": 0.0007504000677727163,
          "train_metric": 0.00010230554471490905,
          "validation_metric": 7.557753997389227e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.807863251000526
        },
        {
          "epoch": 1262,
          "updates": 1262,
          "samples_seen": 1292288,
          "online_loss": 0.001015778980217874,
          "train_loss": 0.0009767726296558976,
          "validation_loss": 0.0008655106648802757,
          "train_metric": 9.837695688474923e-05,
          "validation_metric": 8.717106538824737e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.81557552599952
        },
        {
          "epoch": 1263,
          "updates": 1263,
          "samples_seen": 1293312,
          "online_loss": 0.0009767725132405758,
          "train_loss": 0.000848764437250793,
          "validation_loss": 0.0006561644258908927,
          "train_metric": 8.548444020561874e-05,
          "validation_metric": 6.608648254768923e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.823085097999865
        },
        {
          "epoch": 1264,
          "updates": 1264,
          "samples_seen": 1294336,
          "online_loss": 0.0008487644954584539,
          "train_loss": 0.000761759583838284,
          "validation_loss": 0.0006382816936820745,
          "train_metric": 7.67216261010617e-05,
          "validation_metric": 6.428539199987426e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.83070067799963
        },
        {
          "epoch": 1265,
          "updates": 1265,
          "samples_seen": 1295360,
          "online_loss": 0.0007617594092153013,
          "train_loss": 0.0008063536370173097,
          "validation_loss": 0.0007024866645224392,
          "train_metric": 8.121298014884815e-05,
          "validation_metric": 7.075189205352217e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.837985403999483
        },
        {
          "epoch": 1266,
          "updates": 1266,
          "samples_seen": 1296384,
          "online_loss": 0.0008063535788096488,
          "train_loss": 0.0008948014583438635,
          "validation_loss": 0.0006801821291446686,
          "train_metric": 9.012112423079088e-05,
          "validation_metric": 6.850545469205827e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.845710441999472
        },
        {
          "epoch": 1267,
          "updates": 1267,
          "samples_seen": 1297408,
          "online_loss": 0.0008948013419285417,
          "train_loss": 0.0008879235247150064,
          "validation_loss": 0.0007880888879299164,
          "train_metric": 8.942840213421732e-05,
          "validation_metric": 7.937342888908461e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.853480174000651
        },
        {
          "epoch": 1268,
          "updates": 1268,
          "samples_seen": 1298432,
          "online_loss": 0.0008879235247150064,
          "train_loss": 0.0008142368751578033,
          "validation_loss": 0.0006324917776510119,
          "train_metric": 8.200693991966546e-05,
          "validation_metric": 6.370226037688553e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.861367531999349
        },
        {
          "epoch": 1269,
          "updates": 1269,
          "samples_seen": 1299456,
          "online_loss": 0.0008142368169501424,
          "train_loss": 0.0007547339191660285,
          "validation_loss": 0.0006310927565209568,
          "train_metric": 7.601403922308236e-05,
          "validation_metric": 6.356134690577164e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.869081388000268
        },
        {
          "epoch": 1270,
          "updates": 1270,
          "samples_seen": 1300480,
          "online_loss": 0.0007547338609583676,
          "train_loss": 0.0007736862753517926,
          "validation_loss": 0.000661365338601172,
          "train_metric": 7.79228430474177e-05,
          "validation_metric": 6.661029328824952e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.87711674899947
        },
        {
          "epoch": 1271,
          "updates": 1271,
          "samples_seen": 1301504,
          "online_loss": 0.0007736862753517926,
          "train_loss": 0.0008272714330814779,
          "validation_loss": 0.0006441038567572832,
          "train_metric": 8.331973367603496e-05,
          "validation_metric": 6.487178325187415e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.984743600000002
        },
        {
          "epoch": 1272,
          "updates": 1272,
          "samples_seen": 1302528,
          "online_loss": 0.000827271316666156,
          "train_loss": 0.0008334617596119642,
          "validation_loss": 0.0007270650239661336,
          "train_metric": 8.394321775995195e-05,
          "validation_metric": 7.322733290493488e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 14.993311738000557
        },
        {
          "epoch": 1273,
          "updates": 1273,
          "samples_seen": 1303552,
          "online_loss": 0.000833461876027286,
          "train_loss": 0.0007936313049867749,
          "validation_loss": 0.0006234069587662816,
          "train_metric": 7.993163308128715e-05,
          "validation_metric": 6.278725777519867e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.001174831000753
        },
        {
          "epoch": 1274,
          "updates": 1274,
          "samples_seen": 1304576,
          "online_loss": 0.000793631246779114,
          "train_loss": 0.0007509120041504502,
          "validation_loss": 0.0006279933732002974,
          "train_metric": 7.562910468550399e-05,
          "validation_metric": 6.324919377220795e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.009220871999787
        },
        {
          "epoch": 1275,
          "updates": 1275,
          "samples_seen": 1305600,
          "online_loss": 0.0007509120041504502,
          "train_loss": 0.0007523833774030209,
          "validation_loss": 0.0006320291431620717,
          "train_metric": 7.577730139018968e-05,
          "validation_metric": 6.36556651443243e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.017271590999371
        },
        {
          "epoch": 1276,
          "updates": 1276,
          "samples_seen": 1306624,
          "online_loss": 0.0007523834356106818,
          "train_loss": 0.0007823208579793572,
          "validation_loss": 0.0006161359488032758,
          "train_metric": 7.879249460529536e-05,
          "validation_metric": 6.205495446920395e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.025872273999994
        },
        {
          "epoch": 1277,
          "updates": 1277,
          "samples_seen": 1307648,
          "online_loss": 0.0007823208579793572,
          "train_loss": 0.0007925191312097013,
          "validation_loss": 0.0006861507426947355,
          "train_metric": 7.98196269897744e-05,
          "validation_metric": 6.91065943101421e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.03391377600019
        },
        {
          "epoch": 1278,
          "updates": 1278,
          "samples_seen": 1308672,
          "online_loss": 0.0007925191312097013,
          "train_loss": 0.0007726606563664973,
          "validation_loss": 0.0006137610762380064,
          "train_metric": 7.781955355312675e-05,
          "validation_metric": 6.18157719145529e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.042582292000588
        },
        {
          "epoch": 1279,
          "updates": 1279,
          "samples_seen": 1309696,
          "online_loss": 0.0007726607145741582,
          "train_loss": 0.0007449107943102717,
          "validation_loss": 0.0006266582640819252,
          "train_metric": 7.502468361053616e-05,
          "validation_metric": 6.311471952358261e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.050419913000951
        },
        {
          "epoch": 1280,
          "updates": 1280,
          "samples_seen": 1310720,
          "online_loss": 0.0007449109689332545,
          "train_loss": 0.0007406151853501797,
          "validation_loss": 0.0006191270658746362,
          "train_metric": 7.459205517079681e-05,
          "validation_metric": 6.235621549421921e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.05789091800034
        },
        {
          "epoch": 1281,
          "updates": 1281,
          "samples_seen": 1311744,
          "online_loss": 0.0007406152435578406,
          "train_loss": 0.0007564082043245435,
          "validation_loss": 0.0006050490192137659,
          "train_metric": 7.618266681674868e-05,
          "validation_metric": 6.093832416809164e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.066224481999598
        },
        {
          "epoch": 1282,
          "updates": 1282,
          "samples_seen": 1312768,
          "online_loss": 0.0007564082043245435,
          "train_loss": 0.0007663595606572926,
          "validation_loss": 0.0006585094961337745,
          "train_metric": 7.718492997810245e-05,
          "validation_metric": 6.632267468376085e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.074513313000352
        },
        {
          "epoch": 1283,
          "updates": 1283,
          "samples_seen": 1313792,
          "online_loss": 0.0007663595606572926,
          "train_loss": 0.0007586376741528511,
          "validation_loss": 0.0006056372076272964,
          "train_metric": 7.640721742063761e-05,
          "validation_metric": 6.099756865296513e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.082258570000704
        },
        {
          "epoch": 1284,
          "updates": 1284,
          "samples_seen": 1314816,
          "online_loss": 0.0007586376741528511,
          "train_loss": 0.0007406629738397896,
          "validation_loss": 0.0006267494754865766,
          "train_metric": 7.459685730282217e-05,
          "validation_metric": 6.312390178209171e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.090225272999305
        },
        {
          "epoch": 1285,
          "updates": 1285,
          "samples_seen": 1315840,
          "online_loss": 0.0007406629156321287,
          "train_loss": 0.0007321125594899058,
          "validation_loss": 0.0006083862390369177,
          "train_metric": 7.373569678748026e-05,
          "validation_metric": 6.127443339210004e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.097678042999178
        },
        {
          "epoch": 1286,
          "updates": 1286,
          "samples_seen": 1316864,
          "online_loss": 0.0007321125594899058,
          "train_loss": 0.0007376324501819909,
          "validation_loss": 0.0006000532885082066,
          "train_metric": 7.429163815686479e-05,
          "validation_metric": 6.0435166233219206e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.106674074000694
        },
        {
          "epoch": 1287,
          "updates": 1287,
          "samples_seen": 1317888,
          "online_loss": 0.0007376325665973127,
          "train_loss": 0.0007458883919753134,
          "validation_loss": 0.0006366863381117582,
          "train_metric": 7.512314186897129e-05,
          "validation_metric": 6.412470975192264e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.114565735000724
        },
        {
          "epoch": 1288,
          "updates": 1288,
          "samples_seen": 1318912,
          "online_loss": 0.0007458884501829743,
          "train_loss": 0.0007456235471181571,
          "validation_loss": 0.0005991659127175808,
          "train_metric": 7.50964754843153e-05,
          "validation_metric": 6.0345799283823e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.123139990999334
        },
        {
          "epoch": 1289,
          "updates": 1289,
          "samples_seen": 1319936,
          "online_loss": 0.0007456234889104962,
          "train_loss": 0.0007357464637607336,
          "validation_loss": 0.0006242372328415513,
          "train_metric": 7.41016847314313e-05,
          "validation_metric": 6.28708949079737e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.131273533999774
        },
        {
          "epoch": 1290,
          "updates": 1290,
          "samples_seen": 1320960,
          "online_loss": 0.0007357464055530727,
          "train_loss": 0.0007268495974130929,
          "validation_loss": 0.0005994262173771858,
          "train_metric": 7.320563599932939e-05,
          "validation_metric": 6.03720145591069e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.139140339999358
        },
        {
          "epoch": 1291,
          "updates": 1291,
          "samples_seen": 1321984,
          "online_loss": 0.0007268495974130929,
          "train_loss": 0.0007257044198922813,
          "validation_loss": 0.0005979132256470621,
          "train_metric": 7.309029024327174e-05,
          "validation_metric": 6.0219626902835444e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.147611947999394
        },
        {
          "epoch": 1292,
          "updates": 1292,
          "samples_seen": 1323008,
          "online_loss": 0.0007257044198922813,
          "train_loss": 0.0007300043944269419,
          "validation_loss": 0.0006172357825562358,
          "train_metric": 7.352335524046794e-05,
          "validation_metric": 6.216571637196466e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.155336801000885
        },
        {
          "epoch": 1293,
          "updates": 1293,
          "samples_seen": 1324032,
          "online_loss": 0.0007300042198039591,
          "train_loss": 0.0007326038903556764,
          "validation_loss": 0.000592343567404896,
          "train_metric": 7.378518057521433e-05,
          "validation_metric": 5.965867603663355e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.164140378999946
        },
        {
          "epoch": 1294,
          "updates": 1294,
          "samples_seen": 1325056,
          "online_loss": 0.0007326038321480155,
          "train_loss": 0.0007291247602552176,
          "validation_loss": 0.0006166989915072918,
          "train_metric": 7.343478500843048e-05,
          "validation_metric": 6.21116632828489e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.172022819999256
        },
        {
          "epoch": 1295,
          "updates": 1295,
          "samples_seen": 1326080,
          "online_loss": 0.0007291248766705394,
          "train_loss": 0.000722759694326669,
          "validation_loss": 0.0005911830230616033,
          "train_metric": 7.279372221091762e-05,
          "validation_metric": 5.954179505351931e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.181127935000404
        },
        {
          "epoch": 1296,
          "updates": 1296,
          "samples_seen": 1327104,
          "online_loss": 0.0007227597525343299,
          "train_loss": 0.0007184450514614582,
          "validation_loss": 0.0005972907529212534,
          "train_metric": 7.23591583664529e-05,
          "validation_metric": 6.015693361405283e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.189588941999318
        },
        {
          "epoch": 1297,
          "updates": 1297,
          "samples_seen": 1328128,
          "online_loss": 0.0007184451096691191,
          "train_loss": 0.0007184118730947375,
          "validation_loss": 0.0006011381628923118,
          "train_metric": 7.235583325382322e-05,
          "validation_metric": 6.05444292887114e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.19731692499954
        },
        {
          "epoch": 1298,
          "updates": 1298,
          "samples_seen": 1329152,
          "online_loss": 0.0007184118730947375,
          "train_loss": 0.0007203571731224656,
          "validation_loss": 0.0005884587299078703,
          "train_metric": 7.255173841258511e-05,
          "validation_metric": 5.926741141593084e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.2055273950009
        },
        {
          "epoch": 1299,
          "updates": 1299,
          "samples_seen": 1330176,
          "online_loss": 0.0007203572313301265,
          "train_loss": 0.0007205390138551593,
          "validation_loss": 0.0006080907769501209,
          "train_metric": 7.257005199790001e-05,
          "validation_metric": 6.124467472545803e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.21346244400047
        },
        {
          "epoch": 1300,
          "updates": 1300,
          "samples_seen": 1331200,
          "online_loss": 0.0007205389556474984,
          "train_loss": 0.0007178288651630282,
          "validation_loss": 0.0005860997480340302,
          "train_metric": 7.229709444800392e-05,
          "validation_metric": 5.902981138206087e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.221815036000407
        },
        {
          "epoch": 1301,
          "updates": 1301,
          "samples_seen": 1332224,
          "online_loss": 0.0007178288651630282,
          "train_loss": 0.000713748624548316,
          "validation_loss": 0.000597773352637887,
          "train_metric": 7.188615563791245e-05,
          "validation_metric": 6.020553701091558e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.229702681999697
        },
        {
          "epoch": 1302,
          "updates": 1302,
          "samples_seen": 1333248,
          "online_loss": 0.000713748624548316,
          "train_loss": 0.0007110040169209242,
          "validation_loss": 0.0005899282405152917,
          "train_metric": 7.16097274562344e-05,
          "validation_metric": 5.941541530773975e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.23738323500038
        },
        {
          "epoch": 1303,
          "updates": 1303,
          "samples_seen": 1334272,
          "online_loss": 0.0007110040751285851,
          "train_loss": 0.0007104889373295009,
          "validation_loss": 0.0005868749576620758,
          "train_metric": 7.155784987844527e-05,
          "validation_metric": 5.910789332119748e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.245216151000932
        },
        {
          "epoch": 1304,
          "updates": 1304,
          "samples_seen": 1335296,
          "online_loss": 0.00071048887912184,
          "train_loss": 0.0007109481375664473,
          "validation_loss": 0.0005977354594506323,
          "train_metric": 7.160409586504102e-05,
          "validation_metric": 6.0201724409125745e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.25309269399986
        },
        {
          "epoch": 1305,
          "updates": 1305,
          "samples_seen": 1336320,
          "online_loss": 0.0007109480793587863,
          "train_loss": 0.0007106177508831024,
          "validation_loss": 0.0005825314437970519,
          "train_metric": 7.157080835895613e-05,
          "validation_metric": 5.8670433645602316e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.261820343999716
        },
        {
          "epoch": 1306,
          "updates": 1306,
          "samples_seen": 1337344,
          "online_loss": 0.0007106176344677806,
          "train_loss": 0.0007087023113854229,
          "validation_loss": 0.0005955810775049031,
          "train_metric": 7.137790089473128e-05,
          "validation_metric": 5.998475171509199e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.269684876999236
        },
        {
          "epoch": 1307,
          "updates": 1307,
          "samples_seen": 1338368,
          "online_loss": 0.0007087023113854229,
          "train_loss": 0.0007060234202072024,
          "validation_loss": 0.0005815084441564977,
          "train_metric": 7.110808655852452e-05,
          "validation_metric": 5.856740244780667e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.27850576399942
        },
        {
          "epoch": 1308,
          "updates": 1308,
          "samples_seen": 1339392,
          "online_loss": 0.0007060233619995415,
          "train_loss": 0.0007037770701572299,
          "validation_loss": 0.0005858392105437815,
          "train_metric": 7.088184065651149e-05,
          "validation_metric": 5.900357064092532e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.286541725999996
        },
        {
          "epoch": 1309,
          "updates": 1309,
          "samples_seen": 1340416,
          "online_loss": 0.0007037770701572299,
          "train_loss": 0.0007026335224509239,
          "validation_loss": 0.0005854549235664308,
          "train_metric": 7.076667679939419e-05,
          "validation_metric": 5.896487709833309e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.294252355000935
        },
        {
          "epoch": 1310,
          "updates": 1310,
          "samples_seen": 1341440,
          "online_loss": 0.0007026335806585848,
          "train_loss": 0.0007022393401712179,
          "validation_loss": 0.0005789920105598867,
          "train_metric": 7.07269791746512e-05,
          "validation_metric": 5.831395537825301e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.492236378000598
        },
        {
          "epoch": 1311,
          "updates": 1311,
          "samples_seen": 1342464,
          "online_loss": 0.0007022393401712179,
          "train_loss": 0.0007017192547209561,
          "validation_loss": 0.0005883458070456982,
          "train_metric": 7.067459227982908e-05,
          "validation_metric": 5.925603545620106e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.508145802999934
        },
        {
          "epoch": 1312,
          "updates": 1312,
          "samples_seen": 1343488,
          "online_loss": 0.0007017191965132952,
          "train_loss": 0.0007005163934081793,
          "validation_loss": 0.0005759095656685531,
          "train_metric": 7.055344030959532e-05,
          "validation_metric": 5.800349754281342e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.527169211
        },
        {
          "epoch": 1313,
          "updates": 1313,
          "samples_seen": 1344512,
          "online_loss": 0.0007005163934081793,
          "train_loss": 0.0006986716762185097,
          "validation_loss": 0.0005843075923621655,
          "train_metric": 7.036765600787476e-05,
          "validation_metric": 5.8849323977483436e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.550930322000568
        },
        {
          "epoch": 1314,
          "updates": 1314,
          "samples_seen": 1345536,
          "online_loss": 0.0006986716180108488,
          "train_loss": 0.0006967540248297155,
          "validation_loss": 0.0005758649203926325,
          "train_metric": 7.017451571300626e-05,
          "validation_metric": 5.799900463898666e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.569907491000777
        },
        {
          "epoch": 1315,
          "updates": 1315,
          "samples_seen": 1346560,
          "online_loss": 0.0006967540830373764,
          "train_loss": 0.0006952430121600628,
          "validation_loss": 0.0005773233133368194,
          "train_metric": 7.0022331783548e-05,
          "validation_metric": 5.8145895309280604e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.587715002999175
        },
        {
          "epoch": 1316,
          "updates": 1316,
          "samples_seen": 1347584,
          "online_loss": 0.0006952429539524019,
          "train_loss": 0.000694245332852006,
          "validation_loss": 0.0005784039967693388,
          "train_metric": 6.992185080889612e-05,
          "validation_metric": 5.8254736359231174e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.603263496999716
        },
        {
          "epoch": 1317,
          "updates": 1317,
          "samples_seen": 1348608,
          "online_loss": 0.000694245332852006,
          "train_loss": 0.0006934796110726893,
          "validation_loss": 0.0005723458016291261,
          "train_metric": 6.984472565818578e-05,
          "validation_metric": 5.764457455370575e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.621475776999432
        },
        {
          "epoch": 1318,
          "updates": 1318,
          "samples_seen": 1349632,
          "online_loss": 0.0006934796110726893,
          "train_loss": 0.0006925605121068656,
          "validation_loss": 0.0005791217554360628,
          "train_metric": 6.975216092541814e-05,
          "validation_metric": 5.8327022998128086e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.632673381000131
        },
        {
          "epoch": 1319,
          "updates": 1319,
          "samples_seen": 1350656,
          "online_loss": 0.0006925605121068656,
          "train_loss": 0.0006913110264576972,
          "validation_loss": 0.0005698774475604296,
          "train_metric": 6.962631596252322e-05,
          "validation_metric": 5.739596963394433e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.641306343999531
        },
        {
          "epoch": 1320,
          "updates": 1320,
          "samples_seen": 1351680,
          "online_loss": 0.0006913110846653581,
          "train_loss": 0.0006897979765199125,
          "validation_loss": 0.0005758583429269493,
          "train_metric": 6.947392830625176e-05,
          "validation_metric": 5.799834616482258e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.64876074499989
        },
        {
          "epoch": 1321,
          "updates": 1321,
          "samples_seen": 1352704,
          "online_loss": 0.0006897979183122516,
          "train_loss": 0.000688263913616538,
          "validation_loss": 0.0005698049208149314,
          "train_metric": 6.93194306222722e-05,
          "validation_metric": 5.738866457249969e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.657417596999949
        },
        {
          "epoch": 1322,
          "updates": 1322,
          "samples_seen": 1353728,
          "online_loss": 0.0006882640300318599,
          "train_loss": 0.0006869130302220583,
          "validation_loss": 0.0005709443939849734,
          "train_metric": 6.918336293892935e-05,
          "validation_metric": 5.750342825194821e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.665052447000562
        },
        {
          "epoch": 1323,
          "updates": 1323,
          "samples_seen": 1354752,
          "online_loss": 0.0006869130302220583,
          "train_loss": 0.0006857915432192385,
          "validation_loss": 0.0005709286197088659,
          "train_metric": 6.907041824888438e-05,
          "validation_metric": 5.750184573116712e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.672629148000851
        },
        {
          "epoch": 1324,
          "updates": 1324,
          "samples_seen": 1355776,
          "online_loss": 0.0006857915432192385,
          "train_loss": 0.000684799742884934,
          "validation_loss": 0.000566799018997699,
          "train_metric": 6.897051935084164e-05,
          "validation_metric": 5.708592289010994e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.680851952000012
        },
        {
          "epoch": 1325,
          "updates": 1325,
          "samples_seen": 1356800,
          "online_loss": 0.000684799742884934,
          "train_loss": 0.0006837854161858559,
          "validation_loss": 0.0005709856632165611,
          "train_metric": 6.886835762998089e-05,
          "validation_metric": 5.750759009970352e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.688416037000934
        },
        {
          "epoch": 1326,
          "updates": 1326,
          "samples_seen": 1357824,
          "online_loss": 0.0006837854161858559,
          "train_loss": 0.0006826471071690321,
          "validation_loss": 0.0005642946343868971,
          "train_metric": 6.875371764181182e-05,
          "validation_metric": 5.6833687267499045e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.696262928999204
        },
        {
          "epoch": 1327,
          "updates": 1327,
          "samples_seen": 1358848,
          "online_loss": 0.000682647165376693,
          "train_loss": 0.0006813715444877744,
          "validation_loss": 0.0005689030513167381,
          "train_metric": 6.86252533341758e-05,
          "validation_metric": 5.7297835155623034e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.703351219000979
        },
        {
          "epoch": 1328,
          "updates": 1328,
          "samples_seen": 1359872,
          "online_loss": 0.0006813715444877744,
          "train_loss": 0.0006800319533795118,
          "validation_loss": 0.000563413486815989,
          "train_metric": 6.849033525213599e-05,
          "validation_metric": 5.674493877450004e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.71149663200049
        },
        {
          "epoch": 1329,
          "updates": 1329,
          "samples_seen": 1360896,
          "online_loss": 0.0006800319533795118,
          "train_loss": 0.0006787218735553324,
          "validation_loss": 0.0005652693216688931,
          "train_metric": 6.835838576080278e-05,
          "validation_metric": 5.693185448762961e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.718740914000591
        },
        {
          "epoch": 1330,
          "updates": 1330,
          "samples_seen": 1361920,
          "online_loss": 0.0006787218735553324,
          "train_loss": 0.0006775016663596034,
          "validation_loss": 0.0005631858948618174,
          "train_metric": 6.823549483669922e-05,
          "validation_metric": 5.672202314599417e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.726982107000367
        },
        {
          "epoch": 1331,
          "updates": 1331,
          "samples_seen": 1362944,
          "online_loss": 0.0006775016663596034,
          "train_loss": 0.0006763732526451349,
          "validation_loss": 0.0005613825633190572,
          "train_metric": 6.812183710280806e-05,
          "validation_metric": 5.6540400692028925e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.735263307000423
        },
        {
          "epoch": 1332,
          "updates": 1332,
          "samples_seen": 1363968,
          "online_loss": 0.0006763731362298131,
          "train_loss": 0.000675287505146116,
          "validation_loss": 0.0005627681384794414,
          "train_metric": 6.80124867358245e-05,
          "validation_metric": 5.6679949921090156e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.742548341999282
        },
        {
          "epoch": 1333,
          "updates": 1333,
          "samples_seen": 1364992,
          "online_loss": 0.0006752875633537769,
          "train_loss": 0.0006741992547176778,
          "validation_loss": 0.0005584722384810448,
          "train_metric": 6.790288171032444e-05,
          "validation_metric": 5.624728146358393e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.750568486000702
        },
        {
          "epoch": 1334,
          "updates": 1334,
          "samples_seen": 1366016,
          "online_loss": 0.0006741992547176778,
          "train_loss": 0.0006730679306201637,
          "validation_loss": 0.0005615678965114057,
          "train_metric": 6.778894021408632e-05,
          "validation_metric": 5.6559063523309305e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.75803042500047
        },
        {
          "epoch": 1335,
          "updates": 1335,
          "samples_seen": 1367040,
          "online_loss": 0.0006730680470354855,
          "train_loss": 0.0006718821823596954,
          "validation_loss": 0.0005566935869865119,
          "train_metric": 6.766951992176473e-05,
          "validation_metric": 5.606814374914393e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.766124784999192
        },
        {
          "epoch": 1336,
          "updates": 1336,
          "samples_seen": 1368064,
          "online_loss": 0.0006718821823596954,
          "train_loss": 0.0006706618005409837,
          "validation_loss": 0.0005592243978753686,
          "train_metric": 6.754659989383072e-05,
          "validation_metric": 5.6323031458305195e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.773616297999979
        },
        {
          "epoch": 1337,
          "updates": 1337,
          "samples_seen": 1369088,
          "online_loss": 0.00067066156771034,
          "train_loss": 0.0006694387411698699,
          "validation_loss": 0.0005554876988753676,
          "train_metric": 6.742343248333782e-05,
          "validation_metric": 5.594668255071156e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.782102741999552
        },
        {
          "epoch": 1338,
          "updates": 1338,
          "samples_seen": 1370112,
          "online_loss": 0.0006694387411698699,
          "train_loss": 0.0006682343082502484,
          "validation_loss": 0.0005561478901654482,
          "train_metric": 6.730211316607893e-05,
          "validation_metric": 5.601317752734758e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.789764012000887
        },
        {
          "epoch": 1339,
          "updates": 1339,
          "samples_seen": 1371136,
          "online_loss": 0.0006682343664579093,
          "train_loss": 0.0006670705624856055,
          "validation_loss": 0.0005546180764213204,
          "train_metric": 6.718489748891443e-05,
          "validation_metric": 5.585910184890963e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.797814761999689
        },
        {
          "epoch": 1340,
          "updates": 1340,
          "samples_seen": 1372160,
          "online_loss": 0.0006670705042779446,
          "train_loss": 0.0006659302162006497,
          "validation_loss": 0.0005532607319764793,
          "train_metric": 6.707005377393216e-05,
          "validation_metric": 5.572240115725435e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.806330421999519
        },
        {
          "epoch": 1341,
          "updates": 1341,
          "samples_seen": 1373184,
          "online_loss": 0.0006659302744083107,
          "train_loss": 0.0006648106500506401,
          "validation_loss": 0.0005537621327675879,
          "train_metric": 6.695728370686993e-05,
          "validation_metric": 5.577289994107559e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.81379467700026
        },
        {
          "epoch": 1342,
          "updates": 1342,
          "samples_seen": 1374208,
          "online_loss": 0.0006648105918429792,
          "train_loss": 0.0006636949838139117,
          "validation_loss": 0.0005508197355084121,
          "train_metric": 6.684492836939171e-05,
          "validation_metric": 5.547653927351348e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.822277351999219
        },
        {
          "epoch": 1343,
          "updates": 1343,
          "samples_seen": 1375232,
          "online_loss": 0.0006636949256062508,
          "train_loss": 0.0006625712849199772,
          "validation_loss": 0.0005523572326637805,
          "train_metric": 6.67317581246607e-05,
          "validation_metric": 5.5631397117394954e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.829757564000829
        },
        {
          "epoch": 1344,
          "updates": 1344,
          "samples_seen": 1376256,
          "online_loss": 0.0006625712849199772,
          "train_loss": 0.0006614304147660732,
          "validation_loss": 0.0005487308371812105,
          "train_metric": 6.661686347797513e-05,
          "validation_metric": 5.526616587303579e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.838061075000951
        },
        {
          "epoch": 1345,
          "updates": 1345,
          "samples_seen": 1377280,
          "online_loss": 0.0006614304729737341,
          "train_loss": 0.0006602805224247277,
          "validation_loss": 0.0005503618158400059,
          "train_metric": 6.650103023275733e-05,
          "validation_metric": 5.5430427892133594e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.845296903000417
        },
        {
          "epoch": 1346,
          "updates": 1346,
          "samples_seen": 1378304,
          "online_loss": 0.0006602805224247277,
          "train_loss": 0.0006591211422346532,
          "validation_loss": 0.0005470563773997128,
          "train_metric": 6.638426566496491e-05,
          "validation_metric": 5.509750917553902e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.853464759000417
        },
        {
          "epoch": 1347,
          "updates": 1347,
          "samples_seen": 1379328,
          "online_loss": 0.0006591211422346532,
          "train_loss": 0.0006579618202522397,
          "validation_loss": 0.0005480944528244436,
          "train_metric": 6.626750837313011e-05,
          "validation_metric": 5.520206468645483e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.860626208999747
        },
        {
          "epoch": 1348,
          "updates": 1348,
          "samples_seen": 1380352,
          "online_loss": 0.0006579619948752224,
          "train_loss": 0.0006568038370460272,
          "validation_loss": 0.0005456776125356555,
          "train_metric": 6.615088204853237e-05,
          "validation_metric": 5.495864388649352e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.96928304100038
        },
        {
          "epoch": 1349,
          "updates": 1349,
          "samples_seen": 1381376,
          "online_loss": 0.0006568038952536881,
          "train_loss": 0.0006556568550877273,
          "validation_loss": 0.0005456667277030647,
          "train_metric": 6.603535439353436e-05,
          "validation_metric": 5.4957559768809006e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.977567111000099
        },
        {
          "epoch": 1350,
          "updates": 1350,
          "samples_seen": 1382400,
          "online_loss": 0.0006556569132953882,
          "train_loss": 0.0006545217474922538,
          "validation_loss": 0.0005443020490929484,
          "train_metric": 6.592103454750031e-05,
          "validation_metric": 5.4820109653519467e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.98645529299938
        },
        {
          "epoch": 1351,
          "updates": 1351,
          "samples_seen": 1383424,
          "online_loss": 0.000654521631076932,
          "train_loss": 0.0006533906562253833,
          "validation_loss": 0.0005432633915916085,
          "train_metric": 6.580711487913504e-05,
          "validation_metric": 5.471549593494274e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 15.994914512000832
        },
        {
          "epoch": 1352,
          "updates": 1352,
          "samples_seen": 1384448,
          "online_loss": 0.0006533906562253833,
          "train_loss": 0.0006522724870592356,
          "validation_loss": 0.0005428891745395958,
          "train_metric": 6.56944903312251e-05,
          "validation_metric": 5.467781375045888e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.003687369000545
        },
        {
          "epoch": 1353,
          "updates": 1353,
          "samples_seen": 1385472,
          "online_loss": 0.0006522724288515747,
          "train_loss": 0.0006511565879918635,
          "validation_loss": 0.0005410677404142916,
          "train_metric": 6.558211316587403e-05,
          "validation_metric": 5.449436139315367e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.012469871000576
        },
        {
          "epoch": 1354,
          "updates": 1354,
          "samples_seen": 1386496,
          "online_loss": 0.0006511565879918635,
          "train_loss": 0.0006500385352410376,
          "validation_loss": 0.0005413787439465523,
          "train_metric": 6.54694958939217e-05,
          "validation_metric": 5.452568075270392e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.019765495000684
        },
        {
          "epoch": 1355,
          "updates": 1355,
          "samples_seen": 1387520,
          "online_loss": 0.0006500384770333767,
          "train_loss": 0.0006489231600426137,
          "validation_loss": 0.0005390201695263386,
          "train_metric": 6.535716238431633e-05,
          "validation_metric": 5.428813528851606e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.02810733499973
        },
        {
          "epoch": 1356,
          "updates": 1356,
          "samples_seen": 1388544,
          "online_loss": 0.000648922985419631,
          "train_loss": 0.0006478080758824944,
          "validation_loss": 0.0005396736669354141,
          "train_metric": 6.52448579785414e-05,
          "validation_metric": 5.4353949963115156e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.035790367999653
        },
        {
          "epoch": 1357,
          "updates": 1357,
          "samples_seen": 1389568,
          "online_loss": 0.0006478080758824944,
          "train_loss": 0.0006466938648372889,
          "validation_loss": 0.0005370780709199607,
          "train_metric": 6.513263360830024e-05,
          "validation_metric": 5.4092535719973966e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.044236509000257
        },
        {
          "epoch": 1358,
          "updates": 1358,
          "samples_seen": 1390592,
          "online_loss": 0.0006466938648372889,
          "train_loss": 0.000645574415102601,
          "validation_loss": 0.0005378611967898905,
          "train_metric": 6.501988536911085e-05,
          "validation_metric": 5.4171407100511715e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.05148765100057
        },
        {
          "epoch": 1359,
          "updates": 1359,
          "samples_seen": 1391616,
          "online_loss": 0.0006455742404796183,
          "train_loss": 0.0006444570026360452,
          "validation_loss": 0.0005352828302420676,
          "train_metric": 6.490734813269228e-05,
          "validation_metric": 5.39117245352827e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.05955476700001
        },
        {
          "epoch": 1360,
          "updates": 1360,
          "samples_seen": 1392640,
          "online_loss": 0.0006444569444283843,
          "train_loss": 0.0006433439557440579,
          "validation_loss": 0.0005360239301808178,
          "train_metric": 6.479524745373055e-05,
          "validation_metric": 5.39863649464678e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.06700712699967
        },
        {
          "epoch": 1361,
          "updates": 1361,
          "samples_seen": 1393664,
          "online_loss": 0.0006433439557440579,
          "train_loss": 0.0006422300939448178,
          "validation_loss": 0.0005335238529369235,
          "train_metric": 6.468305946327746e-05,
          "validation_metric": 5.373456951929256e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.074959293000575
        },
        {
          "epoch": 1362,
          "updates": 1362,
          "samples_seen": 1394688,
          "online_loss": 0.0006422300357371569,
          "train_loss": 0.0006411179201677442,
          "validation_loss": 0.0005341160576790571,
          "train_metric": 6.457103881984949e-05,
          "validation_metric": 5.379421781981364e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.08206825600064
        },
        {
          "epoch": 1363,
          "updates": 1363,
          "samples_seen": 1395712,
          "online_loss": 0.0006411178037524223,
          "train_loss": 0.0006400045822374523,
          "validation_loss": 0.0005317304749041796,
          "train_metric": 6.445892358897254e-05,
          "validation_metric": 5.355394387152046e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.089846130000296
        },
        {
          "epoch": 1364,
          "updates": 1364,
          "samples_seen": 1396736,
          "online_loss": 0.0006400045822374523,
          "train_loss": 0.00063889886951074,
          "validation_loss": 0.0005322013748809695,
          "train_metric": 6.434755778172985e-05,
          "validation_metric": 5.360136856324971e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.097242764999464
        },
        {
          "epoch": 1365,
          "updates": 1365,
          "samples_seen": 1397760,
          "online_loss": 0.000638898927718401,
          "train_loss": 0.0006377897225320339,
          "validation_loss": 0.0005299542099237442,
          "train_metric": 6.423584272852167e-05,
          "validation_metric": 5.337503898772411e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.10517851899931
        },
        {
          "epoch": 1366,
          "updates": 1366,
          "samples_seen": 1398784,
          "online_loss": 0.000637789664324373,
          "train_loss": 0.0006366899469867349,
          "validation_loss": 0.0005303466459736228,
          "train_metric": 6.412508082576096e-05,
          "validation_metric": 5.341458017937839e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.112938812999346
        },
        {
          "epoch": 1367,
          "updates": 1367,
          "samples_seen": 1399808,
          "online_loss": 0.0006366899469867349,
          "train_loss": 0.0006355880759656429,
          "validation_loss": 0.0005281495396047831,
          "train_metric": 6.401410792022943e-05,
          "validation_metric": 5.3193289204500616e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.12062305099971
        },
        {
          "epoch": 1368,
          "updates": 1368,
          "samples_seen": 1400832,
          "online_loss": 0.0006355880759656429,
          "train_loss": 0.0006344886496663094,
          "validation_loss": 0.0005285111838020384,
          "train_metric": 6.390338239725679e-05,
          "validation_metric": 5.3229716286296025e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.12834256899987
        },
        {
          "epoch": 1369,
          "updates": 1369,
          "samples_seen": 1401856,
          "online_loss": 0.0006344887660816312,
          "train_loss": 0.0006333943456411362,
          "validation_loss": 0.0005262690829113126,
          "train_metric": 6.379315891535953e-05,
          "validation_metric": 5.300389602780342e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.13640255199971
        },
        {
          "epoch": 1370,
          "updates": 1370,
          "samples_seen": 1402880,
          "online_loss": 0.0006333943456411362,
          "train_loss": 0.0006323025445453823,
          "validation_loss": 0.0005267171072773635,
          "train_metric": 6.368319736793637e-05,
          "validation_metric": 5.304902151692659e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.144198306999897
        },
        {
          "epoch": 1371,
          "updates": 1371,
          "samples_seen": 1403904,
          "online_loss": 0.0006323025445453823,
          "train_loss": 0.0006312108016572893,
          "validation_loss": 0.0005243587074801326,
          "train_metric": 6.357324309647083e-05,
          "validation_metric": 5.281148696667515e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.15226539700052
        },
        {
          "epoch": 1372,
          "updates": 1372,
          "samples_seen": 1404928,
          "online_loss": 0.0006312107434496284,
          "train_loss": 0.000630125286988914,
          "validation_loss": 0.0005250404356047511,
          "train_metric": 6.34639072814025e-05,
          "validation_metric": 5.2880150178680196e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.160300375999213
        },
        {
          "epoch": 1373,
          "updates": 1373,
          "samples_seen": 1405952,
          "online_loss": 0.0006301251705735922,
          "train_loss": 0.0006290449528023601,
          "validation_loss": 0.0005224034539423883,
          "train_metric": 6.335510988719761e-05,
          "validation_metric": 5.261455225991085e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.168173337000553
        },
        {
          "epoch": 1374,
          "updates": 1374,
          "samples_seen": 1406976,
          "online_loss": 0.0006290451274253428,
          "train_loss": 0.0006279691006056964,
          "validation_loss": 0.0005234775599092245,
          "train_metric": 6.324674177449197e-05,
          "validation_metric": 5.272273483569734e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.175575982999362
        },
        {
          "epoch": 1375,
          "updates": 1375,
          "samples_seen": 1408000,
          "online_loss": 0.0006279691006056964,
          "train_loss": 0.0006269036093726754,
          "validation_loss": 0.0005203296313993633,
          "train_metric": 6.313942867564037e-05,
          "validation_metric": 5.2405695896595716e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.183382184999573
        },
        {
          "epoch": 1376,
          "updates": 1376,
          "samples_seen": 1409024,
          "online_loss": 0.0006269035511650145,
          "train_loss": 0.0006258482462726533,
          "validation_loss": 0.0005220751627348363,
          "train_metric": 6.303314148681238e-05,
          "validation_metric": 5.2581490308512e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.190949674000876
        },
        {
          "epoch": 1377,
          "updates": 1377,
          "samples_seen": 1410048,
          "online_loss": 0.0006258482462726533,
          "train_loss": 0.0006248109275475144,
          "validation_loss": 0.0005181199521757662,
          "train_metric": 6.292868056334555e-05,
          "validation_metric": 5.218314254307188e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.19888204000017
        },
        {
          "epoch": 1378,
          "updates": 1378,
          "samples_seen": 1411072,
          "online_loss": 0.0006248109275475144,
          "train_loss": 0.0006237982306629419,
          "validation_loss": 0.0005209813243709505,
          "train_metric": 6.28266716375947e-05,
          "validation_metric": 5.2471332310233265e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.206386104000558
        },
        {
          "epoch": 1379,
          "updates": 1379,
          "samples_seen": 1412096,
          "online_loss": 0.0006237982306629419,
          "train_loss": 0.0006228265701793134,
          "validation_loss": 0.000515706604346633,
          "train_metric": 6.272881000768393e-05,
          "validation_metric": 5.194008190301247e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.214404952999757
        },
        {
          "epoch": 1380,
          "updates": 1380,
          "samples_seen": 1413120,
          "online_loss": 0.0006228265701793134,
          "train_loss": 0.0006219182396307588,
          "validation_loss": 0.0005204016342759132,
          "train_metric": 6.263732211664319e-05,
          "validation_metric": 5.241294638835825e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.22157432200038
        },
        {
          "epoch": 1381,
          "updates": 1381,
          "samples_seen": 1414144,
          "online_loss": 0.0006219182396307588,
          "train_loss": 0.0006211112486198545,
          "validation_loss": 0.0005129809724166989,
          "train_metric": 6.255604239413515e-05,
          "validation_metric": 5.166556729818694e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.229780935000235
        },
        {
          "epoch": 1382,
          "updates": 1382,
          "samples_seen": 1415168,
          "online_loss": 0.0006211111904121935,
          "train_loss": 0.0006204678793437779,
          "validation_loss": 0.0005207953508943319,
          "train_metric": 6.249125726753846e-05,
          "validation_metric": 5.245259671937674e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.236964297999293
        },
        {
          "epoch": 1383,
          "updates": 1383,
          "samples_seen": 1416192,
          "online_loss": 0.0006204679957590997,
          "train_loss": 0.0006201157229952514,
          "validation_loss": 0.0005099375266581774,
          "train_metric": 6.245578697416931e-05,
          "validation_metric": 5.135903847985901e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.24515202600014
        },
        {
          "epoch": 1384,
          "updates": 1384,
          "samples_seen": 1417216,
          "online_loss": 0.0006201157229952514,
          "train_loss": 0.0006202139193192124,
          "validation_loss": 0.0005232659168541431,
          "train_metric": 6.246567500056699e-05,
          "validation_metric": 5.2701434469781816e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.252308675000677
        },
        {
          "epoch": 1385,
          "updates": 1385,
          "samples_seen": 1418240,
          "online_loss": 0.0006202139193192124,
          "train_loss": 0.0006211751606315374,
          "validation_loss": 0.0005069283652119339,
          "train_metric": 6.256249616853893e-05,
          "validation_metric": 5.1055976655334234e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.260325699000532
        },
        {
          "epoch": 1386,
          "updates": 1386,
          "samples_seen": 1419264,
          "online_loss": 0.0006211751606315374,
          "train_loss": 0.0006234316388145089,
          "validation_loss": 0.0005304518272168934,
          "train_metric": 6.278974615270272e-05,
          "validation_metric": 5.342517397366464e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.267419064999558
        },
        {
          "epoch": 1387,
          "updates": 1387,
          "samples_seen": 1420288,
          "online_loss": 0.000623431580606848,
          "train_loss": 0.0006283718976192176,
          "validation_loss": 0.0005058387177996337,
          "train_metric": 6.328731251414865e-05,
          "validation_metric": 5.094621883472428e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.27556644499964
        },
        {
          "epoch": 1388,
          "updates": 1388,
          "samples_seen": 1421312,
          "online_loss": 0.0006283718976192176,
          "train_loss": 0.0006370958872139454,
          "validation_loss": 0.0005493416683748364,
          "train_metric": 6.416597170755267e-05,
          "validation_metric": 5.532768409466371e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.372764526999163
        },
        {
          "epoch": 1389,
          "updates": 1389,
          "samples_seen": 1422336,
          "online_loss": 0.0006370958872139454,
          "train_loss": 0.0006545347278006375,
          "validation_loss": 0.000514296698383987,
          "train_metric": 6.592234421987087e-05,
          "validation_metric": 5.1798073400277644e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.38053904499975
        },
        {
          "epoch": 1390,
          "updates": 1390,
          "samples_seen": 1423360,
          "online_loss": 0.0006545346695929766,
          "train_loss": 0.0006824912270531058,
          "validation_loss": 0.0005991761572659016,
          "train_metric": 6.873801612528041e-05,
          "validation_metric": 6.03468251938466e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.387689003999185
        },
        {
          "epoch": 1391,
          "updates": 1391,
          "samples_seen": 1424384,
          "online_loss": 0.0006824912270531058,
          "train_loss": 0.0007387897931039333,
          "validation_loss": 0.0005586075130850077,
          "train_metric": 7.440819899784401e-05,
          "validation_metric": 5.6260898418258876e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.394813953000266
        },
        {
          "epoch": 1392,
          "updates": 1392,
          "samples_seen": 1425408,
          "online_loss": 0.0007387897931039333,
          "train_loss": 0.0008174495887942612,
          "validation_loss": 0.0007287219632416964,
          "train_metric": 8.233053085859865e-05,
          "validation_metric": 7.339420699281618e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.40225387999999
        },
        {
          "epoch": 1393,
          "updates": 1393,
          "samples_seen": 1426432,
          "online_loss": 0.0008174495887942612,
          "train_loss": 0.000977571471594274,
          "validation_loss": 0.0007089031278155744,
          "train_metric": 9.84574289759621e-05,
          "validation_metric": 7.139812805689871e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.409349014000327
        },
        {
          "epoch": 1394,
          "updates": 1394,
          "samples_seen": 1427456,
          "online_loss": 0.0009775713551789522,
          "train_loss": 0.0011299183825030923,
          "validation_loss": 0.0010083980159834027,
          "train_metric": 0.00011380123032722622,
          "validation_metric": 0.00010156216740142554,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.416333978000694
        },
        {
          "epoch": 1395,
          "updates": 1395,
          "samples_seen": 1428480,
          "online_loss": 0.0011299183825030923,
          "train_loss": 0.0014176091644912958,
          "validation_loss": 0.0010087864939123392,
          "train_metric": 0.00014277642185334116,
          "validation_metric": 0.00010160128294955939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.42352230799952
        },
        {
          "epoch": 1396,
          "updates": 1396,
          "samples_seen": 1429504,
          "online_loss": 0.0014176091644912958,
          "train_loss": 0.0014147164765745401,
          "validation_loss": 0.0012613203143700957,
          "train_metric": 0.0001424850634066388,
          "validation_metric": 0.00012703557149507105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.430359037000017
        },
        {
          "epoch": 1397,
          "updates": 1397,
          "samples_seen": 1430528,
          "online_loss": 0.001414716592989862,
          "train_loss": 0.0013992824824526906,
          "validation_loss": 0.000993889756500721,
          "train_metric": 0.00014093058416619897,
          "validation_metric": 0.0001001009441097267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.436820487999285
        },
        {
          "epoch": 1398,
          "updates": 1398,
          "samples_seen": 1431552,
          "online_loss": 0.001399282249622047,
          "train_loss": 0.0009778091916814446,
          "validation_loss": 0.0008725850493647158,
          "train_metric": 9.848135960055515e-05,
          "validation_metric": 8.78835708135739e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.44367912599955
        },
        {
          "epoch": 1399,
          "updates": 1399,
          "samples_seen": 1432576,
          "online_loss": 0.0009778091916814446,
          "train_loss": 0.0006747526349499822,
          "validation_loss": 0.0005084351869300008,
          "train_metric": 6.795861554564908e-05,
          "validation_metric": 5.120773130329326e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.45066875699922
        },
        {
          "epoch": 1400,
          "updates": 1400,
          "samples_seen": 1433600,
          "online_loss": 0.000674752751365304,
          "train_loss": 0.0006244329852052033,
          "validation_loss": 0.000483419542433694,
          "train_metric": 6.289060547715053e-05,
          "validation_metric": 4.8688245442463085e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.45854543600035
        },
        {
          "epoch": 1401,
          "updates": 1401,
          "samples_seen": 1434624,
          "online_loss": 0.0006244328105822206,
          "train_loss": 0.0008044758578762412,
          "validation_loss": 0.0006944818887859583,
          "train_metric": 8.102386345854029e-05,
          "validation_metric": 6.994567229412496e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.465679229999296
        },
        {
          "epoch": 1402,
          "updates": 1402,
          "samples_seen": 1435648,
          "online_loss": 0.0008044759160839021,
          "train_loss": 0.0010284393792971969,
          "validation_loss": 0.0007379417074844241,
          "train_metric": 0.00010358065628679469,
          "validation_metric": 7.432278653141111e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.472253740999804
        },
        {
          "epoch": 1403,
          "updates": 1403,
          "samples_seen": 1436672,
          "online_loss": 0.0010284393792971969,
          "train_loss": 0.0009879549033939838,
          "validation_loss": 0.0008647118811495602,
          "train_metric": 9.950318781193346e-05,
          "validation_metric": 8.709061512490734e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.47969845099942
        },
        {
          "epoch": 1404,
          "updates": 1404,
          "samples_seen": 1437696,
          "online_loss": 0.0009879546705633402,
          "train_loss": 0.0008317617466673255,
          "validation_loss": 0.0006110119866207242,
          "train_metric": 8.377199264941737e-05,
          "validation_metric": 6.153888534754515e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.486759335999523
        },
        {
          "epoch": 1405,
          "updates": 1405,
          "samples_seen": 1438720,
          "online_loss": 0.0008317618048749864,
          "train_loss": 0.0006363611319102347,
          "validation_loss": 0.0005398562061600387,
          "train_metric": 6.40919606667012e-05,
          "validation_metric": 5.437234358396381e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.49449248900055
        },
        {
          "epoch": 1406,
          "updates": 1406,
          "samples_seen": 1439744,
          "online_loss": 0.0006363611319102347,
          "train_loss": 0.0006156367599032819,
          "validation_loss": 0.0005153222591616213,
          "train_metric": 6.200467032613233e-05,
          "validation_metric": 5.1901370170526206e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.501652413000556
        },
        {
          "epoch": 1407,
          "updates": 1407,
          "samples_seen": 1440768,
          "online_loss": 0.0006156367599032819,
          "train_loss": 0.0007384593482129276,
          "validation_loss": 0.0005439753294922411,
          "train_metric": 7.437492604367435e-05,
          "validation_metric": 5.478720777318813e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.50864204699974
        },
        {
          "epoch": 1408,
          "updates": 1408,
          "samples_seen": 1441792,
          "online_loss": 0.0007384592900052667,
          "train_loss": 0.0008210260421037674,
          "validation_loss": 0.000729615509044379,
          "train_metric": 8.269074169220403e-05,
          "validation_metric": 7.348420331254601e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.515409772000567
        },
        {
          "epoch": 1409,
          "updates": 1409,
          "samples_seen": 1442816,
          "online_loss": 0.0008210259838961065,
          "train_loss": 0.0008022970287129283,
          "validation_loss": 0.0005929536418989301,
          "train_metric": 8.080442057689652e-05,
          "validation_metric": 5.972013605060056e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.521573133999482
        },
        {
          "epoch": 1410,
          "updates": 1410,
          "samples_seen": 1443840,
          "online_loss": 0.0008022969705052674,
          "train_loss": 0.0006692906026728451,
          "validation_loss": 0.000594851968344301,
          "train_metric": 6.74084949423559e-05,
          "validation_metric": 5.9911315474892035e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.528279341000598
        },
        {
          "epoch": 1411,
          "updates": 1411,
          "samples_seen": 1444864,
          "online_loss": 0.0006692905444651842,
          "train_loss": 0.0005951871280558407,
          "validation_loss": 0.0004916703328490257,
          "train_metric": 5.9945075918221846e-05,
          "validation_metric": 4.951923983753659e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.534596920000695
        },
        {
          "epoch": 1412,
          "updates": 1412,
          "samples_seen": 1445888,
          "online_loss": 0.0005951871862635016,
          "train_loss": 0.0006313321646302938,
          "validation_loss": 0.0004903309745714068,
          "train_metric": 6.358546670526266e-05,
          "validation_metric": 4.938434358336963e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.53986188700037
        },
        {
          "epoch": 1413,
          "updates": 1413,
          "samples_seen": 1446912,
          "online_loss": 0.0006313321646302938,
          "train_loss": 0.000705067883245647,
          "validation_loss": 0.0006247326382435858,
          "train_metric": 7.101186201907694e-05,
          "validation_metric": 6.292078614933416e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.544710075000694
        },
        {
          "epoch": 1414,
          "updates": 1414,
          "samples_seen": 1447936,
          "online_loss": 0.000705067883245647,
          "train_loss": 0.0007325534825213253,
          "validation_loss": 0.0005425962735898793,
          "train_metric": 7.378009468084201e-05,
          "validation_metric": 5.4648313380312175e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.549203938000574
        },
        {
          "epoch": 1415,
          "updates": 1415,
          "samples_seen": 1448960,
          "online_loss": 0.0007325534825213253,
          "train_loss": 0.0006697746575810015,
          "validation_loss": 0.0005888464511372149,
          "train_metric": 6.745725113432854e-05,
          "validation_metric": 5.9306454204488546e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.553557798999464
        },
        {
          "epoch": 1416,
          "updates": 1416,
          "samples_seen": 1449984,
          "online_loss": 0.0006697746575810015,
          "train_loss": 0.0006049795192666352,
          "validation_loss": 0.00048716002493165433,
          "train_metric": 6.093132105888799e-05,
          "validation_metric": 4.906497633783147e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.558407431000887
        },
        {
          "epoch": 1417,
          "updates": 1417,
          "samples_seen": 1451008,
          "online_loss": 0.0006049795774742961,
          "train_loss": 0.000590726500377059,
          "validation_loss": 0.00048296735621988773,
          "train_metric": 5.9495803725440055e-05,
          "validation_metric": 4.864270886173472e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.563592140000765
        },
        {
          "epoch": 1418,
          "updates": 1418,
          "samples_seen": 1452032,
          "online_loss": 0.0005907263839617372,
          "train_loss": 0.0006269646692089736,
          "validation_loss": 0.0005396410706453025,
          "train_metric": 6.314558413578197e-05,
          "validation_metric": 5.4350672144209966e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.56826731900037
        },
        {
          "epoch": 1419,
          "updates": 1419,
          "samples_seen": 1453056,
          "online_loss": 0.0006269646110013127,
          "train_loss": 0.000664280669298023,
          "validation_loss": 0.0005040044779889286,
          "train_metric": 6.69039145577699e-05,
          "validation_metric": 5.076147863292135e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.57299632899958
        },
        {
          "epoch": 1420,
          "updates": 1420,
          "samples_seen": 1454080,
          "online_loss": 0.000664280669298023,
          "train_loss": 0.0006515234126709402,
          "validation_loss": 0.0005650803213939071,
          "train_metric": 6.561905320268124e-05,
          "validation_metric": 5.6912820582510903e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.577714619999824
        },
        {
          "epoch": 1421,
          "updates": 1421,
          "samples_seen": 1455104,
          "online_loss": 0.0006515233544632792,
          "train_loss": 0.0006127869710326195,
          "validation_loss": 0.0004810321843251586,
          "train_metric": 6.171765562612563e-05,
          "validation_metric": 4.844780050916597e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.582839071999842
        },
        {
          "epoch": 1422,
          "updates": 1422,
          "samples_seen": 1456128,
          "online_loss": 0.0006127869710326195,
          "train_loss": 0.0005844348925165832,
          "validation_loss": 0.0004889103001914918,
          "train_metric": 5.886214421479963e-05,
          "validation_metric": 4.9241258238907903e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.58753479400002
        },
        {
          "epoch": 1423,
          "updates": 1423,
          "samples_seen": 1457152,
          "online_loss": 0.0005844348925165832,
          "train_loss": 0.0005913804634474218,
          "validation_loss": 0.0005022037075832486,
          "train_metric": 5.9561676607700065e-05,
          "validation_metric": 5.058011447545141e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.592448105000585
        },
        {
          "epoch": 1424,
          "updates": 1424,
          "samples_seen": 1458176,
          "online_loss": 0.0005913804634474218,
          "train_loss": 0.0006163750076666474,
          "validation_loss": 0.0004819714231416583,
          "train_metric": 6.20790378889069e-05,
          "validation_metric": 4.854239523410797e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.59716422100064
        },
        {
          "epoch": 1425,
          "updates": 1425,
          "samples_seen": 1459200,
          "online_loss": 0.0006163749494589865,
          "train_loss": 0.0006249193102121353,
          "validation_loss": 0.0005405727424658835,
          "train_metric": 6.293959449976683e-05,
          "validation_metric": 5.444451016956009e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.60169656799917
        },
        {
          "epoch": 1426,
          "updates": 1426,
          "samples_seen": 1460224,
          "online_loss": 0.0006249193102121353,
          "train_loss": 0.0006107774679549038,
          "validation_loss": 0.0004777427529916167,
          "train_metric": 6.151526758912951e-05,
          "validation_metric": 4.811650069314055e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.606616885999756
        },
        {
          "epoch": 1427,
          "updates": 1427,
          "samples_seen": 1461248,
          "online_loss": 0.0006107774097472429,
          "train_loss": 0.0005867744330316782,
          "validation_loss": 0.0004978964570909739,
          "train_metric": 5.909776518819854e-05,
          "validation_metric": 5.0146307330578566e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.611330863999683
        },
        {
          "epoch": 1428,
          "updates": 1428,
          "samples_seen": 1462272,
          "online_loss": 0.0005867743748240173,
          "train_loss": 0.0005772716249339283,
          "validation_loss": 0.00048042155685834587,
          "train_metric": 5.814068572362885e-05,
          "validation_metric": 4.838630047743209e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.683859087999735
        },
        {
          "epoch": 1429,
          "updates": 1429,
          "samples_seen": 1463296,
          "online_loss": 0.0005772717413492501,
          "train_loss": 0.0005854608025401831,
          "validation_loss": 0.0004726058105006814,
          "train_metric": 5.896546645089984e-05,
          "validation_metric": 4.7599125537090003e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.68891364599949
        },
        {
          "epoch": 1430,
          "updates": 1430,
          "samples_seen": 1464320,
          "online_loss": 0.0005854608025401831,
          "train_loss": 0.0005973573424853384,
          "validation_loss": 0.0005132165970280766,
          "train_metric": 6.0163645684951916e-05,
          "validation_metric": 5.1689294195966795e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.69376047700098
        },
        {
          "epoch": 1431,
          "updates": 1431,
          "samples_seen": 1465344,
          "online_loss": 0.0005973574006929994,
          "train_loss": 0.0005998550332151353,
          "validation_loss": 0.00047317883581854403,
          "train_metric": 6.041520100552589e-05,
          "validation_metric": 4.7656838432885706e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.698276147000797
        },
        {
          "epoch": 1432,
          "updates": 1432,
          "samples_seen": 1466368,
          "online_loss": 0.0005998549750074744,
          "train_loss": 0.000588715891353786,
          "validation_loss": 0.0005023779231123626,
          "train_metric": 5.9293306549079716e-05,
          "validation_metric": 5.059766044723801e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.702829399999246
        },
        {
          "epoch": 1433,
          "updates": 1433,
          "samples_seen": 1467392,
          "online_loss": 0.0005887159495614469,
          "train_loss": 0.0005761160282418132,
          "validation_loss": 0.00046952455886639655,
          "train_metric": 5.802430314361118e-05,
          "validation_metric": 4.7288787754951045e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.70794354999998
        },
        {
          "epoch": 1434,
          "updates": 1434,
          "samples_seen": 1468416,
          "online_loss": 0.0005761159700341523,
          "train_loss": 0.0005717144231311977,
          "validation_loss": 0.00047342132893390954,
          "train_metric": 5.758098268415779e-05,
          "validation_metric": 4.768126382259652e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.712694375000865
        },
        {
          "epoch": 1435,
          "updates": 1435,
          "samples_seen": 1469440,
          "online_loss": 0.0005717143649235368,
          "train_loss": 0.0005762423388659954,
          "validation_loss": 0.00048785554827190936,
          "train_metric": 5.803702151752077e-05,
          "validation_metric": 4.913502561976202e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.717394700999648
        },
        {
          "epoch": 1436,
          "updates": 1436,
          "samples_seen": 1470464,
          "online_loss": 0.0005762423970736563,
          "train_loss": 0.0005823639803566039,
          "validation_loss": 0.00046610404388047755,
          "train_metric": 5.865356797585264e-05,
          "validation_metric": 4.69442929897923e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.72253871599969
        },
        {
          "epoch": 1437,
          "updates": 1437,
          "samples_seen": 1471488,
          "online_loss": 0.0005823639803566039,
          "train_loss": 0.000582348380703479,
          "validation_loss": 0.0004955441108904779,
          "train_metric": 5.8651996369007975e-05,
          "validation_metric": 4.9909387598745525e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.727376241000457
        },
        {
          "epoch": 1438,
          "updates": 1438,
          "samples_seen": 1472512,
          "online_loss": 0.000582348380703479,
          "train_loss": 0.0005764336092397571,
          "validation_loss": 0.00046346746967174113,
          "train_metric": 5.80562737013679e-05,
          "validation_metric": 4.667875327868387e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.732523168000625
        },
        {
          "epoch": 1439,
          "updates": 1439,
          "samples_seen": 1473536,
          "online_loss": 0.0005764333182014525,
          "train_loss": 0.0005691710975952446,
          "validation_loss": 0.0004771188832819462,
          "train_metric": 5.7324828958371654e-05,
          "validation_metric": 4.805366916116327e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.737229427000784
        },
        {
          "epoch": 1440,
          "updates": 1440,
          "samples_seen": 1474560,
          "online_loss": 0.0005691710393875837,
          "train_loss": 0.0005660567549057305,
          "validation_loss": 0.00047006006934680045,
          "train_metric": 5.7011166063603014e-05,
          "validation_metric": 4.7342728066723794e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.742135233000226
        },
        {
          "epoch": 1441,
          "updates": 1441,
          "samples_seen": 1475584,
          "online_loss": 0.000566057045944035,
          "train_loss": 0.0005675919237546623,
          "validation_loss": 0.0004644024302251637,
          "train_metric": 5.716578380088322e-05,
          "validation_metric": 4.677291144616902e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.746908929999336
        },
        {
          "epoch": 1442,
          "updates": 1442,
          "samples_seen": 1476608,
          "online_loss": 0.0005675919237546623,
          "train_loss": 0.0005704102222807705,
          "validation_loss": 0.0004817676090169698,
          "train_metric": 5.744963345932774e-05,
          "validation_metric": 4.852186975767836e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.75149323400001
        },
        {
          "epoch": 1443,
          "updates": 1443,
          "samples_seen": 1477632,
          "online_loss": 0.0005704102222807705,
          "train_loss": 0.0005710858968086541,
          "validation_loss": 0.00046067044604569674,
          "train_metric": 5.7517678214935586e-05,
          "validation_metric": 4.63970354758203e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.75645884200094
        },
        {
          "epoch": 1444,
          "updates": 1444,
          "samples_seen": 1478656,
          "online_loss": 0.0005710858968086541,
          "train_loss": 0.000568134943023324,
          "validation_loss": 0.00047913671005517244,
          "train_metric": 5.722047353629023e-05,
          "validation_metric": 4.8256897571263835e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.761432315999627
        },
        {
          "epoch": 1445,
          "updates": 1445,
          "samples_seen": 1479680,
          "online_loss": 0.0005681350012309849,
          "train_loss": 0.000563828565645963,
          "validation_loss": 0.0004611184122040868,
          "train_metric": 5.678674642695114e-05,
          "validation_metric": 4.6442153688985854e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.76600030100053
        },
        {
          "epoch": 1446,
          "updates": 1446,
          "samples_seen": 1480704,
          "online_loss": 0.0005638286238536239,
          "train_loss": 0.0005607783095911145,
          "validation_loss": 0.0004673853691201657,
          "train_metric": 5.64795482205227e-05,
          "validation_metric": 4.7073346649995074e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.77042923299996
        },
        {
          "epoch": 1447,
          "updates": 1447,
          "samples_seen": 1481728,
          "online_loss": 0.0005607783677987754,
          "train_loss": 0.0005602027522400022,
          "validation_loss": 0.00046820048009976745,
          "train_metric": 5.642156975227408e-05,
          "validation_metric": 4.7155441279755905e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.77510813799927
        },
        {
          "epoch": 1448,
          "updates": 1448,
          "samples_seen": 1482752,
          "online_loss": 0.0005602027522400022,
          "train_loss": 0.0005611796514131129,
          "validation_loss": 0.000459424831205979,
          "train_metric": 5.651995161315426e-05,
          "validation_metric": 4.6271590690594167e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.78032938999968
        },
        {
          "epoch": 1449,
          "updates": 1449,
          "samples_seen": 1483776,
          "online_loss": 0.000561179593205452,
          "train_loss": 0.00056181859690696,
          "validation_loss": 0.0004733136447612196,
          "train_metric": 5.6584318372188136e-05,
          "validation_metric": 4.7670411731814966e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.785029591999773
        },
        {
          "epoch": 1450,
          "updates": 1450,
          "samples_seen": 1484800,
          "online_loss": 0.0005618186551146209,
          "train_loss": 0.0005609557847492397,
          "validation_loss": 0.0004566027200780809,
          "train_metric": 5.6497407058486715e-05,
          "validation_metric": 4.598735540639609e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.79013919099998
        },
        {
          "epoch": 1451,
          "updates": 1451,
          "samples_seen": 1485824,
          "online_loss": 0.0005609557265415788,
          "train_loss": 0.0005586296319961548,
          "validation_loss": 0.0004694405652116984,
          "train_metric": 5.6263124861288816e-05,
          "validation_metric": 4.7280333092203364e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.794961171999603
        },
        {
          "epoch": 1452,
          "updates": 1452,
          "samples_seen": 1486848,
          "online_loss": 0.0005586295737884939,
          "train_loss": 0.0005560973077081144,
          "validation_loss": 0.0004577520303428173,
          "train_metric": 5.600808799499646e-05,
          "validation_metric": 4.610310497810133e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.79967522699917
        },
        {
          "epoch": 1453,
          "updates": 1453,
          "samples_seen": 1487872,
          "online_loss": 0.0005560973077081144,
          "train_loss": 0.0005543678416870534,
          "validation_loss": 0.00046055589336901903,
          "train_metric": 5.583390884567052e-05,
          "validation_metric": 4.638550308300182e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.804212663000726
        },
        {
          "epoch": 1454,
          "updates": 1454,
          "samples_seen": 1488896,
          "online_loss": 0.0005543678998947144,
          "train_loss": 0.0005537724937312305,
          "validation_loss": 0.00046114774886518717,
          "train_metric": 5.5773936765035614e-05,
          "validation_metric": 4.644510772777721e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.808985751000364
        },
        {
          "epoch": 1455,
          "updates": 1455,
          "samples_seen": 1489920,
          "online_loss": 0.0005537725519388914,
          "train_loss": 0.000553830002900213,
          "validation_loss": 0.0004539184446912259,
          "train_metric": 5.5779728427296504e-05,
          "validation_metric": 4.571700264932588e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.81407198099987
        },
        {
          "epoch": 1456,
          "updates": 1456,
          "samples_seen": 1490944,
          "online_loss": 0.0005538299446925521,
          "train_loss": 0.0005536802927963436,
          "validation_loss": 0.00046367940376512706,
          "train_metric": 5.576464900514111e-05,
          "validation_metric": 4.670009366236627e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.818590718999985
        },
        {
          "epoch": 1457,
          "updates": 1457,
          "samples_seen": 1491968,
          "online_loss": 0.0005536802345886827,
          "train_loss": 0.0005528501933440566,
          "validation_loss": 0.0004519923822954297,
          "train_metric": 5.568105189013295e-05,
          "validation_metric": 4.552302198135294e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.823722172999624
        },
        {
          "epoch": 1458,
          "updates": 1458,
          "samples_seen": 1492992,
          "online_loss": 0.0005528501933440566,
          "train_loss": 0.0005513530923053622,
          "validation_loss": 0.00046152074355632067,
          "train_metric": 5.553026494453661e-05,
          "validation_metric": 4.6482680772896856e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.828597211999295
        },
        {
          "epoch": 1459,
          "updates": 1459,
          "samples_seen": 1494016,
          "online_loss": 0.0005513529758900404,
          "train_loss": 0.0005496603553183377,
          "validation_loss": 0.0004522453236859292,
          "train_metric": 5.535976742976345e-05,
          "validation_metric": 4.554849147098139e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.83328796100068
        },
        {
          "epoch": 1460,
          "updates": 1460,
          "samples_seen": 1495040,
          "online_loss": 0.0005496602971106768,
          "train_loss": 0.0005482371198013425,
          "validation_loss": 0.00045499749830923975,
          "train_metric": 5.521644197870046e-05,
          "validation_metric": 4.582567999023013e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.837878278000062
        },
        {
          "epoch": 1461,
          "updates": 1461,
          "samples_seen": 1496064,
          "online_loss": 0.0005482371198013425,
          "train_loss": 0.0005473322817124426,
          "validation_loss": 0.00045375784975476563,
          "train_metric": 5.512531060958281e-05,
          "validation_metric": 4.570082455757074e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.842706707000616
        },
        {
          "epoch": 1462,
          "updates": 1462,
          "samples_seen": 1497088,
          "online_loss": 0.0005473323399201035,
          "train_loss": 0.0005468152230605483,
          "validation_loss": 0.0004499570350162685,
          "train_metric": 5.507322566700168e-05,
          "validation_metric": 4.531802187557332e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.84774441800073
        },
        {
          "epoch": 1463,
          "updates": 1463,
          "samples_seen": 1498112,
          "online_loss": 0.0005468152812682092,
          "train_loss": 0.0005463673733174801,
          "validation_loss": 0.00045613147085532546,
          "train_metric": 5.502811836777255e-05,
          "validation_metric": 4.5939890696899965e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.852179429999524
        },
        {
          "epoch": 1464,
          "updates": 1464,
          "samples_seen": 1499136,
          "online_loss": 0.0005463673733174801,
          "train_loss": 0.000545721675734967,
          "validation_loss": 0.00044821170740760863,
          "train_metric": 5.496308585861698e-05,
          "validation_metric": 4.514223837759346e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.857186765000733
        },
        {
          "epoch": 1465,
          "updates": 1465,
          "samples_seen": 1500160,
          "online_loss": 0.0005457216175273061,
          "train_loss": 0.0005447505973279476,
          "validation_loss": 0.00045545175089500844,
          "train_metric": 5.4865282436367124e-05,
          "validation_metric": 4.5871427573729306e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.86190660899956
        },
        {
          "epoch": 1466,
          "updates": 1466,
          "samples_seen": 1501184,
          "online_loss": 0.0005447505973279476,
          "train_loss": 0.0005435598432086408,
          "validation_loss": 0.0004471616994123906,
          "train_metric": 5.474536010297015e-05,
          "validation_metric": 4.5036489609628916e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.931578286999866
        },
        {
          "epoch": 1467,
          "updates": 1467,
          "samples_seen": 1502208,
          "online_loss": 0.0005435599014163017,
          "train_loss": 0.000542330089956522,
          "validation_loss": 0.00045122497249394655,
          "train_metric": 5.462150147650391e-05,
          "validation_metric": 4.5445722207659855e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.93619876900084
        },
        {
          "epoch": 1468,
          "updates": 1468,
          "samples_seen": 1503232,
          "online_loss": 0.000542330089956522,
          "train_loss": 0.0005412273458205163,
          "validation_loss": 0.000447401573183015,
          "train_metric": 5.4510437621502206e-05,
          "validation_metric": 4.506065306486562e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.94076650399984
        },
        {
          "epoch": 1469,
          "updates": 1469,
          "samples_seen": 1504256,
          "online_loss": 0.0005412273458205163,
          "train_loss": 0.000540319480933249,
          "validation_loss": 0.0004474038432817906,
          "train_metric": 5.4418997024185956e-05,
          "validation_metric": 4.506087498157285e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.9454093380009
        },
        {
          "epoch": 1470,
          "updates": 1470,
          "samples_seen": 1505280,
          "online_loss": 0.0005403194227255881,
          "train_loss": 0.0005395814660005271,
          "validation_loss": 0.0004489124985411763,
          "train_metric": 5.4344669479178265e-05,
          "validation_metric": 4.521282244240865e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.949929337999492
        },
        {
          "epoch": 1471,
          "updates": 1471,
          "samples_seen": 1506304,
          "online_loss": 0.0005395814660005271,
          "train_loss": 0.0005389038706198335,
          "validation_loss": 0.00044468080159276724,
          "train_metric": 5.427641735877842e-05,
          "validation_metric": 4.4786622311221436e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.95479301100022
        },
        {
          "epoch": 1472,
          "updates": 1472,
          "samples_seen": 1507328,
          "online_loss": 0.000538904161658138,
          "train_loss": 0.0005381773808039725,
          "validation_loss": 0.0004488042904995382,
          "train_metric": 5.42032576049678e-05,
          "validation_metric": 4.520192669588141e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.959588421999797
        },
        {
          "epoch": 1473,
          "updates": 1473,
          "samples_seen": 1508352,
          "online_loss": 0.0005381774390116334,
          "train_loss": 0.0005373538588173687,
          "validation_loss": 0.0004423856735229492,
          "train_metric": 5.4120318964123726e-05,
          "validation_metric": 4.455546149984002e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.96462878500097
        },
        {
          "epoch": 1474,
          "updates": 1474,
          "samples_seen": 1509376,
          "online_loss": 0.0005373538588173687,
          "train_loss": 0.0005364184034988284,
          "validation_loss": 0.0004469354171305895,
          "train_metric": 5.402609895099886e-05,
          "validation_metric": 4.5013697672402486e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.969104060000973
        },
        {
          "epoch": 1475,
          "updates": 1475,
          "samples_seen": 1510400,
          "online_loss": 0.0005364183452911675,
          "train_loss": 0.0005354227032512426,
          "validation_loss": 0.00044155243085697293,
          "train_metric": 5.392581078922376e-05,
          "validation_metric": 4.4471547880675644e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.974024073999317
        },
        {
          "epoch": 1476,
          "updates": 1476,
          "samples_seen": 1511424,
          "online_loss": 0.0005354227614589036,
          "train_loss": 0.0005344214150682092,
          "validation_loss": 0.00044449116103351116,
          "train_metric": 5.3824958740733564e-05,
          "validation_metric": 4.47675192845054e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.978689114999725
        },
        {
          "epoch": 1477,
          "updates": 1477,
          "samples_seen": 1512448,
          "online_loss": 0.0005344214732758701,
          "train_loss": 0.0005334655288606882,
          "validation_loss": 0.00044146832078695297,
          "train_metric": 5.3728686907561496e-05,
          "validation_metric": 4.4463064114097506e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.98374514200077
        },
        {
          "epoch": 1478,
          "updates": 1478,
          "samples_seen": 1513472,
          "online_loss": 0.0005334655288606882,
          "train_loss": 0.0005325719830580056,
          "validation_loss": 0.00044142690603621304,
          "train_metric": 5.3638697863789275e-05,
          "validation_metric": 4.445890226634219e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.9885999170001
        },
        {
          "epoch": 1479,
          "updates": 1479,
          "samples_seen": 1514496,
          "online_loss": 0.0005325720412656665,
          "train_loss": 0.0005317418836057186,
          "validation_loss": 0.0004411019035615027,
          "train_metric": 5.355509711080231e-05,
          "validation_metric": 4.4426164095057175e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.99367199900007
        },
        {
          "epoch": 1480,
          "updates": 1480,
          "samples_seen": 1515520,
          "online_loss": 0.0005317418836057186,
          "train_loss": 0.0005309470579959452,
          "validation_loss": 0.0004387258377391845,
          "train_metric": 5.347505066310987e-05,
          "validation_metric": 4.418686148710549e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 16.998691146000056
        },
        {
          "epoch": 1481,
          "updates": 1481,
          "samples_seen": 1516544,
          "online_loss": 0.0005309470579959452,
          "train_loss": 0.0005301591008901596,
          "validation_loss": 0.00044078106293454766,
          "train_metric": 5.339568815543316e-05,
          "validation_metric": 4.439385520527139e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.003062572999625
        },
        {
          "epoch": 1482,
          "updates": 1482,
          "samples_seen": 1517568,
          "online_loss": 0.0005301592173054814,
          "train_loss": 0.0005293604335747659,
          "validation_loss": 0.000437001435784623,
          "train_metric": 5.331524880602956e-05,
          "validation_metric": 4.401318437885493e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.00808438199965
        },
        {
          "epoch": 1483,
          "updates": 1483,
          "samples_seen": 1518592,
          "online_loss": 0.000529360375367105,
          "train_loss": 0.0005285333609208465,
          "validation_loss": 0.0004398902819957584,
          "train_metric": 5.323194636730477e-05,
          "validation_metric": 4.43041353719309e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.012910968000142
        },
        {
          "epoch": 1484,
          "updates": 1484,
          "samples_seen": 1519616,
          "online_loss": 0.0005285335355438292,
          "train_loss": 0.000527676660567522,
          "validation_loss": 0.0004354965058155358,
          "train_metric": 5.314566078595817e-05,
          "validation_metric": 4.386161162983626e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.01780622700062
        },
        {
          "epoch": 1485,
          "updates": 1485,
          "samples_seen": 1520640,
          "online_loss": 0.000527676660567522,
          "train_loss": 0.0005267952801659703,
          "validation_loss": 0.00043807164183817804,
          "train_metric": 5.305689774104394e-05,
          "validation_metric": 4.412097405293025e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.02235722200021
        },
        {
          "epoch": 1486,
          "updates": 1486,
          "samples_seen": 1521664,
          "online_loss": 0.0005267952801659703,
          "train_loss": 0.0005259046447463334,
          "validation_loss": 0.000434350484283641,
          "train_metric": 5.2967188821639866e-05,
          "validation_metric": 4.374618947622366e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.027575223000895
        },
        {
          "epoch": 1487,
          "updates": 1487,
          "samples_seen": 1522688,
          "online_loss": 0.0005259045865386724,
          "train_loss": 0.0005250166868790984,
          "validation_loss": 0.00043614074820652604,
          "train_metric": 5.287776002660394e-05,
          "validation_metric": 4.392649861983955e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.03222620899942
        },
        {
          "epoch": 1488,
          "updates": 1488,
          "samples_seen": 1523712,
          "online_loss": 0.0005250166868790984,
          "train_loss": 0.0005241371691226959,
          "validation_loss": 0.00043363610166125,
          "train_metric": 5.278917160467245e-05,
          "validation_metric": 4.367424116935581e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.037056359000417
        },
        {
          "epoch": 1489,
          "updates": 1489,
          "samples_seen": 1524736,
          "online_loss": 0.0005241371691226959,
          "train_loss": 0.0005232728435657918,
          "validation_loss": 0.00043406765325926244,
          "train_metric": 5.270213296171278e-05,
          "validation_metric": 4.371770774014294e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.041800177999903
        },
        {
          "epoch": 1490,
          "updates": 1490,
          "samples_seen": 1525760,
          "online_loss": 0.0005232729017734528,
          "train_loss": 0.0005224233027547598,
          "validation_loss": 0.00043270751484669745,
          "train_metric": 5.2616564062191173e-05,
          "validation_metric": 4.3580716010183096e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.046789983000053
        },
        {
          "epoch": 1491,
          "updates": 1491,
          "samples_seen": 1526784,
          "online_loss": 0.0005224233027547598,
          "train_loss": 0.0005215857527218759,
          "validation_loss": 0.0004318776773288846,
          "train_metric": 5.253219569567591e-05,
          "validation_metric": 4.349713708506897e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.051574638999227
        },
        {
          "epoch": 1492,
          "updates": 1492,
          "samples_seen": 1527808,
          "online_loss": 0.000521585694514215,
          "train_loss": 0.0005207590293139219,
          "validation_loss": 0.00043176798499189317,
          "train_metric": 5.2448940550675616e-05,
          "validation_metric": 4.348609218141064e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.05641523399936
        },
        {
          "epoch": 1493,
          "updates": 1493,
          "samples_seen": 1528832,
          "online_loss": 0.0005207590875215828,
          "train_loss": 0.0005199371371418238,
          "validation_loss": 0.0004301168955862522,
          "train_metric": 5.2366169256856665e-05,
          "validation_metric": 4.331979653215967e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.06139297899972
        },
        {
          "epoch": 1494,
          "updates": 1494,
          "samples_seen": 1529856,
          "online_loss": 0.0005199371953494847,
          "train_loss": 0.0005191192030906677,
          "validation_loss": 0.00043090121471323073,
          "train_metric": 5.228379450272769e-05,
          "validation_metric": 4.339879160397686e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.065973667000435
        },
        {
          "epoch": 1495,
          "updates": 1495,
          "samples_seen": 1530880,
          "online_loss": 0.0005191192612983286,
          "train_loss": 0.0005183052853681147,
          "validation_loss": 0.0004284843453206122,
          "train_metric": 5.2201812650309876e-05,
          "validation_metric": 4.3155374441994354e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.070833141999174
        },
        {
          "epoch": 1496,
          "updates": 1496,
          "samples_seen": 1531904,
          "online_loss": 0.0005183052853681147,
          "train_loss": 0.0005174851976335049,
          "validation_loss": 0.0004296762344893068,
          "train_metric": 5.211921961745247e-05,
          "validation_metric": 4.327541682869196e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.075426253999467
        },
        {
          "epoch": 1497,
          "updates": 1497,
          "samples_seen": 1532928,
          "online_loss": 0.0005174851976335049,
          "train_loss": 0.0005166665650904179,
          "validation_loss": 0.0004268112243153155,
          "train_metric": 5.203676118981093e-05,
          "validation_metric": 4.298686690162867e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.080507843000305
        },
        {
          "epoch": 1498,
          "updates": 1498,
          "samples_seen": 1533952,
          "online_loss": 0.0005166665650904179,
          "train_loss": 0.0005158449639566243,
          "validation_loss": 0.0004283838497940451,
          "train_metric": 5.195401899982244e-05,
          "validation_metric": 4.3145253584953025e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.084893616000045
        },
        {
          "epoch": 1499,
          "updates": 1499,
          "samples_seen": 1534976,
          "online_loss": 0.0005158446729183197,
          "train_loss": 0.0005150241195224226,
          "validation_loss": 0.0004254216910339892,
          "train_metric": 5.187134229345247e-05,
          "validation_metric": 4.284691021894105e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.08989192899935
        },
        {
          "epoch": 1500,
          "updates": 1500,
          "samples_seen": 1536000,
          "online_loss": 0.0005150240613147616,
          "train_loss": 0.0005142037989571691,
          "validation_loss": 0.00042717866017483175,
          "train_metric": 5.178872015676461e-05,
          "validation_metric": 4.30238687840756e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.09467726100047
        },
        {
          "epoch": 1501,
          "updates": 1501,
          "samples_seen": 1537024,
          "online_loss": 0.0005142037989571691,
          "train_loss": 0.0005133816739544272,
          "validation_loss": 0.0004240309353917837,
          "train_metric": 5.170592339709401e-05,
          "validation_metric": 4.27068407589104e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.09968326300077
        },
        {
          "epoch": 1502,
          "updates": 1502,
          "samples_seen": 1538048,
          "online_loss": 0.0005133816739544272,
          "train_loss": 0.0005125622265040874,
          "validation_loss": 0.00042579026194289327,
          "train_metric": 5.1623384933918715e-05,
          "validation_metric": 4.2884039430646226e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.104055289000826
        },
        {
          "epoch": 1503,
          "updates": 1503,
          "samples_seen": 1539072,
          "online_loss": 0.0005125622265040874,
          "train_loss": 0.0005117473774589598,
          "validation_loss": 0.00042250112164765596,
          "train_metric": 5.1541326683945954e-05,
          "validation_metric": 4.255276871845126e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.10908997200022
        },
        {
          "epoch": 1504,
          "updates": 1504,
          "samples_seen": 1540096,
          "online_loss": 0.0005117474356666207,
          "train_loss": 0.0005109328194521368,
          "validation_loss": 0.00042445058352313936,
          "train_metric": 5.145928662386723e-05,
          "validation_metric": 4.274911043467e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.11366992799958
        },
        {
          "epoch": 1505,
          "updates": 1505,
          "samples_seen": 1541120,
          "online_loss": 0.0005109328194521368,
          "train_loss": 0.000510124082211405,
          "validation_loss": 0.00042106330511160195,
          "train_metric": 5.137783591635525e-05,
          "validation_metric": 4.240795533405617e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.18616083699999
        },
        {
          "epoch": 1506,
          "updates": 1506,
          "samples_seen": 1542144,
          "online_loss": 0.000510124140419066,
          "train_loss": 0.0005093234940432012,
          "validation_loss": 0.0004232875653542578,
          "train_metric": 5.129720375407487e-05,
          "validation_metric": 4.2631971155060455e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.191041441999914
        },
        {
          "epoch": 1507,
          "updates": 1507,
          "samples_seen": 1543168,
          "online_loss": 0.0005093235522508621,
          "train_loss": 0.0005085280281491578,
          "validation_loss": 0.00041952752508223057,
          "train_metric": 5.121706999489106e-05,
          "validation_metric": 4.2253272113157436e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.19614983899919
        },
        {
          "epoch": 1508,
          "updates": 1508,
          "samples_seen": 1544192,
          "online_loss": 0.0005085280281491578,
          "train_loss": 0.0005077425157651305,
          "validation_loss": 0.00042214387212879956,
          "train_metric": 5.113796578370966e-05,
          "validation_metric": 4.251679274602793e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.20057063599961
        },
        {
          "epoch": 1509,
          "updates": 1509,
          "samples_seen": 1545216,
          "online_loss": 0.0005077425157651305,
          "train_loss": 0.0005069756298325956,
          "validation_loss": 0.0004178038507234305,
          "train_metric": 5.10607314936351e-05,
          "validation_metric": 4.207967140246183e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.205391761999635
        },
        {
          "epoch": 1510,
          "updates": 1510,
          "samples_seen": 1546240,
          "online_loss": 0.0005069756298325956,
          "train_loss": 0.0005062247510068119,
          "validation_loss": 0.0004212360945530236,
          "train_metric": 5.098510519019328e-05,
          "validation_metric": 4.242535578669049e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.21006412600036
        },
        {
          "epoch": 1511,
          "updates": 1511,
          "samples_seen": 1547264,
          "online_loss": 0.0005062247510068119,
          "train_loss": 0.0005055056535638869,
          "validation_loss": 0.00041602374403737485,
          "train_metric": 5.091267667012289e-05,
          "validation_metric": 4.190039180684835e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.21506115900047
        },
        {
          "epoch": 1512,
          "updates": 1512,
          "samples_seen": 1548288,
          "online_loss": 0.0005055056535638869,
          "train_loss": 0.0005048306775279343,
          "validation_loss": 0.0004207231104373932,
          "train_metric": 5.0844697398133576e-05,
          "validation_metric": 4.237368921167217e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.219521015000282
        },
        {
          "epoch": 1513,
          "updates": 1513,
          "samples_seen": 1549312,
          "online_loss": 0.0005048306193202734,
          "train_loss": 0.0005042211269028485,
          "validation_loss": 0.0004140275705140084,
          "train_metric": 5.078329922980629e-05,
          "validation_metric": 4.169934254605323e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.22443055000076
        },
        {
          "epoch": 1514,
          "updates": 1514,
          "samples_seen": 1550336,
          "online_loss": 0.0005042210686951876,
          "train_loss": 0.0005037054070271552,
          "validation_loss": 0.0004207442398183048,
          "train_metric": 5.0731367082335055e-05,
          "validation_metric": 4.2375824705231935e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.229057601000022
        },
        {
          "epoch": 1515,
          "updates": 1515,
          "samples_seen": 1551360,
          "online_loss": 0.0005037054070271552,
          "train_loss": 0.0005033475463278592,
          "validation_loss": 0.00041181081905961037,
          "train_metric": 5.069531835033558e-05,
          "validation_metric": 4.147607978666201e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.233925532000285
        },
        {
          "epoch": 1516,
          "updates": 1516,
          "samples_seen": 1552384,
          "online_loss": 0.0005033475463278592,
          "train_loss": 0.0005032083136029541,
          "validation_loss": 0.0004219458205625415,
          "train_metric": 5.068130121799186e-05,
          "validation_metric": 4.2496838432271034e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.23835388000043
        },
        {
          "epoch": 1517,
          "updates": 1517,
          "samples_seen": 1553408,
          "online_loss": 0.0005032082553952932,
          "train_loss": 0.000503474788274616,
          "validation_loss": 0.00040959176840260625,
          "train_metric": 5.0708138587651774e-05,
          "validation_metric": 4.125258419662714e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.243551257000945
        },
        {
          "epoch": 1518,
          "updates": 1518,
          "samples_seen": 1554432,
          "online_loss": 0.000503474788274616,
          "train_loss": 0.0005042847478762269,
          "validation_loss": 0.00042544398456811905,
          "train_metric": 5.078971298644319e-05,
          "validation_metric": 4.2849165765801445e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.248134441999355
        },
        {
          "epoch": 1519,
          "updates": 1519,
          "samples_seen": 1555456,
          "online_loss": 0.0005042847478762269,
          "train_loss": 0.0005061643314547837,
          "validation_loss": 0.00040785642340779305,
          "train_metric": 5.0979018851649016e-05,
          "validation_metric": 4.107780841877684e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.25292657099999
        },
        {
          "epoch": 1520,
          "updates": 1520,
          "samples_seen": 1556480,
          "online_loss": 0.0005061643314547837,
          "train_loss": 0.0005094252992421389,
          "validation_loss": 0.0004337677964940667,
          "train_metric": 5.130744830239564e-05,
          "validation_metric": 4.3687505240086466e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.257387091000055
        },
        {
          "epoch": 1521,
          "updates": 1521,
          "samples_seen": 1557504,
          "online_loss": 0.000509425241034478,
          "train_loss": 0.0005156989791430533,
          "validation_loss": 0.0004088868445251137,
          "train_metric": 5.193931428948417e-05,
          "validation_metric": 4.118159267818555e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.262010361000648
        },
        {
          "epoch": 1522,
          "updates": 1522,
          "samples_seen": 1558528,
          "online_loss": 0.0005156989209353924,
          "train_loss": 0.0005254875286482275,
          "validation_loss": 0.0004534648614935577,
          "train_metric": 5.292517744237557e-05,
          "validation_metric": 4.5671324187424034e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.266552487999434
        },
        {
          "epoch": 1523,
          "updates": 1523,
          "samples_seen": 1559552,
          "online_loss": 0.0005254874704405665,
          "train_loss": 0.0005440208478830755,
          "validation_loss": 0.000420108757680282,
          "train_metric": 5.479178435052745e-05,
          "validation_metric": 4.231182538205758e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.270907132000502
        },
        {
          "epoch": 1524,
          "updates": 1524,
          "samples_seen": 1560576,
          "online_loss": 0.0005440208478830755,
          "train_loss": 0.0005708049866370857,
          "validation_loss": 0.0005004117265343666,
          "train_metric": 5.748939292971045e-05,
          "validation_metric": 5.03996379848104e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.27555468600076
        },
        {
          "epoch": 1525,
          "updates": 1525,
          "samples_seen": 1561600,
          "online_loss": 0.0005708049866370857,
          "train_loss": 0.000622880703303963,
          "validation_loss": 0.0004628786991816014,
          "train_metric": 6.273425969993696e-05,
          "validation_metric": 4.661945058614947e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.280122042000585
        },
        {
          "epoch": 1526,
          "updates": 1526,
          "samples_seen": 1562624,
          "online_loss": 0.000622880645096302,
          "train_loss": 0.0006880131550133228,
          "validation_loss": 0.0006095843273214996,
          "train_metric": 6.929416122147813e-05,
          "validation_metric": 6.139510514913127e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.28447074200085
        },
        {
          "epoch": 1527,
          "updates": 1527,
          "samples_seen": 1563648,
          "online_loss": 0.0006880130968056619,
          "train_loss": 0.0008179120486602187,
          "validation_loss": 0.0005847122520208359,
          "train_metric": 8.237711881520227e-05,
          "validation_metric": 5.8890080254059285e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.288834864999444
        },
        {
          "epoch": 1528,
          "updates": 1528,
          "samples_seen": 1564672,
          "online_loss": 0.0008179121650755405,
          "train_loss": 0.0009281906532123685,
          "validation_loss": 0.0008206518250517547,
          "train_metric": 9.348396270070225e-05,
          "validation_metric": 8.265305223176256e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.29348602800019
        },
        {
          "epoch": 1529,
          "updates": 1529,
          "samples_seen": 1565696,
          "online_loss": 0.0009281905950047076,
          "train_loss": 0.0011433782055974007,
          "validation_loss": 0.0008020823006518185,
          "train_metric": 0.00011515685764607042,
          "validation_metric": 8.078279643086717e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.298163367999223
        },
        {
          "epoch": 1530,
          "updates": 1530,
          "samples_seen": 1566720,
          "online_loss": 0.001143377972766757,
          "train_loss": 0.001148001290857792,
          "validation_loss": 0.0010113156167790294,
          "train_metric": 0.00011562246800167486,
          "validation_metric": 0.00010185599967371672,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.302693435000037
        },
        {
          "epoch": 1531,
          "updates": 1531,
          "samples_seen": 1567744,
          "online_loss": 0.001148001290857792,
          "train_loss": 0.0011784780072048306,
          "validation_loss": 0.0008231350220739841,
          "train_metric": 0.00011869200534420088,
          "validation_metric": 8.290314144687727e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.307186011999875
        },
        {
          "epoch": 1532,
          "updates": 1532,
          "samples_seen": 1568768,
          "online_loss": 0.0011784780072048306,
          "train_loss": 0.000883614644408226,
          "validation_loss": 0.0007797103025950491,
          "train_metric": 8.899442764231935e-05,
          "validation_metric": 7.852956332499161e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.311893817000964
        },
        {
          "epoch": 1533,
          "updates": 1533,
          "samples_seen": 1569792,
          "online_loss": 0.0008836147026158869,
          "train_loss": 0.0006375927478075027,
          "validation_loss": 0.000458461552625522,
          "train_metric": 6.421601574402303e-05,
          "validation_metric": 4.617456943378784e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.316364766000333
        },
        {
          "epoch": 1534,
          "updates": 1534,
          "samples_seen": 1570816,
          "online_loss": 0.0006375926895998418,
          "train_loss": 0.0004970065783709288,
          "validation_loss": 0.0004012722347397357,
          "train_metric": 5.005668936064467e-05,
          "validation_metric": 4.0414674003841355e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.321202021999852
        },
        {
          "epoch": 1535,
          "updates": 1535,
          "samples_seen": 1571840,
          "online_loss": 0.0004970065783709288,
          "train_loss": 0.0005451742908917367,
          "validation_loss": 0.0004569370357785374,
          "train_metric": 5.490795592777431e-05,
          "validation_metric": 4.6021024900255725e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.3260331879992
        },
        {
          "epoch": 1536,
          "updates": 1536,
          "samples_seen": 1572864,
          "online_loss": 0.0005451742326840758,
          "train_loss": 0.0007055729511193931,
          "validation_loss": 0.0005066493758931756,
          "train_metric": 7.106272096280009e-05,
          "validation_metric": 5.1027865993091837e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.330648200000724
        },
        {
          "epoch": 1537,
          "updates": 1537,
          "samples_seen": 1573888,
          "online_loss": 0.0007055728347040713,
          "train_loss": 0.0008025987190194428,
          "validation_loss": 0.0006962225888855755,
          "train_metric": 8.083481225185096e-05,
          "validation_metric": 7.012098649283871e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.334926202000133
        },
        {
          "epoch": 1538,
          "updates": 1538,
          "samples_seen": 1574912,
          "online_loss": 0.0008025987772271037,
          "train_loss": 0.0008302486385218799,
          "validation_loss": 0.0005934017826803029,
          "train_metric": 8.36195977171883e-05,
          "validation_metric": 5.97652506257873e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.339362231999985
        },
        {
          "epoch": 1539,
          "updates": 1539,
          "samples_seen": 1575936,
          "online_loss": 0.000830248580314219,
          "train_loss": 0.0006780032999813557,
          "validation_loss": 0.0005952009232714772,
          "train_metric": 6.828601181041449e-05,
          "validation_metric": 5.994646198814735e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.343994483000643
        },
        {
          "epoch": 1540,
          "updates": 1540,
          "samples_seen": 1576960,
          "online_loss": 0.0006780032999813557,
          "train_loss": 0.0005397980567067862,
          "validation_loss": 0.0004054719174746424,
          "train_metric": 5.436647916212678e-05,
          "validation_metric": 4.0837643609847873e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.348345398000674
        },
        {
          "epoch": 1541,
          "updates": 1541,
          "samples_seen": 1577984,
          "online_loss": 0.0005397980567067862,
          "train_loss": 0.0004851394332945347,
          "validation_loss": 0.00039105696487240493,
          "train_metric": 4.886146416538395e-05,
          "validation_metric": 3.9385820855386555e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.353312368000843
        },
        {
          "epoch": 1542,
          "updates": 1542,
          "samples_seen": 1579008,
          "online_loss": 0.0004851394332945347,
          "train_loss": 0.0005370716680772603,
          "validation_loss": 0.0004642627318389714,
          "train_metric": 5.409189179772511e-05,
          "validation_metric": 4.675883610616438e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.35814762500013
        },
        {
          "epoch": 1543,
          "updates": 1543,
          "samples_seen": 1580032,
          "online_loss": 0.0005370717262849212,
          "train_loss": 0.0006295662024058402,
          "validation_loss": 0.00046238029608502984,
          "train_metric": 6.340761319734156e-05,
          "validation_metric": 4.656925375456922e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.3633074570007
        },
        {
          "epoch": 1544,
          "updates": 1544,
          "samples_seen": 1581056,
          "online_loss": 0.0006295662024058402,
          "train_loss": 0.0006531259277835488,
          "validation_loss": 0.0005855928175151348,
          "train_metric": 6.578044849447906e-05,
          "validation_metric": 5.897876326343976e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.435636010000962
        },
        {
          "epoch": 1545,
          "updates": 1545,
          "samples_seen": 1582080,
          "online_loss": 0.0006531259859912097,
          "train_loss": 0.0006180330528877676,
          "validation_loss": 0.000456235051387921,
          "train_metric": 6.22460211161524e-05,
          "validation_metric": 4.59503207821399e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.44017560499924
        },
        {
          "epoch": 1546,
          "updates": 1546,
          "samples_seen": 1583104,
          "online_loss": 0.0006180329946801066,
          "train_loss": 0.0005311911227181554,
          "validation_loss": 0.00046447059139609337,
          "train_metric": 5.349961793399416e-05,
          "validation_metric": 4.6779772674199194e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.444783801000085
        },
        {
          "epoch": 1547,
          "updates": 1547,
          "samples_seen": 1584128,
          "online_loss": 0.0005311910645104945,
          "train_loss": 0.0004828257951885462,
          "validation_loss": 0.00038747701910324395,
          "train_metric": 4.862844798481092e-05,
          "validation_metric": 3.90252644137945e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.44984372699946
        },
        {
          "epoch": 1548,
          "updates": 1548,
          "samples_seen": 1585152,
          "online_loss": 0.00048282582429237664,
          "train_loss": 0.0004946818808093667,
          "validation_loss": 0.00038709936779923737,
          "train_metric": 4.982254176866263e-05,
          "validation_metric": 3.898723662132397e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.454626287999417
        },
        {
          "epoch": 1549,
          "updates": 1549,
          "samples_seen": 1586176,
          "online_loss": 0.0004946818226017058,
          "train_loss": 0.000540065229870379,
          "validation_loss": 0.0004691201029345393,
          "train_metric": 5.4393392929341644e-05,
          "validation_metric": 4.7248056944226846e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.45932424900093
        },
        {
          "epoch": 1550,
          "updates": 1550,
          "samples_seen": 1587200,
          "online_loss": 0.000540065229870379,
          "train_loss": 0.0005786194815300405,
          "validation_loss": 0.0004365675267763436,
          "train_metric": 5.827642962685786e-05,
          "validation_metric": 4.3969484977424145e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.463872790000096
        },
        {
          "epoch": 1551,
          "updates": 1551,
          "samples_seen": 1588224,
          "online_loss": 0.0005786194815300405,
          "train_loss": 0.0005653667030856013,
          "validation_loss": 0.0004946199478581548,
          "train_metric": 5.6941666116472334e-05,
          "validation_metric": 4.9816309910966083e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.46823922099975
        },
        {
          "epoch": 1552,
          "updates": 1552,
          "samples_seen": 1589248,
          "online_loss": 0.0005653667030856013,
          "train_loss": 0.0005272201378829777,
          "validation_loss": 0.000404076388804242,
          "train_metric": 5.309967673383653e-05,
          "validation_metric": 4.0697090298635885e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.472578143000646
        },
        {
          "epoch": 1553,
          "updates": 1553,
          "samples_seen": 1590272,
          "online_loss": 0.0005272200796753168,
          "train_loss": 0.0004868111864197999,
          "validation_loss": 0.0004074200114700943,
          "train_metric": 4.902984801447019e-05,
          "validation_metric": 4.1033847082871944e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.47728893400017
        },
        {
          "epoch": 1554,
          "updates": 1554,
          "samples_seen": 1591296,
          "online_loss": 0.00048681124462746084,
          "train_loss": 0.00047629070468246937,
          "validation_loss": 0.00038919170037843287,
          "train_metric": 4.797025030711666e-05,
          "validation_metric": 3.919796290574595e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.48181050200037
        },
        {
          "epoch": 1555,
          "updates": 1555,
          "samples_seen": 1592320,
          "online_loss": 0.0004762907337862998,
          "train_loss": 0.0004938258789479733,
          "validation_loss": 0.00038421055069193244,
          "train_metric": 4.973632894689217e-05,
          "validation_metric": 3.86962819902692e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.486528065999664
        },
        {
          "epoch": 1556,
          "updates": 1556,
          "samples_seen": 1593344,
          "online_loss": 0.0004938258789479733,
          "train_loss": 0.0005159570719115436,
          "validation_loss": 0.00044589940807782114,
          "train_metric": 5.196530764806084e-05,
          "validation_metric": 4.490935680223629e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.491026900999714
        },
        {
          "epoch": 1557,
          "updates": 1557,
          "samples_seen": 1594368,
          "online_loss": 0.0005159570719115436,
          "train_loss": 0.0005241628387011588,
          "validation_loss": 0.00040233152685686946,
          "train_metric": 5.27917654835619e-05,
          "validation_metric": 4.052135773235932e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.49560847400062
        },
        {
          "epoch": 1558,
          "updates": 1558,
          "samples_seen": 1595392,
          "online_loss": 0.0005241628969088197,
          "train_loss": 0.000507468415889889,
          "validation_loss": 0.00043821035069413483,
          "train_metric": 5.111036080052145e-05,
          "validation_metric": 4.413494389154948e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.499980611000865
        },
        {
          "epoch": 1559,
          "updates": 1559,
          "samples_seen": 1596416,
          "online_loss": 0.0005074684740975499,
          "train_loss": 0.0004853799764532596,
          "validation_loss": 0.00038258451968431473,
          "train_metric": 4.888569310423918e-05,
          "validation_metric": 3.853251473628916e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.504819023999517
        },
        {
          "epoch": 1560,
          "updates": 1560,
          "samples_seen": 1597440,
          "online_loss": 0.00048537994734942913,
          "train_loss": 0.00047157585504464805,
          "validation_loss": 0.000389284425182268,
          "train_metric": 4.7495399485342205e-05,
          "validation_metric": 3.9207305235322565e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.50934962000065
        },
        {
          "epoch": 1561,
          "updates": 1561,
          "samples_seen": 1598464,
          "online_loss": 0.00047157585504464805,
          "train_loss": 0.0004732075904030353,
          "validation_loss": 0.00039466426824219525,
          "train_metric": 4.7659730626037344e-05,
          "validation_metric": 3.9749145798850805e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.513779086000795
        },
        {
          "epoch": 1562,
          "updates": 1562,
          "samples_seen": 1599488,
          "online_loss": 0.0004732075904030353,
          "train_loss": 0.0004844364302698523,
          "validation_loss": 0.0003828491608146578,
          "train_metric": 4.879066182184033e-05,
          "validation_metric": 3.855916656902991e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.518135341000743
        },
        {
          "epoch": 1563,
          "updates": 1563,
          "samples_seen": 1600512,
          "online_loss": 0.0004844364011660218,
          "train_loss": 0.0004930243012495339,
          "validation_loss": 0.0004239623376633972,
          "train_metric": 4.965560219716281e-05,
          "validation_metric": 4.269993587513454e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.522448521000115
        },
        {
          "epoch": 1564,
          "updates": 1564,
          "samples_seen": 1601536,
          "online_loss": 0.000493024243041873,
          "train_loss": 0.000493287923745811,
          "validation_loss": 0.0003860858560074121,
          "train_metric": 4.968215216649696e-05,
          "validation_metric": 3.8885158573975787e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.527007589000277
        },
        {
          "epoch": 1565,
          "updates": 1565,
          "samples_seen": 1602560,
          "online_loss": 0.0004932878655381501,
          "train_loss": 0.00048360085929743946,
          "validation_loss": 0.00041149338358081877,
          "train_metric": 4.870650445809588e-05,
          "validation_metric": 4.1444101952947676e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.531471648999286
        },
        {
          "epoch": 1566,
          "updates": 1566,
          "samples_seen": 1603584,
          "online_loss": 0.00048360085929743946,
          "train_loss": 0.00047255365643650293,
          "validation_loss": 0.0003779536345973611,
          "train_metric": 4.7593868657713756e-05,
          "validation_metric": 3.806610402534716e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.536297922000813
        },
        {
          "epoch": 1567,
          "updates": 1567,
          "samples_seen": 1604608,
          "online_loss": 0.0004725536273326725,
          "train_loss": 0.0004660185950342566,
          "validation_loss": 0.0003845491446554661,
          "train_metric": 4.693568916991353e-05,
          "validation_metric": 3.873038076562807e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.540802309000355
        },
        {
          "epoch": 1568,
          "updates": 1568,
          "samples_seen": 1605632,
          "online_loss": 0.00046601862413808703,
          "train_loss": 0.0004666982276830822,
          "validation_loss": 0.00038833467988297343,
          "train_metric": 4.700414137914777e-05,
          "validation_metric": 3.911164458259009e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.545404561000396
        },
        {
          "epoch": 1569,
          "updates": 1569,
          "samples_seen": 1606656,
          "online_loss": 0.00046669828589074314,
          "train_loss": 0.0004716552793979645,
          "validation_loss": 0.0003768737078644335,
          "train_metric": 4.750338848680258e-05,
          "validation_metric": 3.795733937295154e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.550348875000054
        },
        {
          "epoch": 1570,
          "updates": 1570,
          "samples_seen": 1607680,
          "online_loss": 0.00047165530850179493,
          "train_loss": 0.0004755709378514439,
          "validation_loss": 0.0004027692775707692,
          "train_metric": 4.789776721736416e-05,
          "validation_metric": 4.056544275954366e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.554734986999392
        },
        {
          "epoch": 1571,
          "updates": 1571,
          "samples_seen": 1608704,
          "online_loss": 0.00047557096695527434,
          "train_loss": 0.000475820415886119,
          "validation_loss": 0.0003761108382605016,
          "train_metric": 4.7922894736984745e-05,
          "validation_metric": 3.788050526054576e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.559840149999218
        },
        {
          "epoch": 1572,
          "updates": 1572,
          "samples_seen": 1609728,
          "online_loss": 0.0004758204740937799,
          "train_loss": 0.0004715419199783355,
          "validation_loss": 0.000397261930629611,
          "train_metric": 4.749196887132712e-05,
          "validation_metric": 4.0010767406784e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.564405267999973
        },
        {
          "epoch": 1573,
          "updates": 1573,
          "samples_seen": 1610752,
          "online_loss": 0.00047154189087450504,
          "train_loss": 0.00046596149331890047,
          "validation_loss": 0.00037504339707084,
          "train_metric": 4.6929930249461904e-05,
          "validation_metric": 3.77730029867962e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.569107904000703
        },
        {
          "epoch": 1574,
          "updates": 1574,
          "samples_seen": 1611776,
          "online_loss": 0.00046596149331890047,
          "train_loss": 0.0004616275255102664,
          "validation_loss": 0.0003828479675576091,
          "train_metric": 4.6493438276229426e-05,
          "validation_metric": 3.855904651572928e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.573411306000708
        },
        {
          "epoch": 1575,
          "updates": 1575,
          "samples_seen": 1612800,
          "online_loss": 0.0004616275546140969,
          "train_loss": 0.00046022998867556453,
          "validation_loss": 0.00037957183667458594,
          "train_metric": 4.635268123820424e-05,
          "validation_metric": 3.822909275186248e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.57785211800001
        },
        {
          "epoch": 1576,
          "updates": 1576,
          "samples_seen": 1613824,
          "online_loss": 0.000460230017779395,
          "train_loss": 0.0004613582859747112,
          "validation_loss": 0.00037377397529780865,
          "train_metric": 4.646631714422256e-05,
          "validation_metric": 3.764514985959977e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.582682878000924
        },
        {
          "epoch": 1577,
          "updates": 1577,
          "samples_seen": 1614848,
          "online_loss": 0.0004613583441823721,
          "train_loss": 0.0004632353666238487,
          "validation_loss": 0.00038746357313357294,
          "train_metric": 4.6655368350911885e-05,
          "validation_metric": 3.9023911085678264e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.587158050999278
        },
        {
          "epoch": 1578,
          "updates": 1578,
          "samples_seen": 1615872,
          "online_loss": 0.00046323530841618776,
          "train_loss": 0.00046420301077887416,
          "validation_loss": 0.0003720985259860754,
          "train_metric": 4.675281888921745e-05,
          "validation_metric": 3.7476394936675206e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.59217928900034
        },
        {
          "epoch": 1579,
          "updates": 1579,
          "samples_seen": 1616896,
          "online_loss": 0.0004642029816750437,
          "train_loss": 0.00046321385889314115,
          "validation_loss": 0.00038961050449870527,
          "train_metric": 4.6653203753521666e-05,
          "validation_metric": 3.924013799405657e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.59682024200083
        },
        {
          "epoch": 1580,
          "updates": 1580,
          "samples_seen": 1617920,
          "online_loss": 0.00046321385889314115,
          "train_loss": 0.0004608819726854563,
          "validation_loss": 0.0003722720721270889,
          "train_metric": 4.6418346755672246e-05,
          "validation_metric": 3.749387542484328e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.601171689000694
        },
        {
          "epoch": 1581,
          "updates": 1581,
          "samples_seen": 1618944,
          "online_loss": 0.0004608819435816258,
          "train_loss": 0.00045799213694408536,
          "validation_loss": 0.0003822874859906733,
          "train_metric": 4.612729389918968e-05,
          "validation_metric": 3.850259236060083e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.605440583000018
        },
        {
          "epoch": 1582,
          "updates": 1582,
          "samples_seen": 1619968,
          "online_loss": 0.00045799213694408536,
          "train_loss": 0.0004557404899969697,
          "validation_loss": 0.00037241136305965483,
          "train_metric": 4.590051685227081e-05,
          "validation_metric": 3.7507907109102234e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.678638117000446
        },
        {
          "epoch": 1583,
          "updates": 1583,
          "samples_seen": 1620992,
          "online_loss": 0.0004557404899969697,
          "train_loss": 0.0004546426935121417,
          "validation_loss": 0.00037314899964258075,
          "train_metric": 4.578994048642926e-05,
          "validation_metric": 3.758219463634305e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.68296364699927
        },
        {
          "epoch": 1584,
          "updates": 1584,
          "samples_seen": 1622016,
          "online_loss": 0.00045464272261597216,
          "train_loss": 0.00045454956125468016,
          "validation_loss": 0.00037627361598424613,
          "train_metric": 4.578056905302219e-05,
          "validation_metric": 3.7896897993050516e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.68722109600094
        },
        {
          "epoch": 1585,
          "updates": 1585,
          "samples_seen": 1623040,
          "online_loss": 0.0004545498522929847,
          "train_loss": 0.00045493332436308265,
          "validation_loss": 0.00036932367947883904,
          "train_metric": 4.581922257784754e-05,
          "validation_metric": 3.719692904269323e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.692294661999767
        },
        {
          "epoch": 1586,
          "updates": 1586,
          "samples_seen": 1624064,
          "online_loss": 0.00045493326615542173,
          "train_loss": 0.00045513815712183714,
          "validation_loss": 0.00037989148404449224,
          "train_metric": 4.583984991768375e-05,
          "validation_metric": 3.826128158834763e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.69698996599982
        },
        {
          "epoch": 1587,
          "updates": 1587,
          "samples_seen": 1625088,
          "online_loss": 0.00045513815712183714,
          "train_loss": 0.0004548075667116791,
          "validation_loss": 0.0003675063489936292,
          "train_metric": 4.5806558773620054e-05,
          "validation_metric": 3.701389505295083e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.701879510999788
        },
        {
          "epoch": 1588,
          "updates": 1588,
          "samples_seen": 1626112,
          "online_loss": 0.00045480753760784864,
          "train_loss": 0.00045377996866591275,
          "validation_loss": 0.00037794289528392255,
          "train_metric": 4.5703058276558295e-05,
          "validation_metric": 3.806502718362026e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.706422086999737
        },
        {
          "epoch": 1589,
          "updates": 1589,
          "samples_seen": 1627136,
          "online_loss": 0.0004537799395620823,
          "train_loss": 0.0004523546085692942,
          "validation_loss": 0.0003660216461867094,
          "train_metric": 4.555950727080926e-05,
          "validation_metric": 3.6864355934085324e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.711366743000326
        },
        {
          "epoch": 1590,
          "updates": 1590,
          "samples_seen": 1628160,
          "online_loss": 0.00045235463767312467,
          "train_loss": 0.000450806604931131,
          "validation_loss": 0.000373163289623335,
          "train_metric": 4.540358713711612e-05,
          "validation_metric": 3.758364255190827e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.7157430139996
        },
        {
          "epoch": 1591,
          "updates": 1591,
          "samples_seen": 1629184,
          "online_loss": 0.00045080657582730055,
          "train_loss": 0.0004494788299780339,
          "validation_loss": 0.00036691667628474534,
          "train_metric": 4.526986231212504e-05,
          "validation_metric": 3.695450141094625e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.72007191600096
        },
        {
          "epoch": 1592,
          "updates": 1592,
          "samples_seen": 1630208,
          "online_loss": 0.00044947880087420344,
          "train_loss": 0.00044852265273220837,
          "validation_loss": 0.000368516833987087,
          "train_metric": 4.517356137512252e-05,
          "validation_metric": 3.711566751007922e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.72447509400081
        },
        {
          "epoch": 1593,
          "updates": 1593,
          "samples_seen": 1631232,
          "online_loss": 0.00044852268183603883,
          "train_loss": 0.0004479378112591803,
          "validation_loss": 0.0003687800490297377,
          "train_metric": 4.5114658860256895e-05,
          "validation_metric": 3.71421774616465e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.728982721999273
        },
        {
          "epoch": 1594,
          "updates": 1594,
          "samples_seen": 1632256,
          "online_loss": 0.0004479378112591803,
          "train_loss": 0.0004475984605960548,
          "validation_loss": 0.00036485923919826746,
          "train_metric": 4.508047641138546e-05,
          "validation_metric": 3.674728577607311e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.733860788999664
        },
        {
          "epoch": 1595,
          "updates": 1595,
          "samples_seen": 1633280,
          "online_loss": 0.00044759848969988525,
          "train_loss": 0.0004473196167964488,
          "validation_loss": 0.00037032278487458825,
          "train_metric": 4.505239849095233e-05,
          "validation_metric": 3.729755189851858e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.73818678799944
        },
        {
          "epoch": 1596,
          "updates": 1596,
          "samples_seen": 1634304,
          "online_loss": 0.00044731958769261837,
          "train_loss": 0.00044695736141875386,
          "validation_loss": 0.0003631155414041132,
          "train_metric": 4.501589501160197e-05,
          "validation_metric": 3.657167326309718e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.743215349999446
        },
        {
          "epoch": 1597,
          "updates": 1597,
          "samples_seen": 1635328,
          "online_loss": 0.0004469572741072625,
          "train_loss": 0.00044641163549385965,
          "validation_loss": 0.0003706836432684213,
          "train_metric": 4.4960946979699656e-05,
          "validation_metric": 3.7333906220737845e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.74763014599921
        },
        {
          "epoch": 1598,
          "updates": 1598,
          "samples_seen": 1636352,
          "online_loss": 0.00044641163549385965,
          "train_loss": 0.00044569093734025955,
          "validation_loss": 0.00036208186065778136,
          "train_metric": 4.4888358388561755e-05,
          "validation_metric": 3.6467557947617024e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.752343881000343
        },
        {
          "epoch": 1599,
          "updates": 1599,
          "samples_seen": 1637376,
          "online_loss": 0.00044569099554792047,
          "train_loss": 0.0004448083636816591,
          "validation_loss": 0.0003687065909616649,
          "train_metric": 4.479946437641047e-05,
          "validation_metric": 3.713477781275287e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.75670079200063
        },
        {
          "epoch": 1600,
          "updates": 1600,
          "samples_seen": 1638400,
          "online_loss": 0.00044480833457782865,
          "train_loss": 0.0004438625182956457,
          "validation_loss": 0.000361256388714537,
          "train_metric": 4.470421117730439e-05,
          "validation_metric": 3.638441921793856e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 17.761587385000894
        }
      ],
      "best": {
        "epoch": 1600,
        "updates": 1600
      },
      "best_validation_metric": 3.638441921793856e-05,
      "config": {
        "label": "Full batch",
        "batch": 1024,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mse",
        "outliers": false,
        "max_updates": 1600
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.006031949868652637
    },
    "13e54c648cc2_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.27039185469038785,
          "train_loss": 0.10831719636917114,
          "validation_loss": 0.09477120637893677,
          "train_metric": 0.010909311473369598,
          "validation_metric": 0.009545008651912212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.020402255999215413
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.08604311244562268,
          "train_loss": 0.08449298143386841,
          "validation_loss": 0.08675695210695267,
          "train_metric": 0.008509824052453041,
          "validation_metric": 0.008737842552363873,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18000449399914942
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.060514786979183555,
          "train_loss": 0.047580424696207047,
          "validation_loss": 0.03620985895395279,
          "train_metric": 0.00479212636128068,
          "validation_metric": 0.0036469243932515383,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.20004843799870287
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.03882622008677572,
          "train_loss": 0.03523484990000725,
          "validation_loss": 0.036904942244291306,
          "train_metric": 0.003548725275322795,
          "validation_metric": 0.0037169307470321655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2182747049992031
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.03007221716688946,
          "train_loss": 0.02236410602927208,
          "validation_loss": 0.01739375852048397,
          "train_metric": 0.0022524306550621986,
          "validation_metric": 0.0017518356908112764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23565095699996164
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.03099765721708536,
          "train_loss": 0.024014728143811226,
          "validation_loss": 0.01601189374923706,
          "train_metric": 0.00241867545992136,
          "validation_metric": 0.0016126594273373485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2530987699992693
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.027296714659314603,
          "train_loss": 0.02134745381772518,
          "validation_loss": 0.012599899433553219,
          "train_metric": 0.0021500373259186745,
          "validation_metric": 0.0012690158328041434,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2711201419988356
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.021192819200223312,
          "train_loss": 0.02039089985191822,
          "validation_loss": 0.012964699417352676,
          "train_metric": 0.0020536966621875763,
          "validation_metric": 0.0013057570904493332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3008773459987424
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.030007397668669,
          "train_loss": 0.03706315532326698,
          "validation_loss": 0.03235498070716858,
          "train_metric": 0.0037328654434531927,
          "validation_metric": 0.0032586753368377686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3304754109994974
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.030091354914475232,
          "train_loss": 0.02330632321536541,
          "validation_loss": 0.016952436417341232,
          "train_metric": 0.0023473273031413555,
          "validation_metric": 0.0017073872731998563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.423282133999237
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.03209615338710137,
          "train_loss": 0.024077557027339935,
          "validation_loss": 0.0241977758705616,
          "train_metric": 0.0024250035639852285,
          "validation_metric": 0.002437111223116517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5855980679989443
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.024692562234122306,
          "train_loss": 0.019976388663053513,
          "validation_loss": 0.015801269561052322,
          "train_metric": 0.0020119487307965755,
          "validation_metric": 0.0015914459945634007,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6644102439986455
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.016155207675183192,
          "train_loss": 0.011977332644164562,
          "validation_loss": 0.009538279846310616,
          "train_metric": 0.0012063130270689726,
          "validation_metric": 0.0009606605744920671,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7478512629986653
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.014945351402275264,
          "train_loss": 0.009783338755369186,
          "validation_loss": 0.007489394396543503,
          "train_metric": 0.0009853420779109001,
          "validation_metric": 0.0007543043466284871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8268141499993362
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.02490110631333664,
          "train_loss": 0.039069853723049164,
          "validation_loss": 0.03355080634355545,
          "train_metric": 0.003934972919523716,
          "validation_metric": 0.003379114205017686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9027383359989472
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.02574498814647086,
          "train_loss": 0.027779800817370415,
          "validation_loss": 0.023996975272893906,
          "train_metric": 0.002797879744321108,
          "validation_metric": 0.002416887553408742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9770610429986846
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.014243885641917586,
          "train_loss": 0.011660939082503319,
          "validation_loss": 0.007793628145009279,
          "train_metric": 0.001174447126686573,
          "validation_metric": 0.0007849456742405891,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0493032049998874
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.014597671921364963,
          "train_loss": 0.012175031937658787,
          "validation_loss": 0.009340544231235981,
          "train_metric": 0.0012262248201295733,
          "validation_metric": 0.0009407453471794724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1248911109996698
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.011154008767334744,
          "train_loss": 0.00809113122522831,
          "validation_loss": 0.006103809457272291,
          "train_metric": 0.0008149091736413538,
          "validation_metric": 0.0006147532840259373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2030778999996983
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.01245289073267486,
          "train_loss": 0.01601196639239788,
          "validation_loss": 0.015226091258227825,
          "train_metric": 0.0016126667615026236,
          "validation_metric": 0.0015335161006078124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2750265039994702
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.014955991646274924,
          "train_loss": 0.009401122108101845,
          "validation_loss": 0.006404159590601921,
          "train_metric": 0.0009468466741964221,
          "validation_metric": 0.0006450035143643618,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3471830599992245
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.016770219488535076,
          "train_loss": 0.022339433431625366,
          "validation_loss": 0.02121872827410698,
          "train_metric": 0.0022499458864331245,
          "validation_metric": 0.0021370723843574524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4308458269988478
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.018702914705500007,
          "train_loss": 0.008285162039101124,
          "validation_loss": 0.006565423216670752,
          "train_metric": 0.0008344511734321713,
          "validation_metric": 0.0006612454308196902,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.511602534999838
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.010724682273576036,
          "train_loss": 0.006374263670295477,
          "validation_loss": 0.004598214756697416,
          "train_metric": 0.0006419925484806299,
          "validation_metric": 0.00046311537153087556,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.595984633999251
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.006870915822219104,
          "train_loss": 0.007550710812211037,
          "validation_loss": 0.005002892576158047,
          "train_metric": 0.0007604800048284233,
          "validation_metric": 0.000503873045090586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.673311966998881
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.006254155043279752,
          "train_loss": 0.00838354229927063,
          "validation_loss": 0.006725983694195747,
          "train_metric": 0.0008443596307188272,
          "validation_metric": 0.000677416508551687,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7473008739998477
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.01153010233247187,
          "train_loss": 0.012380441650748253,
          "validation_loss": 0.00788271427154541,
          "train_metric": 0.0012469131033867598,
          "validation_metric": 0.0007939182105474174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8232600009996531
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.01210236651240848,
          "train_loss": 0.012877147644758224,
          "validation_loss": 0.010447257198393345,
          "train_metric": 0.0012969390954822302,
          "validation_metric": 0.001052209408953786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8963251019995369
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.00964202600880526,
          "train_loss": 0.007029639557003975,
          "validation_loss": 0.006290952675044537,
          "train_metric": 0.0007079995120875537,
          "validation_metric": 0.0006336017395369709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.970940884999436
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.005755072925239801,
          "train_loss": 0.007404249627143145,
          "validation_loss": 0.006765091326087713,
          "train_metric": 0.0007457288447767496,
          "validation_metric": 0.0006813552463427186,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.046344565000254
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.0072822551446734,
          "train_loss": 0.0044168513268232346,
          "validation_loss": 0.004020411055535078,
          "train_metric": 0.00044484907994046807,
          "validation_metric": 0.0004049211274832487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1228656849998515
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.0059974856703775,
          "train_loss": 0.003360067494213581,
          "validation_loss": 0.0027285295072942972,
          "train_metric": 0.00033841366530396044,
          "validation_metric": 0.0002748075348790735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2014742179999303
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.0036758054775418714,
          "train_loss": 0.007153711281716824,
          "validation_loss": 0.004597029648721218,
          "train_metric": 0.0007204957073554397,
          "validation_metric": 0.00046299598761834204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2777507819992024
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.014916526299202815,
          "train_loss": 0.019613247364759445,
          "validation_loss": 0.017338203266263008,
          "train_metric": 0.0019753742963075638,
          "validation_metric": 0.0017462404211983085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3508583469993027
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.014810927532380447,
          "train_loss": 0.006818210706114769,
          "validation_loss": 0.004920503124594688,
          "train_metric": 0.0006867052288725972,
          "validation_metric": 0.000495575019158423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.426932747999672
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.0063115113152889535,
          "train_loss": 0.0066928197629749775,
          "validation_loss": 0.004379418212920427,
          "train_metric": 0.0006740762619301677,
          "validation_metric": 0.0004410789697431028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5009367459988425
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.00683075851702597,
          "train_loss": 0.003941405098885298,
          "validation_loss": 0.002759720664471388,
          "train_metric": 0.0003969638200942427,
          "validation_metric": 0.0002779489441309124,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5761065959995904
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.010884807867114432,
          "train_loss": 0.0069356816820800304,
          "validation_loss": 0.005292749498039484,
          "train_metric": 0.0006985364598222077,
          "validation_metric": 0.0005330663407221437,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.832592986000236
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.014509179018205032,
          "train_loss": 0.013422197662293911,
          "validation_loss": 0.011587749235332012,
          "train_metric": 0.0013518345076590776,
          "validation_metric": 0.0011670757085084915,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.90917465199891
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.012970162235433236,
          "train_loss": 0.005950959399342537,
          "validation_loss": 0.0036894117947667837,
          "train_metric": 0.0005993589293211699,
          "validation_metric": 0.00037158402847126126,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9831429240002763
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.008675849130668212,
          "train_loss": 0.011716296896338463,
          "validation_loss": 0.011606852523982525,
          "train_metric": 0.0011800226056948304,
          "validation_metric": 0.0011689997045323253,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0592342489999282
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.008912016564863734,
          "train_loss": 0.011386807076632977,
          "validation_loss": 0.011526749469339848,
          "train_metric": 0.001146837486885488,
          "validation_metric": 0.0011609321227297187,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.132642685999599
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.008542801624571439,
          "train_loss": 0.012400511652231216,
          "validation_loss": 0.00755611527711153,
          "train_metric": 0.0012489340733736753,
          "validation_metric": 0.0007610243046656251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.207019342999047
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.007273746850842144,
          "train_loss": 0.00538557069376111,
          "validation_loss": 0.004754276014864445,
          "train_metric": 0.0005424149567261338,
          "validation_metric": 0.000478833302622661,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.2215137369985314
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.006097309320466593,
          "train_loss": 0.005181498825550079,
          "validation_loss": 0.004026203881949186,
          "train_metric": 0.0005218615406192839,
          "validation_metric": 0.0004055044846609235,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.235469748999094
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.0031967391187208705,
          "train_loss": 0.0027258347254246473,
          "validation_loss": 0.0018735986668616533,
          "train_metric": 0.00027453608345240355,
          "validation_metric": 0.0001887019898276776,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.249745675999293
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.00351145256718155,
          "train_loss": 0.0025766517501324415,
          "validation_loss": 0.0020014429464936256,
          "train_metric": 0.0002595109399408102,
          "validation_metric": 0.0002015780337387696,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.2635991099996318
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.0044786485232179984,
          "train_loss": 0.006316338200122118,
          "validation_loss": 0.005975830834358931,
          "train_metric": 0.0006361583946272731,
          "validation_metric": 0.0006018637795932591,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.277507554999829
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.006223092568689026,
          "train_loss": 0.002850273856893182,
          "validation_loss": 0.002155404770746827,
          "train_metric": 0.00028706915327347815,
          "validation_metric": 0.00021708446729462594,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.291262665999966
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.0034540895067038946,
          "train_loss": 0.004735914058983326,
          "validation_loss": 0.0021765073761343956,
          "train_metric": 0.0004769838706124574,
          "validation_metric": 0.00021920989092905074,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.304831560999446
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.005932461019256152,
          "train_loss": 0.006641009822487831,
          "validation_loss": 0.004162976518273354,
          "train_metric": 0.000668858177959919,
          "validation_metric": 0.0004192797641735524,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.3182833249993564
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0035993847050121985,
          "train_loss": 0.00285546132363379,
          "validation_loss": 0.002161769662052393,
          "train_metric": 0.000287591596134007,
          "validation_metric": 0.00021772552281618118,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.3315823289995024
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.002213327603385551,
          "train_loss": 0.0022293992806226015,
          "validation_loss": 0.0015023990999907255,
          "train_metric": 0.00022453693964052945,
          "validation_metric": 0.0001513161405455321,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.346242423000149
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.0020752526907017455,
          "train_loss": 0.005454767029732466,
          "validation_loss": 0.004867353476583958,
          "train_metric": 0.0005493841017596424,
          "validation_metric": 0.0004902220098301768,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.3600670599989826
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.007380514005490113,
          "train_loss": 0.012254022061824799,
          "validation_loss": 0.0089866379275918,
          "train_metric": 0.0012341801775619388,
          "validation_metric": 0.0009051012457348406,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.373983079000027
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.0076319071267789695,
          "train_loss": 0.0056738462299108505,
          "validation_loss": 0.0038891916628926992,
          "train_metric": 0.0005714489379897714,
          "validation_metric": 0.0003917051653843373,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.387909742999909
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.0042775771507876925,
          "train_loss": 0.0031999628990888596,
          "validation_loss": 0.0025880630128085613,
          "train_metric": 0.00032228854252025485,
          "validation_metric": 0.0002606602502055466,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.4015591819988913
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.002596281836304115,
          "train_loss": 0.0022443158086389303,
          "validation_loss": 0.0017601564759388566,
          "train_metric": 0.00022603927936870605,
          "validation_metric": 0.0001772765099303797,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.4151570449994324
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.002543121125199832,
          "train_loss": 0.0022101118229329586,
          "validation_loss": 0.0017761127091944218,
          "train_metric": 0.00022259434626903385,
          "validation_metric": 0.0001788835652405396,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.4295639659994777
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.0023379655613098294,
          "train_loss": 0.0033856737427413464,
          "validation_loss": 0.003225049702450633,
          "train_metric": 0.00034099267213605344,
          "validation_metric": 0.00032481522066518664,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.4434532089999266
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.0029188027328928,
          "train_loss": 0.002116111107170582,
          "validation_loss": 0.0016223047859966755,
          "train_metric": 0.00021312695753294975,
          "validation_metric": 0.0001633925858186558,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.4569909879992338
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.0021843492941115983,
          "train_loss": 0.00338875874876976,
          "validation_loss": 0.00324048544280231,
          "train_metric": 0.00034130338463000953,
          "validation_metric": 0.00032636983087286353,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.470564708999518
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.00479234377053217,
          "train_loss": 0.00338185322470963,
          "validation_loss": 0.002430734457448125,
          "train_metric": 0.00034060789039358497,
          "validation_metric": 0.00024481466971337795,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.4841328019992943
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.006234061831491999,
          "train_loss": 0.0028306427411735058,
          "validation_loss": 0.0019649851601570845,
          "train_metric": 0.00028509198455139995,
          "validation_metric": 0.00019790611986536533,
          "lr_first": 0.0015,
          "lr_last": 0.0015,
          "seconds": 3.4972352479999245
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.002711169858230278,
          "train_loss": 0.002257874934002757,
          "validation_loss": 0.0014556334353983402,
          "train_metric": 0.00022740490385331213,
          "validation_metric": 0.00014660607848782092,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.511542770000233
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.002314257850230206,
          "train_loss": 0.0020077116787433624,
          "validation_loss": 0.0014190052170306444,
          "train_metric": 0.00020220933947712183,
          "validation_metric": 0.00014291700790636241,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.5263013530002354
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.0023956389486556873,
          "train_loss": 0.0014912657206878066,
          "validation_loss": 0.0010089848656207323,
          "train_metric": 0.00015019481361377984,
          "validation_metric": 0.00010162125545321032,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.5405447049997747
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.0030163879855535924,
          "train_loss": 0.0019829142838716507,
          "validation_loss": 0.0014882433461025357,
          "train_metric": 0.00019971186702605337,
          "validation_metric": 0.0001498903875472024,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.5541900629996235
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.0017976215240196325,
          "train_loss": 0.0020751538686454296,
          "validation_loss": 0.0016681808046996593,
          "train_metric": 0.00020900191157124937,
          "validation_metric": 0.00016801306628622115,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.567604316998768
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.001801738118956564,
          "train_loss": 0.0018705448601394892,
          "validation_loss": 0.001087781973183155,
          "train_metric": 0.00018839442054741085,
          "validation_metric": 0.00010955744073726237,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.5810823809988506
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.0018823555074050091,
          "train_loss": 0.0014083439018577337,
          "validation_loss": 0.0009411974460817873,
          "train_metric": 0.0001418432657374069,
          "validation_metric": 9.479395521339029e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.595845620999171
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.0019586169182730373,
          "train_loss": 0.0012949926313012838,
          "validation_loss": 0.000843672314658761,
          "train_metric": 0.00013042690989095718,
          "validation_metric": 8.49715870572254e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.61045852999996
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.0022908596947672777,
          "train_loss": 0.001621784525923431,
          "validation_loss": 0.0012018565321341157,
          "train_metric": 0.0001633402134757489,
          "validation_metric": 0.000121046592539642,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.623953102000087
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.0029540694995375816,
          "train_loss": 0.0014911520993337035,
          "validation_loss": 0.0009854196105152369,
          "train_metric": 0.00015018336125649512,
          "validation_metric": 9.924786718329415e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.713551680999444
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.0025843462281045504,
          "train_loss": 0.0022231494076550007,
          "validation_loss": 0.0020098425447940826,
          "train_metric": 0.000223907467443496,
          "validation_metric": 0.0002024240093305707,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.7275578759999917
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.004214374152070377,
          "train_loss": 0.0025995387695729733,
          "validation_loss": 0.00207234313711524,
          "train_metric": 0.0002618160506244749,
          "validation_metric": 0.00020871884771622717,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.7410156360001565
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.0020741753542097285,
          "train_loss": 0.0014953636564314365,
          "validation_loss": 0.001063522300682962,
          "train_metric": 0.00015060753503348678,
          "validation_metric": 0.00010711405775509775,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.7546040380002523
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.0015665074715798255,
          "train_loss": 0.002059738850221038,
          "validation_loss": 0.001982865622267127,
          "train_metric": 0.00020744935318361968,
          "validation_metric": 0.0001997069630306214,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.7679058619996795
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.001285772701521637,
          "train_loss": 0.0010984080145135522,
          "validation_loss": 0.0007492186850868165,
          "train_metric": 0.00011062763223890215,
          "validation_metric": 7.545855623902753e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.782395070998973
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.001317807857049047,
          "train_loss": 0.0011548821348696947,
          "validation_loss": 0.0009371885098516941,
          "train_metric": 0.00011631549568846822,
          "validation_metric": 9.439019777346402e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.796001044998775
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.0016294590059260372,
          "train_loss": 0.0029947711154818535,
          "validation_loss": 0.003115281695500016,
          "train_metric": 0.00030162237817421556,
          "validation_metric": 0.0003137597523164004,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.8095421219986747
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.0019673897932079853,
          "train_loss": 0.003256085328757763,
          "validation_loss": 0.0016720914281904697,
          "train_metric": 0.00032794103026390076,
          "validation_metric": 0.00016840694297570735,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.8227766009986226
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.0034911314214696176,
          "train_loss": 0.004939116537570953,
          "validation_loss": 0.004539459943771362,
          "train_metric": 0.000497449713293463,
          "validation_metric": 0.00045719777699559927,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.8365908989999298
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0024556624666729476,
          "train_loss": 0.0042329332791268826,
          "validation_loss": 0.002737905830144882,
          "train_metric": 0.0004263255978003144,
          "validation_metric": 0.00027575186686590314,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.8500882549997186
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.0033729345741448924,
          "train_loss": 0.0019649409223347902,
          "validation_loss": 0.0014314119471237063,
          "train_metric": 0.00019790166697930545,
          "validation_metric": 0.00014416658086702228,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.8636725079995813
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.0016562887576583307,
          "train_loss": 0.001752770389430225,
          "validation_loss": 0.0011970781488344073,
          "train_metric": 0.00017653260147199035,
          "validation_metric": 0.00012056532432325184,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.878004807000252
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.0013243104513094295,
          "train_loss": 0.001182821928523481,
          "validation_loss": 0.00106570846401155,
          "train_metric": 0.0001191294941236265,
          "validation_metric": 0.00010733426461229101,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.891307402998791
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.001274044578167377,
          "train_loss": 0.0019282102584838867,
          "validation_loss": 0.0009022144367918372,
          "train_metric": 0.00019420227908995003,
          "validation_metric": 9.086773206945509e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.9049921149999136
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.001534095075840014,
          "train_loss": 0.001416330342181027,
          "validation_loss": 0.0007926099351607263,
          "train_metric": 0.00014264760829973966,
          "validation_metric": 7.982877286849543e-05,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.9186352669985354
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.0012601830203493591,
          "train_loss": 0.0013599090743809938,
          "validation_loss": 0.0012348615564405918,
          "train_metric": 0.00013696507085114717,
          "validation_metric": 0.00012437073746696115,
          "lr_first": 0.00075,
          "lr_last": 0.00075,
          "seconds": 3.9320041579994722
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.0017020137620420428,
          "train_loss": 0.0008792192675173283,
          "validation_loss": 0.0006112584378570318,
          "train_metric": 8.85517438291572e-05,
          "validation_metric": 6.156371819088235e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.946205453999937
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.0014439222322835121,
          "train_loss": 0.0014128442853689194,
          "validation_loss": 0.0007650675834156573,
          "train_metric": 0.00014229652879294008,
          "validation_metric": 7.705479947617278e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.95917339999869
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.0022793409843870904,
          "train_loss": 0.001407533884048462,
          "validation_loss": 0.0015574261778965592,
          "train_metric": 0.00014176165859680623,
          "validation_metric": 0.00015685826656408608,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.9724804289999156
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.0010564502736087888,
          "train_loss": 0.0009462501620873809,
          "validation_loss": 0.0008683642372488976,
          "train_metric": 9.530285751679912e-05,
          "validation_metric": 8.745846571400762e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.9856306040001073
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.0009260230417567072,
          "train_loss": 0.0008279129979200661,
          "validation_loss": 0.0006555364816449583,
          "train_metric": 8.338437328347936e-05,
          "validation_metric": 6.602323264814913e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 3.999452862999533
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.0009793487861315953,
          "train_loss": 0.0008683389169164002,
          "validation_loss": 0.0005781560321338475,
          "train_metric": 8.745592640480027e-05,
          "validation_metric": 5.822976163472049e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.013589725000202
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.0009419480466021923,
          "train_loss": 0.0010685602901503444,
          "validation_loss": 0.0010676626116037369,
          "train_metric": 0.0001076214830391109,
          "validation_metric": 0.00010753106471383944,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.0265034929998365
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.000977919402430416,
          "train_loss": 0.0011735205771401525,
          "validation_loss": 0.0006951753166504204,
          "train_metric": 0.00011819269275292754,
          "validation_metric": 7.001551421126351e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.039136264000263
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.0013109370975143975,
          "train_loss": 0.0010005509248003364,
          "validation_loss": 0.0008657614816911519,
          "train_metric": 0.00010077182378154248,
          "validation_metric": 8.719632751308382e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.0520899269995425
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.00124524182137975,
          "train_loss": 0.002730892738327384,
          "validation_loss": 0.0013844100758433342,
          "train_metric": 0.00027504548779688776,
          "validation_metric": 0.0001394327264279127,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.065012073999242
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.0015836811617191415,
          "train_loss": 0.0008212419343180954,
          "validation_loss": 0.0005816194461658597,
          "train_metric": 8.27124749775976e-05,
          "validation_metric": 5.8578581956680864e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.0779066429986415
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0012844397970184218,
          "train_loss": 0.0023291120305657387,
          "validation_loss": 0.0020122637506574392,
          "train_metric": 0.00023457960924133658,
          "validation_metric": 0.00020266782667022198,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.090750550998564
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.0010753828028100543,
          "train_loss": 0.0008706925436854362,
          "validation_loss": 0.0005170069634914398,
          "train_metric": 8.769296982791275e-05,
          "validation_metric": 5.2071045502088964e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.106630975998996
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0009256051052943803,
          "train_loss": 0.0008441637037321925,
          "validation_loss": 0.0007620723336003721,
          "train_metric": 8.502107812091708e-05,
          "validation_metric": 7.675313827348873e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.12110971599941
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.0010388831251475494,
          "train_loss": 0.0008578147389926016,
          "validation_loss": 0.00048559371498413384,
          "train_metric": 8.639595762360841e-05,
          "validation_metric": 4.890721902484074e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.136397800999475
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.0008264107582363067,
          "train_loss": 0.0007968554273247719,
          "validation_loss": 0.0005233577685430646,
          "train_metric": 8.025635906960815e-05,
          "validation_metric": 5.271068221190944e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.150418527999136
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.0008568362973164767,
          "train_loss": 0.0011822390370070934,
          "validation_loss": 0.0006437514093704522,
          "train_metric": 0.00011907076986972243,
          "validation_metric": 6.483628385467455e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.165158054998756
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.0008986302054836415,
          "train_loss": 0.0022852884139865637,
          "validation_loss": 0.0020967568270862103,
          "train_metric": 0.0002301658969372511,
          "validation_metric": 0.00021117765572853386,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.17935733000013
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.0026267729936080286,
          "train_loss": 0.0039892373606562614,
          "validation_loss": 0.002479719230905175,
          "train_metric": 0.0004017814062535763,
          "validation_metric": 0.0002497482346370816,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.193060005000007
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.003942587623896543,
          "train_loss": 0.0017432188615202904,
          "validation_loss": 0.0010347424540668726,
          "train_metric": 0.00017557060346007347,
          "validation_metric": 0.00010421546903671697,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.293397321000157
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.004702469024778111,
          "train_loss": 0.0082984184846282,
          "validation_loss": 0.006229918450117111,
          "train_metric": 0.0008357863407582045,
          "validation_metric": 0.0006274546030908823,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.308583398998962
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.00291396991815418,
          "train_loss": 0.0014876980567350984,
          "validation_loss": 0.0012439556885510683,
          "train_metric": 0.00014983551227487624,
          "validation_metric": 0.000125286664115265,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.322621767998498
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.001392832102283137,
          "train_loss": 0.0009635768947191536,
          "validation_loss": 0.0006139585166238248,
          "train_metric": 9.70479377429001e-05,
          "validation_metric": 6.183564983075485e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.336755466998511
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.0012680575346166734,
          "train_loss": 0.001989844022318721,
          "validation_loss": 0.001435345271602273,
          "train_metric": 0.00020040982053615153,
          "validation_metric": 0.00014456271310336888,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.350646158998643
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.0015530542132182745,
          "train_loss": 0.0009647554252296686,
          "validation_loss": 0.000722775817848742,
          "train_metric": 9.716663771541789e-05,
          "validation_metric": 7.279534474946558e-05,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.364462774999993
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.0019978164818894584,
          "train_loss": 0.00463605672121048,
          "validation_loss": 0.003945913165807724,
          "train_metric": 0.00046692672185599804,
          "validation_metric": 0.0003974179271608591,
          "lr_first": 0.000375,
          "lr_last": 0.000375,
          "seconds": 4.377980322999065
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.0014610846774303354,
          "train_loss": 0.0009420500136911869,
          "validation_loss": 0.0007517232443206012,
          "train_metric": 9.48798333411105e-05,
          "validation_metric": 7.571079913759604e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.39123361899874
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.0009325676692242268,
          "train_loss": 0.0007882557110860944,
          "validation_loss": 0.0005339266499504447,
          "train_metric": 7.939021452330053e-05,
          "validation_metric": 5.3775132982991636e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.404338311998799
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.0008333768091688398,
          "train_loss": 0.0009167770040221512,
          "validation_loss": 0.0005757063627243042,
          "train_metric": 9.23344268812798e-05,
          "validation_metric": 5.798303391202353e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.4179435890000605
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.0008263180498033762,
          "train_loss": 0.0007387551595456898,
          "validation_loss": 0.00048502368736080825,
          "train_metric": 7.440471381414682e-05,
          "validation_metric": 4.8849811719264835e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.432206726998629
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0007633808163518552,
          "train_loss": 0.0008198108407668769,
          "validation_loss": 0.0005455089849419892,
          "train_metric": 8.2568330981303e-05,
          "validation_metric": 5.494166543940082e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.446798352999394
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.0007870137560530566,
          "train_loss": 0.0007119330111891031,
          "validation_loss": 0.0005320522468537092,
          "train_metric": 7.170328171923757e-05,
          "validation_metric": 5.358635098673403e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.460109712999838
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.0007288602228072705,
          "train_loss": 0.0009481186862103641,
          "validation_loss": 0.0006058833096176386,
          "train_metric": 9.549103560857475e-05,
          "validation_metric": 6.102235056459904e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.473629015999904
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.0008220407962653553,
          "train_loss": 0.0007448068354278803,
          "validation_loss": 0.00046490170643664896,
          "train_metric": 7.501422078348696e-05,
          "validation_metric": 4.682319922721945e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.487569478998921
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0007083185791998403,
          "train_loss": 0.0006698018405586481,
          "validation_loss": 0.0005172903765924275,
          "train_metric": 6.745998689439148e-05,
          "validation_metric": 5.20995854458306e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.501378379998641
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.0006846816540928558,
          "train_loss": 0.0007021906785666943,
          "validation_loss": 0.0005928267491981387,
          "train_metric": 7.072206790326163e-05,
          "validation_metric": 5.97073340031784e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.516538202999072
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.0006658678566964227,
          "train_loss": 0.0006763300625607371,
          "validation_loss": 0.0004817787848878652,
          "train_metric": 6.811749335611239e-05,
          "validation_metric": 4.852299753110856e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.530140349999783
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.0006928018428880023,
          "train_loss": 0.0006425674073398113,
          "validation_loss": 0.0004721706500276923,
          "train_metric": 6.471703090937808e-05,
          "validation_metric": 4.7555291530443355e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.543595695999102
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.0007251664119394263,
          "train_loss": 0.0007631265325471759,
          "validation_loss": 0.000480257673189044,
          "train_metric": 7.685932359891012e-05,
          "validation_metric": 4.836979860556312e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.557359197999176
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.0009287359098379966,
          "train_loss": 0.0011667445069178939,
          "validation_loss": 0.0006812980282120407,
          "train_metric": 0.00011751022975658998,
          "validation_metric": 6.861783913336694e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.570648531000188
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0009269606416637544,
          "train_loss": 0.0015462682349607348,
          "validation_loss": 0.0008027695002965629,
          "train_metric": 0.000155734465806745,
          "validation_metric": 8.085199806373566e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.584429777998594
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.0011951181431868463,
          "train_loss": 0.0010011253179982305,
          "validation_loss": 0.0005494765937328339,
          "train_metric": 0.00010082965309266001,
          "validation_metric": 5.53412719455082e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.598330644999805
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.0007853879669710295,
          "train_loss": 0.0006155035807751119,
          "validation_loss": 0.0004805239150300622,
          "train_metric": 6.1991268012207e-05,
          "validation_metric": 4.839660687139258e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.612062485999559
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.0006542095816257643,
          "train_loss": 0.0006089349626563489,
          "validation_loss": 0.0004556811763904989,
          "train_metric": 6.132970156613737e-05,
          "validation_metric": 4.589453965309076e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.625928366000153
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.0007268115205079084,
          "train_loss": 0.0006693254108540714,
          "validation_loss": 0.00043953247950412333,
          "train_metric": 6.741199467796832e-05,
          "validation_metric": 4.4268101191846654e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.64045913299924
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.0008578254291933263,
          "train_loss": 0.000941129052080214,
          "validation_loss": 0.0008579926798120141,
          "train_metric": 9.478706488152966e-05,
          "validation_metric": 8.641387830721214e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.653864639998574
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.0012567935173137812,
          "train_loss": 0.0010376513237133622,
          "validation_loss": 0.0009790732292458415,
          "train_metric": 0.00010450844274600968,
          "validation_metric": 9.860866703093052e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.667661321998821
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.0008445746898360085,
          "train_loss": 0.0014405252877622843,
          "validation_loss": 0.0013544770190492272,
          "train_metric": 0.00014508444292005152,
          "validation_metric": 0.00013641796249430627,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.6816673269986495
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.0011147153491037898,
          "train_loss": 0.001329685328528285,
          "validation_loss": 0.0012244242243468761,
          "train_metric": 0.00013392104301601648,
          "validation_metric": 0.00012331952166277915,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.695185069998843
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.0007203482200566214,
          "train_loss": 0.0006225103861652315,
          "validation_loss": 0.0004898698534816504,
          "train_metric": 6.269696314120665e-05,
          "validation_metric": 4.9337901145918295e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.708802612998625
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.0006286714333327836,
          "train_loss": 0.000614881340879947,
          "validation_loss": 0.00042280639172531664,
          "train_metric": 6.192859291331843e-05,
          "validation_metric": 4.2583509639371186e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.724569103998874
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0006899255786265712,
          "train_loss": 0.0009200612548738718,
          "validation_loss": 0.0009029224747791886,
          "train_metric": 9.266519919037819e-05,
          "validation_metric": 9.09390437300317e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.73821042399868
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.0006370557512127561,
          "train_loss": 0.0006095089484006166,
          "validation_loss": 0.0005585826002061367,
          "train_metric": 6.138750177342445e-05,
          "validation_metric": 5.6258395488839597e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.752312433000043
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.0006150783701741602,
          "train_loss": 0.0005888034356757998,
          "validation_loss": 0.00040406198240816593,
          "train_metric": 5.9302114095771685e-05,
          "validation_metric": 4.069564238307066e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.7670037019997835
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.0006405900676327292,
          "train_loss": 0.0005520398844964802,
          "validation_loss": 0.0004231816856190562,
          "train_metric": 5.5599433835595846e-05,
          "validation_metric": 4.2621315515134484e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.780589871999837
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.0008247027981269639,
          "train_loss": 0.0005514878430403769,
          "validation_loss": 0.00046329115866683424,
          "train_metric": 5.5543830967508256e-05,
          "validation_metric": 4.666098902816884e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.876235547999386
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.0006036862869223114,
          "train_loss": 0.000555630074813962,
          "validation_loss": 0.00044822023482993245,
          "train_metric": 5.596102710114792e-05,
          "validation_metric": 4.514310057857074e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.8885542299994995
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.0006137484970167861,
          "train_loss": 0.0005777671467512846,
          "validation_loss": 0.00046909714001230896,
          "train_metric": 5.819059151690453e-05,
          "validation_metric": 4.724575046566315e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.902266610999504
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.0006243530569918221,
          "train_loss": 0.0005467726150527596,
          "validation_loss": 0.0004242890572641045,
          "train_metric": 5.506893648998812e-05,
          "validation_metric": 4.273283775546588e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.915320231999431
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.0007795616238581715,
          "train_loss": 0.0007737851119600236,
          "validation_loss": 0.0005895443027839065,
          "train_metric": 7.793280383339152e-05,
          "validation_metric": 5.937673995504156e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.9282563050001045
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.0009155681618722156,
          "train_loss": 0.0007984485710039735,
          "validation_loss": 0.0005544377490878105,
          "train_metric": 8.041680848691612e-05,
          "validation_metric": 5.584093742072582e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.940536168000108
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.0005898551189602586,
          "train_loss": 0.0005492949276231229,
          "validation_loss": 0.0003870583022944629,
          "train_metric": 5.532296927412972e-05,
          "validation_metric": 3.89830966014415e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.9534099840002455
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.0006657890335191041,
          "train_loss": 0.0007591308676637709,
          "validation_loss": 0.0006157043389976025,
          "train_metric": 7.645689038326964e-05,
          "validation_metric": 6.20114806224592e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.966028188999189
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0005878552747162757,
          "train_loss": 0.0005779396160505712,
          "validation_loss": 0.0005272873095236719,
          "train_metric": 5.82079665036872e-05,
          "validation_metric": 5.3106450650375336e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.978969039999356
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.0006473475332313683,
          "train_loss": 0.0005470873438753188,
          "validation_loss": 0.00044551375322043896,
          "train_metric": 5.510063419933431e-05,
          "validation_metric": 4.4870510464534163e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 4.991465660999893
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.0007058894734655041,
          "train_loss": 0.0005772561999037862,
          "validation_loss": 0.00037337816320359707,
          "train_metric": 5.8139125030720606e-05,
          "validation_metric": 3.760528488783166e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.004279551998479
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.0005861909830855438,
          "train_loss": 0.0005877104704268277,
          "validation_loss": 0.0005018806550651789,
          "train_metric": 5.9192047046963125e-05,
          "validation_metric": 5.0547583668958396e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.0172613249997084
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.0005500318275153404,
          "train_loss": 0.0004957277560606599,
          "validation_loss": 0.0003791491035372019,
          "train_metric": 4.992788308300078e-05,
          "validation_metric": 3.818651020992547e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.0300040429992805
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0005373520089051453,
          "train_loss": 0.0004929770366288722,
          "validation_loss": 0.00041122466791421175,
          "train_metric": 4.9650850996840745e-05,
          "validation_metric": 4.141703902860172e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.042605040000126
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.0005331047741492512,
          "train_loss": 0.0004921421059407294,
          "validation_loss": 0.0003957028384320438,
          "train_metric": 4.956675547873601e-05,
          "validation_metric": 3.985373768955469e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.054967913998553
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.0005518693214980885,
          "train_loss": 0.0009011764777824283,
          "validation_loss": 0.00048166757915169,
          "train_metric": 9.076319838641211e-05,
          "validation_metric": 4.8511792556382716e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.067903698998634
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.0009654544828663347,
          "train_loss": 0.0010850002290681005,
          "validation_loss": 0.0004825092910323292,
          "train_metric": 0.00010927725816145539,
          "validation_metric": 4.859656473854557e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.080826929999603
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.000861174165038392,
          "train_loss": 0.000521055655553937,
          "validation_loss": 0.000487484474433586,
          "train_metric": 5.247882290859707e-05,
          "validation_metric": 4.909765993943438e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.094219439999506
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.0006301538633124437,
          "train_loss": 0.0005173426470719278,
          "validation_loss": 0.0004502434458117932,
          "train_metric": 5.210486051510088e-05,
          "validation_metric": 4.534687104751356e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.106285384999865
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.0005198322360229213,
          "train_loss": 0.0005074110231362283,
          "validation_loss": 0.00048128922935575247,
          "train_metric": 5.110458005219698e-05,
          "validation_metric": 4.847368836635724e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.119261852998534
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.0005211836678427062,
          "train_loss": 0.0008848614525049925,
          "validation_loss": 0.0008596036350354552,
          "train_metric": 8.912000339478254e-05,
          "validation_metric": 8.657613943796605e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.131990231999225
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.000574767067519133,
          "train_loss": 0.0004924155073240399,
          "validation_loss": 0.0004047191177960485,
          "train_metric": 4.959428770234808e-05,
          "validation_metric": 4.076182813150808e-05,
          "lr_first": 0.0001875,
          "lr_last": 0.0001875,
          "seconds": 5.14476760999969
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.0005081579856778262,
          "train_loss": 0.00049402384320274,
          "validation_loss": 0.000402391015086323,
          "train_metric": 4.9756265070755035e-05,
          "validation_metric": 4.05273494834546e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.157104919999256
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.0004911189171252772,
          "train_loss": 0.0005047902232035995,
          "validation_loss": 0.00034278136445209384,
          "train_metric": 5.084063013782725e-05,
          "validation_metric": 3.4523680369602516e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.170244265998917
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.0005745814905822044,
          "train_loss": 0.0006790000479668379,
          "validation_loss": 0.00038757361471652985,
          "train_metric": 6.838639819761738e-05,
          "validation_metric": 3.903498873114586e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.183151090999672
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.0005710955074391677,
          "train_loss": 0.0005535850650630891,
          "validation_loss": 0.00035416026366874576,
          "train_metric": 5.5755059293005615e-05,
          "validation_metric": 3.566972736734897e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.195691898999939
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.00048075777158373967,
          "train_loss": 0.00046890543308109045,
          "validation_loss": 0.00042270298581570387,
          "train_metric": 4.722643279819749e-05,
          "validation_metric": 4.257309774402529e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.207773448999433
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.00048738964142103214,
          "train_loss": 0.0004616224905475974,
          "validation_loss": 0.000368866982171312,
          "train_metric": 4.649291804526001e-05,
          "validation_metric": 3.715093407663517e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.2200933570002235
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.0004800596652785316,
          "train_loss": 0.0004643156426027417,
          "validation_loss": 0.00040587509283795953,
          "train_metric": 4.676416938309558e-05,
          "validation_metric": 4.087825072929263e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.232698371999504
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.0004976954278390622,
          "train_loss": 0.0005823108367621899,
          "validation_loss": 0.0005426775314845145,
          "train_metric": 5.86482128710486e-05,
          "validation_metric": 5.465649155667052e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.245151661998534
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.000507233414282382,
          "train_loss": 0.0004458669282030314,
          "validation_loss": 0.00036355495103634894,
          "train_metric": 4.4906082621309906e-05,
          "validation_metric": 3.661592199932784e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.25691340099911
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.000530345087099704,
          "train_loss": 0.0005951587809249759,
          "validation_loss": 0.0005350284045562148,
          "train_metric": 5.994221646687947e-05,
          "validation_metric": 5.3886102250544354e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.2690490989989485
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.0004370557517177076,
          "train_loss": 0.00045455689541995525,
          "validation_loss": 0.00033643352799117565,
          "train_metric": 4.578131120069884e-05,
          "validation_metric": 3.388435288798064e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.282328553999832
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.0004935674105581711,
          "train_loss": 0.0005535155651159585,
          "validation_loss": 0.0003407163603696972,
          "train_metric": 5.5748067097738385e-05,
          "validation_metric": 3.43157043971587e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.295484053000109
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.0005044676581746899,
          "train_loss": 0.00044881858048029244,
          "validation_loss": 0.00033743749372661114,
          "train_metric": 4.520336369751021e-05,
          "validation_metric": 3.398546687094495e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.307221606999519
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.00045442680948326597,
          "train_loss": 0.0004345641646068543,
          "validation_loss": 0.00034638712531886995,
          "train_metric": 4.376770812086761e-05,
          "validation_metric": 3.4886845241999254e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.319789127999684
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.00044488948151411023,
          "train_loss": 0.00043269165325909853,
          "validation_loss": 0.00034498446621000767,
          "train_metric": 4.357911893748678e-05,
          "validation_metric": 3.474557161098346e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.398641361000045
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.00044715044714394026,
          "train_loss": 0.00044416956370696425,
          "validation_loss": 0.00039454910438507795,
          "train_metric": 4.473513035918586e-05,
          "validation_metric": 3.973754428443499e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.411475797998719
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0004896683039987693,
          "train_loss": 0.000570710573811084,
          "validation_loss": 0.0005412119207903743,
          "train_metric": 5.747988325310871e-05,
          "validation_metric": 5.450887692859396e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.423092944000018
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.0004755832087539602,
          "train_loss": 0.0004522438393905759,
          "validation_loss": 0.00041296257404610515,
          "train_metric": 4.554833867587149e-05,
          "validation_metric": 4.159208037890494e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.435784681998484
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.00045410730672301725,
          "train_loss": 0.0004386646905913949,
          "validation_loss": 0.00033436762169003487,
          "train_metric": 4.4180698751006275e-05,
          "validation_metric": 3.3676285966066644e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.448897590998968
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.00044004292158206226,
          "train_loss": 0.0004527380224317312,
          "validation_loss": 0.0003227265551686287,
          "train_metric": 4.559811350191012e-05,
          "validation_metric": 3.2503838156117126e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.4621504609986005
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.000500216967338929,
          "train_loss": 0.000484871823573485,
          "validation_loss": 0.0003202769730705768,
          "train_metric": 4.883451765635982e-05,
          "validation_metric": 3.225712134735659e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.47505707199889
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.0004867843417741824,
          "train_loss": 0.0004231053462717682,
          "validation_loss": 0.00033718920894898474,
          "train_metric": 4.261361755197868e-05,
          "validation_metric": 3.3960459404625e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.4874481839997316
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.00046320617730089,
          "train_loss": 0.0005271767149679363,
          "validation_loss": 0.00047185138100758195,
          "train_metric": 5.309530388331041e-05,
          "validation_metric": 4.752314634970389e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.500576925998757
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.0004790739039890468,
          "train_loss": 0.00043759087566286325,
          "validation_loss": 0.00036449363688006997,
          "train_metric": 4.407254891702905e-05,
          "validation_metric": 3.671046579256654e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.5131151329987915
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.000426483852606907,
          "train_loss": 0.00047087823622860014,
          "validation_loss": 0.00031924978247843683,
          "train_metric": 4.7425135562662035e-05,
          "validation_metric": 3.215367178199813e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.525865221999993
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.0004654811091313604,
          "train_loss": 0.0004162804107181728,
          "validation_loss": 0.0003369271580595523,
          "train_metric": 4.192623600829393e-05,
          "validation_metric": 3.3934069506358355e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.5379602599987265
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.00044791386699216673,
          "train_loss": 0.0005566140171140432,
          "validation_loss": 0.0005139369750395417,
          "train_metric": 5.6060118367895484e-05,
          "validation_metric": 5.176185004529543e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.550346672998785
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.0004945004329783842,
          "train_loss": 0.0004404694482218474,
          "validation_loss": 0.00037963438080623746,
          "train_metric": 4.436247036210261e-05,
          "validation_metric": 3.823538281721994e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.562692053999854
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.000489812340674689,
          "train_loss": 0.0007870064000599086,
          "validation_loss": 0.000704663572832942,
          "train_metric": 7.926440594019368e-05,
          "validation_metric": 7.097113848431036e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.574667250999482
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.0005188627619645558,
          "train_loss": 0.0006467502098530531,
          "validation_loss": 0.000575163634493947,
          "train_metric": 6.51383088552393e-05,
          "validation_metric": 5.7928376918425784e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.586790981999002
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.0005558930970437359,
          "train_loss": 0.0006822713185101748,
          "validation_loss": 0.0006460783770307899,
          "train_metric": 6.871586811030284e-05,
          "validation_metric": 6.507064972538501e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.599361691998638
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.0006238827882043552,
          "train_loss": 0.0005099262343719602,
          "validation_loss": 0.0004288783820811659,
          "train_metric": 5.13579034304712e-05,
          "validation_metric": 4.319506115280092e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.611770176999926
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.0004944497686665272,
          "train_loss": 0.0005602865130640566,
          "validation_loss": 0.00048247736413031816,
          "train_metric": 5.643000622512773e-05,
          "validation_metric": 4.85933487652801e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.623768785999346
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.00043735248982557096,
          "train_loss": 0.0004678351106122136,
          "validation_loss": 0.0004494445165619254,
          "train_metric": 4.7118639486143366e-05,
          "validation_metric": 4.526640987023711e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.636166681999384
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.0004613627152139088,
          "train_loss": 0.0004244947340339422,
          "validation_loss": 0.00031465364736504853,
          "train_metric": 4.275355240679346e-05,
          "validation_metric": 3.169076080666855e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.649458299998514
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.0004273855702194851,
          "train_loss": 0.00042724268860183656,
          "validation_loss": 0.00030340885859914124,
          "train_metric": 4.303032619645819e-05,
          "validation_metric": 3.055823253816925e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.6631359469993185
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.0004229938695061719,
          "train_loss": 0.0004100508231204003,
          "validation_loss": 0.00033377087675035,
          "train_metric": 4.129881563130766e-05,
          "validation_metric": 3.361617564223707e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.675343233999229
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.0004391639849927742,
          "train_loss": 0.00039672685670666397,
          "validation_loss": 0.000311166571918875,
          "train_metric": 3.995687438873574e-05,
          "validation_metric": 3.133955760858953e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.687693602998479
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.0004636430476239184,
          "train_loss": 0.0005337524344213307,
          "validation_loss": 0.00030185183277353644,
          "train_metric": 5.375758701120503e-05,
          "validation_metric": 3.040141200472135e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.702231146999111
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.0004966794322172063,
          "train_loss": 0.00039846845902502537,
          "validation_loss": 0.0003447423514444381,
          "train_metric": 4.013228681287728e-05,
          "validation_metric": 3.472118987701833e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.714990327998748
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.00046033445414650487,
          "train_loss": 0.00039892594213597476,
          "validation_loss": 0.0002960850251838565,
          "train_metric": 4.017835817649029e-05,
          "validation_metric": 2.982060323120095e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.7280584889995225
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.00041211352890968556,
          "train_loss": 0.0004882235953118652,
          "validation_loss": 0.000304011715343222,
          "train_metric": 4.917208934784867e-05,
          "validation_metric": 3.0618946766480803e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.740413841998816
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.00041997796506620944,
          "train_loss": 0.0004521913069766015,
          "validation_loss": 0.00029500937671400607,
          "train_metric": 4.5543056330643594e-05,
          "validation_metric": 2.9712267860304564e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.753051591000258
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.0004806744764209725,
          "train_loss": 0.00042749717249535024,
          "validation_loss": 0.00029849846032448113,
          "train_metric": 4.3055952119175345e-05,
          "validation_metric": 3.0063671147217974e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.765678500998547
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.0004356630188340205,
          "train_loss": 0.0003856720868498087,
          "validation_loss": 0.0003305445425212383,
          "train_metric": 3.884348552674055e-05,
          "validation_metric": 3.329123501316644e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.777977405999991
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.0004725265116576338,
          "train_loss": 0.0003844928869511932,
          "validation_loss": 0.0003075732383877039,
          "train_metric": 3.872471643262543e-05,
          "validation_metric": 3.097764783888124e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.789703817999907
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.00042991039117623586,
          "train_loss": 0.0003965286014135927,
          "validation_loss": 0.0003462503955233842,
          "train_metric": 3.993690916104242e-05,
          "validation_metric": 3.48730682162568e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.802264085999923
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.0005104528190713609,
          "train_loss": 0.0005995709216222167,
          "validation_loss": 0.00036643550265580416,
          "train_metric": 6.038659194018692e-05,
          "validation_metric": 3.6906047171214595e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.814752906999274
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.00045439806672220584,
          "train_loss": 0.0004827219236176461,
          "validation_loss": 0.00028927557286806405,
          "train_metric": 4.861799243371934e-05,
          "validation_metric": 2.9134780561435036e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.827598545000001
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.0005404595558502479,
          "train_loss": 0.0013710764469578862,
          "validation_loss": 0.0007509591523557901,
          "train_metric": 0.00013808978837914765,
          "validation_metric": 7.563385588582605e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.839442215999952
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.000815811723441584,
          "train_loss": 0.0005822476814500988,
          "validation_loss": 0.0003858324489556253,
          "train_metric": 5.86418536840938e-05,
          "validation_metric": 3.8859630876686424e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.918217009999353
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.00046122765616019024,
          "train_loss": 0.0003852869267575443,
          "validation_loss": 0.00033840458490885794,
          "train_metric": 3.880468648276292e-05,
          "validation_metric": 3.408287375350483e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.931381225998848
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.0003650441303761909,
          "train_loss": 0.00044829497346654534,
          "validation_loss": 0.0002960817073471844,
          "train_metric": 4.515062755672261e-05,
          "validation_metric": 2.982026489917189e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.943604607999077
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.00041498539576423354,
          "train_loss": 0.0003912628744728863,
          "validation_loss": 0.00027807726291939616,
          "train_metric": 3.9406560972565785e-05,
          "validation_metric": 2.8006923457724042e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.955981922999854
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.0004326074486016296,
          "train_loss": 0.000375264702597633,
          "validation_loss": 0.000316939054755494,
          "train_metric": 3.779528560698964e-05,
          "validation_metric": 3.192093572579324e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.968900842999574
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.00043043527603003895,
          "train_loss": 0.00037568859988823533,
          "validation_loss": 0.00030430196784436703,
          "train_metric": 3.7837980926269665e-05,
          "validation_metric": 3.064817428821698e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.981672005998917
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.0003903599972545635,
          "train_loss": 0.00037308025639504194,
          "validation_loss": 0.0003018516581505537,
          "train_metric": 3.757527520065196e-05,
          "validation_metric": 3.0401391995837912e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 5.993719599999167
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.00040716522198636085,
          "train_loss": 0.0003683382528834045,
          "validation_loss": 0.00031137646874412894,
          "train_metric": 3.7097684980835766e-05,
          "validation_metric": 3.136069790343754e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.005278609000015
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.00040686106876819395,
          "train_loss": 0.0004456151800695807,
          "validation_loss": 0.00041111206519417465,
          "train_metric": 4.488073318498209e-05,
          "validation_metric": 4.140570308663882e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.017435459998524
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.00047548816473863553,
          "train_loss": 0.00038628443144261837,
          "validation_loss": 0.00026655677356757224,
          "train_metric": 3.8905149267520756e-05,
          "validation_metric": 2.6846621040022e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.030529800000295
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.00041051823427551426,
          "train_loss": 0.00042629867675714195,
          "validation_loss": 0.0003061905736103654,
          "train_metric": 4.293524398235604e-05,
          "validation_metric": 3.083839692408219e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.042731810999612
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.00040550847734266426,
          "train_loss": 0.00037055183202028275,
          "validation_loss": 0.00029196380637586117,
          "train_metric": 3.7320620322134346e-05,
          "validation_metric": 2.9405526220216416e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.054497429999174
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.0004450246933629387,
          "train_loss": 0.0005208317306824028,
          "validation_loss": 0.00047131211613304913,
          "train_metric": 5.245626380201429e-05,
          "validation_metric": 4.7468831326114014e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.066939901998921
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.00047570312563038897,
          "train_loss": 0.0004004082875326276,
          "validation_loss": 0.0002731317945290357,
          "train_metric": 4.032765355077572e-05,
          "validation_metric": 2.750883504631929e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.079196561999197
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.0004987935335520888,
          "train_loss": 0.00042213546112179756,
          "validation_loss": 0.00031820550793781877,
          "train_metric": 4.251594145898707e-05,
          "validation_metric": 3.20484941767063e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.090867520999382
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.00043338778687029844,
          "train_loss": 0.00045710388803854585,
          "validation_loss": 0.0003108917735517025,
          "train_metric": 4.603782872436568e-05,
          "validation_metric": 3.131188350380398e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.102518857000177
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.0004903434710286092,
          "train_loss": 0.0004621543630491942,
          "validation_loss": 0.0002973369264509529,
          "train_metric": 4.6546490921173245e-05,
          "validation_metric": 2.994669193867594e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.115344260999336
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.00046342654695763486,
          "train_loss": 0.0003653688763733953,
          "validation_loss": 0.00029946703580208123,
          "train_metric": 3.679861401906237e-05,
          "validation_metric": 3.0161221729940735e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.127222204999271
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.0004108685780011001,
          "train_loss": 0.00041985500138252974,
          "validation_loss": 0.0003456507110968232,
          "train_metric": 4.228625766700134e-05,
          "validation_metric": 3.481267049210146e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.1384137329987425
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.0004262535594534711,
          "train_loss": 0.0003500965831335634,
          "validation_loss": 0.0002896819787565619,
          "train_metric": 3.526044383761473e-05,
          "validation_metric": 2.9175707823014818e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.1504189309998765
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.0003868026060445118,
          "train_loss": 0.00035308158840052783,
          "validation_loss": 0.00027804463752545416,
          "train_metric": 3.556108276825398e-05,
          "validation_metric": 2.8003636543871835e-05,
          "lr_first": 9.375e-05,
          "lr_last": 9.375e-05,
          "seconds": 6.162435842999912
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.0003699590406540665,
          "train_loss": 0.00041294493712484837,
          "validation_loss": 0.00034400203730911016,
          "train_metric": 4.159030140726827e-05,
          "validation_metric": 3.464662586338818e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 6.173986979998517
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.00038873395260452526,
          "train_loss": 0.0004061054205521941,
          "validation_loss": 0.0003626094257924706,
          "train_metric": 4.0901450120145455e-05,
          "validation_metric": 3.6520694266073406e-05,
          "lr_first": 4.6875e-05,
          "lr_last": 4.6875e-05,
          "seconds": 6.1855608519999805
        }
      ],
      "best": {
        "epoch": 227,
        "updates": 3632
      },
      "best_validation_metric": 2.6846621040022e-05,
      "config": {
        "label": "LayerNorm + tanh + plateau LR",
        "batch": 64,
        "lr": 0.003,
        "schedule": "plateau",
        "normalization": "layer",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "tanh",
        "loss": "mse",
        "outliers": false,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2369,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.005181372505429618
    },
    "fc71c16e3acd_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.8095524124801159,
          "train_loss": 0.503296971321106,
          "validation_loss": 0.34211865067481995,
          "train_metric": 0.0690932646393776,
          "validation_metric": 0.020801134407520294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.017665756000496913
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.42814991995692253,
          "train_loss": 0.4000873267650604,
          "validation_loss": 0.23568782210350037,
          "train_metric": 0.05990787222981453,
          "validation_metric": 0.012743889354169369,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0357528620006633
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.35065119434148073,
          "train_loss": 0.31924280524253845,
          "validation_loss": 0.17192628979682922,
          "train_metric": 0.05539403855800629,
          "validation_metric": 0.012066059745848179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05422017099954246
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.3039058158174157,
          "train_loss": 0.2904966473579407,
          "validation_loss": 0.13696494698524475,
          "train_metric": 0.05374162644147873,
          "validation_metric": 0.009040376171469688,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07259096099915041
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.2854176159016788,
          "train_loss": 0.2817190885543823,
          "validation_loss": 0.13000249862670898,
          "train_metric": 0.051405150443315506,
          "validation_metric": 0.006706669926643372,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.09080291499958548
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.28044577315449715,
          "train_loss": 0.2802561819553375,
          "validation_loss": 0.12973041832447052,
          "train_metric": 0.05272042751312256,
          "validation_metric": 0.006925154943019152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.108837121999386
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.2775714402087033,
          "train_loss": 0.27370429039001465,
          "validation_loss": 0.12614277005195618,
          "train_metric": 0.05252084508538246,
          "validation_metric": 0.007259009405970573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.1263457129989547
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.2713735783472657,
          "train_loss": 0.26402413845062256,
          "validation_loss": 0.11854177713394165,
          "train_metric": 0.05047035962343216,
          "validation_metric": 0.006464026402682066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14455133699993894
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.2600850220769644,
          "train_loss": 0.2576831877231598,
          "validation_loss": 0.11175406724214554,
          "train_metric": 0.04931291192770004,
          "validation_metric": 0.005275217350572348,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16288972300026217
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.2530387924052775,
          "train_loss": 0.25070303678512573,
          "validation_loss": 0.10492081195116043,
          "train_metric": 0.05039924755692482,
          "validation_metric": 0.0058966027572751045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18005372100014938
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.24110363237559795,
          "train_loss": 0.23335248231887817,
          "validation_loss": 0.08721688389778137,
          "train_metric": 0.04876479133963585,
          "validation_metric": 0.004033987410366535,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19787710899981903
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.23286512214690447,
          "train_loss": 0.22540971636772156,
          "validation_loss": 0.07558617740869522,
          "train_metric": 0.04849616438150406,
          "validation_metric": 0.0033279925119131804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21612812300008954
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.2266788287088275,
          "train_loss": 0.23010914027690887,
          "validation_loss": 0.07574989646673203,
          "train_metric": 0.048734769225120544,
          "validation_metric": 0.0023964056745171547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23432966700056568
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.22589328698813915,
          "train_loss": 0.21906565129756927,
          "validation_loss": 0.07007743418216705,
          "train_metric": 0.047786615788936615,
          "validation_metric": 0.00265680905431509,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2512623259990505
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.21959985606372356,
          "train_loss": 0.22032737731933594,
          "validation_loss": 0.0718291848897934,
          "train_metric": 0.046991005539894104,
          "validation_metric": 0.002353315008804202,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2688857499997539
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.21552641643211246,
          "train_loss": 0.21259544789791107,
          "validation_loss": 0.06326719373464584,
          "train_metric": 0.047477662563323975,
          "validation_metric": 0.0022766322363168,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2899695720007003
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.21405377169139683,
          "train_loss": 0.2163914144039154,
          "validation_loss": 0.06805943697690964,
          "train_metric": 0.047599226236343384,
          "validation_metric": 0.0023117666132748127,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33335280699975556
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.2134090238250792,
          "train_loss": 0.21236813068389893,
          "validation_loss": 0.060884371399879456,
          "train_metric": 0.0473230741918087,
          "validation_metric": 0.001678701606579125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.35959413799901085
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.20909020956605673,
          "train_loss": 0.21014700829982758,
          "validation_loss": 0.06135375052690506,
          "train_metric": 0.04687667638063431,
          "validation_metric": 0.0022506974637508392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3777390909999667
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.2144334358163178,
          "train_loss": 0.20710834860801697,
          "validation_loss": 0.057873792946338654,
          "train_metric": 0.04673932120203972,
          "validation_metric": 0.001907133380882442,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3949189029990521
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.21852638479322195,
          "train_loss": 0.21964909136295319,
          "validation_loss": 0.06584838032722473,
          "train_metric": 0.046270642429590225,
          "validation_metric": 0.0018722284585237503,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41155962999982876
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.21329667512327433,
          "train_loss": 0.21514858305454254,
          "validation_loss": 0.06287891417741776,
          "train_metric": 0.046220552176237106,
          "validation_metric": 0.001579748815856874,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.42935840300015116
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.20885114325210452,
          "train_loss": 0.2016110122203827,
          "validation_loss": 0.05079096555709839,
          "train_metric": 0.04679933935403824,
          "validation_metric": 0.0015316666103899479,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4474775380003848
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.20572142279706895,
          "train_loss": 0.20265993475914001,
          "validation_loss": 0.05162736028432846,
          "train_metric": 0.046816304326057434,
          "validation_metric": 0.0013654802460223436,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.46508156200070516
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.2046785894781351,
          "train_loss": 0.2007627636194229,
          "validation_loss": 0.04919073358178139,
          "train_metric": 0.04659951105713844,
          "validation_metric": 0.0015285566914826632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4905235949991038
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.204103444237262,
          "train_loss": 0.20189592242240906,
          "validation_loss": 0.050645362585783005,
          "train_metric": 0.04711925610899925,
          "validation_metric": 0.0016657061642035842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5099387299997034
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.20223903842270374,
          "train_loss": 0.20422004163265228,
          "validation_loss": 0.05432863160967827,
          "train_metric": 0.04725322127342224,
          "validation_metric": 0.001421953784301877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5286579200001142
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.20026368787512183,
          "train_loss": 0.1965322643518448,
          "validation_loss": 0.04333508759737015,
          "train_metric": 0.04657568037509918,
          "validation_metric": 0.0012996428413316607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5482678039988969
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.20400555385276675,
          "train_loss": 0.20200982689857483,
          "validation_loss": 0.04830662161111832,
          "train_metric": 0.04632716253399849,
          "validation_metric": 0.0010169845772907138,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6681566619990917
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.2035908573307097,
          "train_loss": 0.20012520253658295,
          "validation_loss": 0.04790673777461052,
          "train_metric": 0.046368613839149475,
          "validation_metric": 0.0016767524648457766,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6855938859989692
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.19685873109847307,
          "train_loss": 0.19424982368946075,
          "validation_loss": 0.04543166980147362,
          "train_metric": 0.04683529585599899,
          "validation_metric": 0.001006284961476922,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7040567359999841
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.19613979291170835,
          "train_loss": 0.19411781430244446,
          "validation_loss": 0.04561963677406311,
          "train_metric": 0.04635981097817421,
          "validation_metric": 0.001227084780111909,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7210558900005708
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.19570578262209892,
          "train_loss": 0.19725728034973145,
          "validation_loss": 0.0442090705037117,
          "train_metric": 0.04695027694106102,
          "validation_metric": 0.0008605129551142454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7382014569993771
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.19636108307167888,
          "train_loss": 0.2021588236093521,
          "validation_loss": 0.048853691667318344,
          "train_metric": 0.047345492988824844,
          "validation_metric": 0.0008732607821002603,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7550216300005559
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.201477930881083,
          "train_loss": 0.20641465485095978,
          "validation_loss": 0.05480189248919487,
          "train_metric": 0.047068167477846146,
          "validation_metric": 0.0013743836898356676,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7719616880003741
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.19895631074905396,
          "train_loss": 0.1950559914112091,
          "validation_loss": 0.04439372941851616,
          "train_metric": 0.046752043068408966,
          "validation_metric": 0.0008731411653570831,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7885636749997502
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.1932050953619182,
          "train_loss": 0.18894325196743011,
          "validation_loss": 0.03710576891899109,
          "train_metric": 0.04623089358210564,
          "validation_metric": 0.0009812171338126063,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8053595519995724
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.1899584592320025,
          "train_loss": 0.18963654339313507,
          "validation_loss": 0.037973709404468536,
          "train_metric": 0.04640081152319908,
          "validation_metric": 0.0009549788082949817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8220423349994235
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.193441872484982,
          "train_loss": 0.19650433957576752,
          "validation_loss": 0.04369427636265755,
          "train_metric": 0.04605713114142418,
          "validation_metric": 0.0008027316653169692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8394585160003771
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.19249785924330354,
          "train_loss": 0.19287937879562378,
          "validation_loss": 0.042038679122924805,
          "train_metric": 0.04690000042319298,
          "validation_metric": 0.00069382757646963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.857196020999254
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.19440086791291833,
          "train_loss": 0.1996990144252777,
          "validation_loss": 0.05189061537384987,
          "train_metric": 0.04673641920089722,
          "validation_metric": 0.00140459428075701,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8741936259993963
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.19133672816678882,
          "train_loss": 0.18976537883281708,
          "validation_loss": 0.037349555641412735,
          "train_metric": 0.046417541801929474,
          "validation_metric": 0.0006408683839254081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.891307293999489
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.19516094122081995,
          "train_loss": 0.21130862832069397,
          "validation_loss": 0.06173202022910118,
          "train_metric": 0.04589970409870148,
          "validation_metric": 0.0015762177063152194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.908016872999724
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.19859229028224945,
          "train_loss": 0.19509151577949524,
          "validation_loss": 0.043307114392519,
          "train_metric": 0.04576491191983223,
          "validation_metric": 0.0006504353368654847,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9247562410000683
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.1935715761501342,
          "train_loss": 0.19244469702243805,
          "validation_loss": 0.03886822611093521,
          "train_metric": 0.04644128307700157,
          "validation_metric": 0.0005919078830629587,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9436926229991514
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.19501249794848263,
          "train_loss": 0.19307395815849304,
          "validation_loss": 0.03960715979337692,
          "train_metric": 0.046373289078474045,
          "validation_metric": 0.0006948398076929152,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9618255700006557
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.19093380775302649,
          "train_loss": 0.2132163792848587,
          "validation_loss": 0.06191194802522659,
          "train_metric": 0.0465075857937336,
          "validation_metric": 0.0020858945790678263,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9785838409989083
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.200002646073699,
          "train_loss": 0.19463412463665009,
          "validation_loss": 0.04533137381076813,
          "train_metric": 0.045847389847040176,
          "validation_metric": 0.0008758407202549279,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9955976359997294
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.1930690344888717,
          "train_loss": 0.1919638216495514,
          "validation_loss": 0.03870808333158493,
          "train_metric": 0.04560469463467598,
          "validation_metric": 0.0007846801890991628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.013237551000202
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.19311727676540613,
          "train_loss": 0.1928967833518982,
          "validation_loss": 0.04459219053387642,
          "train_metric": 0.046715524047613144,
          "validation_metric": 0.0007236461970023811,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0294531780000398
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.18697846680879593,
          "train_loss": 0.18444743752479553,
          "validation_loss": 0.03482243791222572,
          "train_metric": 0.0461115725338459,
          "validation_metric": 0.0007840113248676062,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0413979919994745
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.18714532814919949,
          "train_loss": 0.1880456507205963,
          "validation_loss": 0.03426426649093628,
          "train_metric": 0.046614281833171844,
          "validation_metric": 0.0004790577513631433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0532423320000817
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.1863470054231584,
          "train_loss": 0.18404518067836761,
          "validation_loss": 0.03200717642903328,
          "train_metric": 0.04628157615661621,
          "validation_metric": 0.0006048297509551048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.064737893999336
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.1847479077987373,
          "train_loss": 0.18337607383728027,
          "validation_loss": 0.03232048824429512,
          "train_metric": 0.04623808339238167,
          "validation_metric": 0.0005086925812065601,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.076110403999337
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.18740401836112142,
          "train_loss": 0.18551526963710785,
          "validation_loss": 0.03096904419362545,
          "train_metric": 0.045860376209020615,
          "validation_metric": 0.0004762744065374136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0876177259997348
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.18850488308817148,
          "train_loss": 0.19853833317756653,
          "validation_loss": 0.049810927361249924,
          "train_metric": 0.04586923122406006,
          "validation_metric": 0.0010819460730999708,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0989910819989746
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.19217094406485558,
          "train_loss": 0.1935335397720337,
          "validation_loss": 0.037986382842063904,
          "train_metric": 0.04656055197119713,
          "validation_metric": 0.0005154329701326787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1101975589990616
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.1893217775505036,
          "train_loss": 0.1879819631576538,
          "validation_loss": 0.03289933502674103,
          "train_metric": 0.04683693125844002,
          "validation_metric": 0.0004959665820933878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1210612419999961
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.18836284847930074,
          "train_loss": 0.1836491823196411,
          "validation_loss": 0.03555195406079292,
          "train_metric": 0.046002671122550964,
          "validation_metric": 0.000615968368947506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1322920940001495
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.18247899133712053,
          "train_loss": 0.18027625977993011,
          "validation_loss": 0.02792087011039257,
          "train_metric": 0.04628244787454605,
          "validation_metric": 0.00046269813901744783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.144064227999479
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.18561590765602887,
          "train_loss": 0.17912203073501587,
          "validation_loss": 0.027200572192668915,
          "train_metric": 0.046027302742004395,
          "validation_metric": 0.00043951801490038633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1557699269997102
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.18708778591826558,
          "train_loss": 0.18594171106815338,
          "validation_loss": 0.03430885076522827,
          "train_metric": 0.04628922417759895,
          "validation_metric": 0.00045700540067628026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1670649209991097
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.1831446154974401,
          "train_loss": 0.18174773454666138,
          "validation_loss": 0.030637066811323166,
          "train_metric": 0.04606049135327339,
          "validation_metric": 0.0004252612416166812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1794414050000341
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.18265868350863457,
          "train_loss": 0.18250475823879242,
          "validation_loss": 0.03369138017296791,
          "train_metric": 0.04588451609015465,
          "validation_metric": 0.0004158992669545114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.191217252999195
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.18332429183647037,
          "train_loss": 0.1828826516866684,
          "validation_loss": 0.03151002526283264,
          "train_metric": 0.046599648892879486,
          "validation_metric": 0.0003629957791417837,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2031204359991534
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.18709478713572025,
          "train_loss": 0.18289566040039062,
          "validation_loss": 0.033846329897642136,
          "train_metric": 0.045826416462659836,
          "validation_metric": 0.000558700121473521,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2144716500006325
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.18426694651134312,
          "train_loss": 0.19429554045200348,
          "validation_loss": 0.04176895320415497,
          "train_metric": 0.047140322625637054,
          "validation_metric": 0.0005578277632594109,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2259360030002426
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.18708814796991646,
          "train_loss": 0.18666163086891174,
          "validation_loss": 0.03320077434182167,
          "train_metric": 0.04614511504769325,
          "validation_metric": 0.0009112187544815242,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2370078759995522
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.18568309536203742,
          "train_loss": 0.18054862320423126,
          "validation_loss": 0.028046609833836555,
          "train_metric": 0.04652594029903412,
          "validation_metric": 0.0003459216095507145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2488250930000504
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.18211004277691245,
          "train_loss": 0.18782560527324677,
          "validation_loss": 0.03669188916683197,
          "train_metric": 0.046120546758174896,
          "validation_metric": 0.0003809825284406543,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2601453440001933
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.1874646432697773,
          "train_loss": 0.1882735639810562,
          "validation_loss": 0.03331874683499336,
          "train_metric": 0.04660949110984802,
          "validation_metric": 0.00046291504986584187,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2714351039994654
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.1887186556123197,
          "train_loss": 0.18696697056293488,
          "validation_loss": 0.033128008246421814,
          "train_metric": 0.04663146659731865,
          "validation_metric": 0.0003833965747617185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2828047650000372
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.1830790478270501,
          "train_loss": 0.1853451281785965,
          "validation_loss": 0.029716622084379196,
          "train_metric": 0.04655943438410759,
          "validation_metric": 0.00039488152833655477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2942744649990345
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.18456867756322026,
          "train_loss": 0.18700271844863892,
          "validation_loss": 0.036308370530605316,
          "train_metric": 0.045763712376356125,
          "validation_metric": 0.00044127978617325425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3053207819993986
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.18516194121912122,
          "train_loss": 0.18248726427555084,
          "validation_loss": 0.027584517374634743,
          "train_metric": 0.04600561037659645,
          "validation_metric": 0.0005678217858076096,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3167668209989642
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.18165161204524338,
          "train_loss": 0.1764351725578308,
          "validation_loss": 0.025401093065738678,
          "train_metric": 0.045782674103975296,
          "validation_metric": 0.0003847129992209375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3280008669989911
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.18014511791989207,
          "train_loss": 0.188447043299675,
          "validation_loss": 0.033245254307985306,
          "train_metric": 0.04682064801454544,
          "validation_metric": 0.00041963905096054077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3392628749988944
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.18253053352236748,
          "train_loss": 0.1764126420021057,
          "validation_loss": 0.02592327632009983,
          "train_metric": 0.045886434614658356,
          "validation_metric": 0.0003710537275765091,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3506585439990886
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.17874291131738573,
          "train_loss": 0.1777917593717575,
          "validation_loss": 0.025807097554206848,
          "train_metric": 0.045772161334753036,
          "validation_metric": 0.000417192408349365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3623265410005843
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.17751795379444957,
          "train_loss": 0.17450502514839172,
          "validation_loss": 0.022528236731886864,
          "train_metric": 0.04612113907933235,
          "validation_metric": 0.0002807738783303648,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3744590829992376
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.17713332502171397,
          "train_loss": 0.17792025208473206,
          "validation_loss": 0.026839643716812134,
          "train_metric": 0.04619131609797478,
          "validation_metric": 0.00028005600324831903,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3863614420006343
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.18277525145094842,
          "train_loss": 0.1780724674463272,
          "validation_loss": 0.028147384524345398,
          "train_metric": 0.04634648934006691,
          "validation_metric": 0.00027456454699859023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3981712410004548
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.18278086534701288,
          "train_loss": 0.1788630485534668,
          "validation_loss": 0.02519955486059189,
          "train_metric": 0.046386417001485825,
          "validation_metric": 0.0003288722364231944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4095271410005807
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.17999267578125,
          "train_loss": 0.1879163384437561,
          "validation_loss": 0.03342049568891525,
          "train_metric": 0.046207405626773834,
          "validation_metric": 0.0008366444963030517,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4206488159998116
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.1785701857879758,
          "train_loss": 0.1758878380060196,
          "validation_loss": 0.023263800889253616,
          "train_metric": 0.046194300055503845,
          "validation_metric": 0.00024346521240659058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4330960999996023
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.17613146919757128,
          "train_loss": 0.17386966943740845,
          "validation_loss": 0.020433174446225166,
          "train_metric": 0.04596059024333954,
          "validation_metric": 0.0003116356092505157,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.444310487999246
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.17554350942373276,
          "train_loss": 0.18144471943378448,
          "validation_loss": 0.027585336938500404,
          "train_metric": 0.04649984836578369,
          "validation_metric": 0.00032795898732729256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4562505619996955
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.17996767675504088,
          "train_loss": 0.17299866676330566,
          "validation_loss": 0.022108029574155807,
          "train_metric": 0.04595820978283882,
          "validation_metric": 0.00023195888206828386,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4680089369994676
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.1807665901724249,
          "train_loss": 0.1734246462583542,
          "validation_loss": 0.020113343372941017,
          "train_metric": 0.04618888720870018,
          "validation_metric": 0.0002376454067416489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4792121020000195
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.18564244464505464,
          "train_loss": 0.18944643437862396,
          "validation_loss": 0.034884147346019745,
          "train_metric": 0.04691489413380623,
          "validation_metric": 0.0002925847074948251,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.490119923999373
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.18167051952332258,
          "train_loss": 0.18389584124088287,
          "validation_loss": 0.03083626925945282,
          "train_metric": 0.045663021504879,
          "validation_metric": 0.000483242969494313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.501292076000027
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.18268524098675698,
          "train_loss": 0.1932493895292282,
          "validation_loss": 0.0379895344376564,
          "train_metric": 0.0462564192712307,
          "validation_metric": 0.0008486853912472725,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5123666919989773
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.19213656522333622,
          "train_loss": 0.18853168189525604,
          "validation_loss": 0.03210437297821045,
          "train_metric": 0.04656246304512024,
          "validation_metric": 0.00037021070602349937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5233988479994878
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.1873335288837552,
          "train_loss": 0.18342876434326172,
          "validation_loss": 0.0260202344506979,
          "train_metric": 0.04686271771788597,
          "validation_metric": 0.0003388433833606541,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.534203339000669
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.18138316459953785,
          "train_loss": 0.17661814391613007,
          "validation_loss": 0.022660348564386368,
          "train_metric": 0.045956239104270935,
          "validation_metric": 0.00040271514444611967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5449709999993502
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.17680354497861117,
          "train_loss": 0.1778627634048462,
          "validation_loss": 0.02836446464061737,
          "train_metric": 0.045821208506822586,
          "validation_metric": 0.00024173129349946976,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5553056720000313
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.17918944580014795,
          "train_loss": 0.17684654891490936,
          "validation_loss": 0.023521315306425095,
          "train_metric": 0.04591412469744682,
          "validation_metric": 0.0002787256962619722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5660028519996558
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.17947422177530825,
          "train_loss": 0.1765485256910324,
          "validation_loss": 0.02181379124522209,
          "train_metric": 0.046469494700431824,
          "validation_metric": 0.00022742067812941968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5768184640000982
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.17656340054236352,
          "train_loss": 0.18310041725635529,
          "validation_loss": 0.0324169360101223,
          "train_metric": 0.04588647559285164,
          "validation_metric": 0.00048711104318499565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6598649450006633
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.1783178278710693,
          "train_loss": 0.18248099088668823,
          "validation_loss": 0.03304802626371384,
          "train_metric": 0.045897092670202255,
          "validation_metric": 0.00027046806644648314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6705265780001355
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.17778121354058385,
          "train_loss": 0.17292258143424988,
          "validation_loss": 0.022720761597156525,
          "train_metric": 0.04609634354710579,
          "validation_metric": 0.00020568212494254112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6818364089995157
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.18026228435337543,
          "train_loss": 0.18543948233127594,
          "validation_loss": 0.03165287896990776,
          "train_metric": 0.04572153836488724,
          "validation_metric": 0.000338792015099898,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6925050459994964
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.1770916546229273,
          "train_loss": 0.17194028198719025,
          "validation_loss": 0.018568575382232666,
          "train_metric": 0.04602022469043732,
          "validation_metric": 0.00025408578221686184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7029539729992393
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.17550251143984497,
          "train_loss": 0.1779135763645172,
          "validation_loss": 0.023631054908037186,
          "train_metric": 0.046379204839468,
          "validation_metric": 0.00027293761377222836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7135135070002434
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.17575468809809536,
          "train_loss": 0.17673444747924805,
          "validation_loss": 0.022399289533495903,
          "train_metric": 0.045985929667949677,
          "validation_metric": 0.00030768310534767807,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7239482679997309
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.1764399913372472,
          "train_loss": 0.18470776081085205,
          "validation_loss": 0.028772303834557533,
          "train_metric": 0.04636969789862633,
          "validation_metric": 0.00060031667817384,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7344946169996547
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.1774472389370203,
          "train_loss": 0.17205922305583954,
          "validation_loss": 0.021929100155830383,
          "train_metric": 0.046039313077926636,
          "validation_metric": 0.0001909505808725953,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7456083619999845
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.17248857207596302,
          "train_loss": 0.17401644587516785,
          "validation_loss": 0.021857287734746933,
          "train_metric": 0.04582095518708229,
          "validation_metric": 0.00035776683944277465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7562572269998782
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.17390790162608027,
          "train_loss": 0.16983821988105774,
          "validation_loss": 0.019566524773836136,
          "train_metric": 0.04598115012049675,
          "validation_metric": 0.00017775919695850462,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7676526279992686
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.17297281505307183,
          "train_loss": 0.17596158385276794,
          "validation_loss": 0.025495605543255806,
          "train_metric": 0.04588846117258072,
          "validation_metric": 0.00019678761600516737,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7783462229999714
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.17482607625424862,
          "train_loss": 0.17383788526058197,
          "validation_loss": 0.020016029477119446,
          "train_metric": 0.04643157869577408,
          "validation_metric": 0.00017984087753575295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7886768709995522
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.17440902534872293,
          "train_loss": 0.1790410429239273,
          "validation_loss": 0.025155549868941307,
          "train_metric": 0.04595557972788811,
          "validation_metric": 0.00036406225990504026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.799246194999796
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.18030031654052436,
          "train_loss": 0.17543578147888184,
          "validation_loss": 0.022952858358621597,
          "train_metric": 0.04580947011709213,
          "validation_metric": 0.00018099386943504214,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8098401129991544
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.17501553753390908,
          "train_loss": 0.17315541207790375,
          "validation_loss": 0.024557579308748245,
          "train_metric": 0.04579544812440872,
          "validation_metric": 0.00018948150682263076,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8203185969996412
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.17576697818003595,
          "train_loss": 0.1745883971452713,
          "validation_loss": 0.021413424983620644,
          "train_metric": 0.04653562977910042,
          "validation_metric": 0.00025006578653119504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.831140444999619
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.1766075538471341,
          "train_loss": 0.18165144324302673,
          "validation_loss": 0.02831798791885376,
          "train_metric": 0.04592331871390343,
          "validation_metric": 0.0005193338147364557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8415382179991866
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.17852599429897964,
          "train_loss": 0.1771792769432068,
          "validation_loss": 0.02338910475373268,
          "train_metric": 0.04660964012145996,
          "validation_metric": 0.00018878397531807423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8520509190002485
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.1743130257818848,
          "train_loss": 0.17163586616516113,
          "validation_loss": 0.02064710110425949,
          "train_metric": 0.04579804837703705,
          "validation_metric": 0.000274558988166973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8626576379992912
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.17300913948565722,
          "train_loss": 0.1720498502254486,
          "validation_loss": 0.01817944645881653,
          "train_metric": 0.04603791981935501,
          "validation_metric": 0.00022383764735423028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8732485780001298
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.1719330595806241,
          "train_loss": 0.17711521685123444,
          "validation_loss": 0.02181408926844597,
          "train_metric": 0.04612863063812256,
          "validation_metric": 0.00034039714955724776,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8837927149997995
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.1716181004885584,
          "train_loss": 0.17215661704540253,
          "validation_loss": 0.01961151510477066,
          "train_metric": 0.04618016257882118,
          "validation_metric": 0.00017206899065058678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8946432979992096
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.17377329850569367,
          "train_loss": 0.17059139907360077,
          "validation_loss": 0.020743828266859055,
          "train_metric": 0.04570351541042328,
          "validation_metric": 0.00020075826614629477,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.904675297999347
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.17931296536698937,
          "train_loss": 0.17524605989456177,
          "validation_loss": 0.021186793223023415,
          "train_metric": 0.04627763107419014,
          "validation_metric": 0.0001722044253256172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9151069540002936
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.17499706184025854,
          "train_loss": 0.17422379553318024,
          "validation_loss": 0.02383853867650032,
          "train_metric": 0.045660652220249176,
          "validation_metric": 0.0001824225764721632,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.92551519799963
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.1719327389728278,
          "train_loss": 0.1714417189359665,
          "validation_loss": 0.020781615749001503,
          "train_metric": 0.045839644968509674,
          "validation_metric": 0.00016326623153872788,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9363750100001198
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.17675163131207228,
          "train_loss": 0.1794685423374176,
          "validation_loss": 0.02844197116792202,
          "train_metric": 0.04572272673249245,
          "validation_metric": 0.00019918070756830275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9467688989989256
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.17380222398787737,
          "train_loss": 0.1748346984386444,
          "validation_loss": 0.021782996132969856,
          "train_metric": 0.046200066804885864,
          "validation_metric": 0.00017305910296272486,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9567909849993157
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.1812941781245172,
          "train_loss": 0.19283699989318848,
          "validation_loss": 0.037766288965940475,
          "train_metric": 0.04646644368767738,
          "validation_metric": 0.0005430062301456928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9670225640002172
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.18526205769740045,
          "train_loss": 0.18037499487400055,
          "validation_loss": 0.021394990384578705,
          "train_metric": 0.0467502661049366,
          "validation_metric": 0.0002059886755887419,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9774150220000593
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.17853111308068037,
          "train_loss": 0.17281851172447205,
          "validation_loss": 0.021597670391201973,
          "train_metric": 0.0458337776362896,
          "validation_metric": 0.00015195593005046248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.988053351000417
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.17563258530572057,
          "train_loss": 0.17679336667060852,
          "validation_loss": 0.02046404778957367,
          "train_metric": 0.046249374747276306,
          "validation_metric": 0.00017303841013927013,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9987319600004412
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.17191968811675906,
          "train_loss": 0.17539922893047333,
          "validation_loss": 0.019272208213806152,
          "train_metric": 0.04605020582675934,
          "validation_metric": 0.00024097329878713936,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0091407370000525
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.174180765170604,
          "train_loss": 0.17279985547065735,
          "validation_loss": 0.01832282356917858,
          "train_metric": 0.04584266617894173,
          "validation_metric": 0.00017812031728681177,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0194564220000757
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.17325618816539645,
          "train_loss": 0.1679355949163437,
          "validation_loss": 0.014036828652024269,
          "train_metric": 0.04597832262516022,
          "validation_metric": 0.0001831494300859049,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.029880800999308
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.1707703418796882,
          "train_loss": 0.17185702919960022,
          "validation_loss": 0.020444108173251152,
          "train_metric": 0.045924797654151917,
          "validation_metric": 0.00014219297736417502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0404699280006753
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.17679552268236876,
          "train_loss": 0.17703726887702942,
          "validation_loss": 0.02192036807537079,
          "train_metric": 0.046457745134830475,
          "validation_metric": 0.00015872198855504394,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.050993646000279
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.17190238274633884,
          "train_loss": 0.17108316719532013,
          "validation_loss": 0.02140193060040474,
          "train_metric": 0.04568861052393913,
          "validation_metric": 0.00017027260037139058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.061556434000522
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.17473557218909264,
          "train_loss": 0.1741541177034378,
          "validation_loss": 0.019459588453173637,
          "train_metric": 0.0464538112282753,
          "validation_metric": 0.0001403448695782572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0721508229999017
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.1756809069775045,
          "train_loss": 0.17493227124214172,
          "validation_loss": 0.025451015681028366,
          "train_metric": 0.04577106609940529,
          "validation_metric": 0.00019257907115388662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0826769119994424
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.17725541698746383,
          "train_loss": 0.17711597681045532,
          "validation_loss": 0.02568616159260273,
          "train_metric": 0.0467073991894722,
          "validation_metric": 0.0002942392020486295,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0926420750001853
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.17592899780720472,
          "train_loss": 0.17297612130641937,
          "validation_loss": 0.018609628081321716,
          "train_metric": 0.04611554741859436,
          "validation_metric": 0.00015770544996485114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1028045479997672
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.17184310406446457,
          "train_loss": 0.18238167464733124,
          "validation_loss": 0.029892073944211006,
          "train_metric": 0.04579778388142586,
          "validation_metric": 0.0005348895210772753,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.113102142999196
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.18443874828517437,
          "train_loss": 0.17577941715717316,
          "validation_loss": 0.02211681939661503,
          "train_metric": 0.04605143889784813,
          "validation_metric": 0.000239177854382433,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1232602229993063
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.1763279251754284,
          "train_loss": 0.17272087931632996,
          "validation_loss": 0.02049512043595314,
          "train_metric": 0.04605776444077492,
          "validation_metric": 0.00014537479728460312,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1336902339990047
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.18040405865758657,
          "train_loss": 0.1771344095468521,
          "validation_loss": 0.025001324713230133,
          "train_metric": 0.04640527442097664,
          "validation_metric": 0.00022124974930193275,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.143991974999153
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.17995937098748982,
          "train_loss": 0.18041639029979706,
          "validation_loss": 0.02646465413272381,
          "train_metric": 0.045691292732954025,
          "validation_metric": 0.0003983660426456481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1540144169994164
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.17440139781683683,
          "train_loss": 0.17022360861301422,
          "validation_loss": 0.016872957348823547,
          "train_metric": 0.04628705978393555,
          "validation_metric": 0.00013033680443186313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1648699050001596
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.1737856671679765,
          "train_loss": 0.16974987089633942,
          "validation_loss": 0.019941112026572227,
          "train_metric": 0.0458516888320446,
          "validation_metric": 0.00015240501670632511,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.175434946000678
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.17526126629672945,
          "train_loss": 0.1780283898115158,
          "validation_loss": 0.02275076135993004,
          "train_metric": 0.04631606489419937,
          "validation_metric": 0.00017452023166697472,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.185850710999148
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.17417097371071577,
          "train_loss": 0.1712881177663803,
          "validation_loss": 0.019727393984794617,
          "train_metric": 0.046020008623600006,
          "validation_metric": 0.00012675518519245088,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1965615899989643
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.1728018547873944,
          "train_loss": 0.1690632700920105,
          "validation_loss": 0.017973659560084343,
          "train_metric": 0.045721959322690964,
          "validation_metric": 0.00017816883337218314,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.206681858999218
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.17106289672665298,
          "train_loss": 0.1728588044643402,
          "validation_loss": 0.018879476934671402,
          "train_metric": 0.046325843781232834,
          "validation_metric": 0.00013645520084537566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.216908451000563
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.17970769945532084,
          "train_loss": 0.18074899911880493,
          "validation_loss": 0.02549842745065689,
          "train_metric": 0.04616092890501022,
          "validation_metric": 0.00035123422276228666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.227168112000072
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.18272819463163614,
          "train_loss": 0.17968860268592834,
          "validation_loss": 0.02982119843363762,
          "train_metric": 0.04588013142347336,
          "validation_metric": 0.00016357468848582357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2371683199999097
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.17704202979803085,
          "train_loss": 0.18252931535243988,
          "validation_loss": 0.0251657385379076,
          "train_metric": 0.04686876758933067,
          "validation_metric": 0.00021483353339135647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.247500928999216
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.17687420919537544,
          "train_loss": 0.17218315601348877,
          "validation_loss": 0.01773699000477791,
          "train_metric": 0.04646854102611542,
          "validation_metric": 0.00012999346654396504,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2575177059989073
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.1750293313525617,
          "train_loss": 0.17987504601478577,
          "validation_loss": 0.031048282980918884,
          "train_metric": 0.04526415467262268,
          "validation_metric": 0.00021717917115893215,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.267718717999742
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.17838219134137034,
          "train_loss": 0.1770324409008026,
          "validation_loss": 0.021035490557551384,
          "train_metric": 0.046628206968307495,
          "validation_metric": 0.00013588680303655565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.278065438998965
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.17936175223439932,
          "train_loss": 0.18962594866752625,
          "validation_loss": 0.03761548921465874,
          "train_metric": 0.0461854413151741,
          "validation_metric": 0.0008334656013175845,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.288114537999718
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.1838164433138445,
          "train_loss": 0.17644181847572327,
          "validation_loss": 0.02168370969593525,
          "train_metric": 0.046289730817079544,
          "validation_metric": 0.00019829483062494546,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2984996889990725
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.17333912709727883,
          "train_loss": 0.17046847939491272,
          "validation_loss": 0.01711377128958702,
          "train_metric": 0.04576709121465683,
          "validation_metric": 0.00012644070375245064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3092287220006256
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.17300240078475326,
          "train_loss": 0.17240485548973083,
          "validation_loss": 0.019516589120030403,
          "train_metric": 0.04622959718108177,
          "validation_metric": 0.00013939804921392351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.319613584999388
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.17278520087711513,
          "train_loss": 0.16804930567741394,
          "validation_loss": 0.015806401148438454,
          "train_metric": 0.04618014395236969,
          "validation_metric": 0.00011582578008528799,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3306753269989713
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.16938022756949067,
          "train_loss": 0.170451357960701,
          "validation_loss": 0.019936971366405487,
          "train_metric": 0.04577171429991722,
          "validation_metric": 0.00013621660764329135,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3411721190004755
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.16981621412560344,
          "train_loss": 0.17299683392047882,
          "validation_loss": 0.019692901521921158,
          "train_metric": 0.04578583315014839,
          "validation_metric": 0.00022195558995008469,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3515717739992397
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.1723293410614133,
          "train_loss": 0.16979457437992096,
          "validation_loss": 0.01604301854968071,
          "train_metric": 0.04603325203061104,
          "validation_metric": 0.0001728119677864015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3619611029989755
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.1683110105805099,
          "train_loss": 0.17258089780807495,
          "validation_loss": 0.020093543455004692,
          "train_metric": 0.046441853046417236,
          "validation_metric": 0.00015486497431993484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3720641899999464
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.17064533079974353,
          "train_loss": 0.16738051176071167,
          "validation_loss": 0.013460020534694195,
          "train_metric": 0.04603172838687897,
          "validation_metric": 0.00014553127402905375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3823137109993695
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.16955772740766406,
          "train_loss": 0.17067568004131317,
          "validation_loss": 0.017567327246069908,
          "train_metric": 0.045786887407302856,
          "validation_metric": 0.0001498060446465388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3927756729990506
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.17439955892041326,
          "train_loss": 0.18672454357147217,
          "validation_loss": 0.03239935263991356,
          "train_metric": 0.04708629474043846,
          "validation_metric": 0.00043483884655870497,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.40323306999926
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.17879456048831344,
          "train_loss": 0.17153537273406982,
          "validation_loss": 0.018093077465891838,
          "train_metric": 0.04618304967880249,
          "validation_metric": 0.00014929348253645003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4135325730003387
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.17047580890357494,
          "train_loss": 0.17011365294456482,
          "validation_loss": 0.01608801633119583,
          "train_metric": 0.046342384070158005,
          "validation_metric": 0.00011170306243002415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4241321419995074
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.1713050256948918,
          "train_loss": 0.17064058780670166,
          "validation_loss": 0.016485512256622314,
          "train_metric": 0.045891761779785156,
          "validation_metric": 0.00024375130305998027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.434596683999189
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.17119976540561765,
          "train_loss": 0.17076046764850616,
          "validation_loss": 0.018268197774887085,
          "train_metric": 0.04587741568684578,
          "validation_metric": 0.000272526522167027,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4447640359994693
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.17030131164938211,
          "train_loss": 0.17234402894973755,
          "validation_loss": 0.01864190399646759,
          "train_metric": 0.04627007991075516,
          "validation_metric": 0.00012787111336365342,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4545809289993485
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.1759753441438079,
          "train_loss": 0.17499373853206635,
          "validation_loss": 0.020816287025809288,
          "train_metric": 0.04636292904615402,
          "validation_metric": 0.00012507544306572527,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.464778932000627
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.17455246578902006,
          "train_loss": 0.17695067822933197,
          "validation_loss": 0.022950923070311546,
          "train_metric": 0.04587004706263542,
          "validation_metric": 0.00015941314632073045,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.474973565000255
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.1753339935094118,
          "train_loss": 0.1764882355928421,
          "validation_loss": 0.022707883268594742,
          "train_metric": 0.045993078500032425,
          "validation_metric": 0.00015758855442982167,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.485213656000269
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.1782651492394507,
          "train_loss": 0.16513125598430634,
          "validation_loss": 0.01318000815808773,
          "train_metric": 0.045873887836933136,
          "validation_metric": 0.0001406912342645228,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4955946509999194
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.1707252247724682,
          "train_loss": 0.16601988673210144,
          "validation_loss": 0.01500315684825182,
          "train_metric": 0.04594866558909416,
          "validation_metric": 0.00011627800267888233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5056597019993205
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.1688995385193266,
          "train_loss": 0.16909855604171753,
          "validation_loss": 0.013808224350214005,
          "train_metric": 0.04587370902299881,
          "validation_metric": 0.00022333115339279175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5161093839997193
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.17688853025902063,
          "train_loss": 0.17435286939144135,
          "validation_loss": 0.023067370057106018,
          "train_metric": 0.0456635020673275,
          "validation_metric": 0.00035323353949934244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5262461969996366
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.1777124828658998,
          "train_loss": 0.19680145382881165,
          "validation_loss": 0.04760411009192467,
          "train_metric": 0.04493303596973419,
          "validation_metric": 0.0004429970867931843,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.536326363999251
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.1851313286460936,
          "train_loss": 0.17825202643871307,
          "validation_loss": 0.025157395750284195,
          "train_metric": 0.045652296394109726,
          "validation_metric": 0.00021767719590570778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5465977909989306
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.17223667027428746,
          "train_loss": 0.17140893638134003,
          "validation_loss": 0.018523134291172028,
          "train_metric": 0.04579917713999748,
          "validation_metric": 0.00011783462832681835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.556503445999624
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.17196841700933874,
          "train_loss": 0.1672191023826599,
          "validation_loss": 0.015926921740174294,
          "train_metric": 0.046090755611658096,
          "validation_metric": 9.838310506893322e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5672605609997845
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.17034576158039272,
          "train_loss": 0.16913051903247833,
          "validation_loss": 0.015494954772293568,
          "train_metric": 0.046122584491968155,
          "validation_metric": 9.294209303334355e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5782099520001793
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.17296522902324796,
          "train_loss": 0.17053204774856567,
          "validation_loss": 0.016564389690756798,
          "train_metric": 0.04631327465176582,
          "validation_metric": 9.601532656233758e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5882778969989886
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.17187194107100368,
          "train_loss": 0.18388456106185913,
          "validation_loss": 0.034553226083517075,
          "train_metric": 0.04536711797118187,
          "validation_metric": 0.00047923417878337204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5984757559999707
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.17497961269691586,
          "train_loss": 0.16539257764816284,
          "validation_loss": 0.011112421751022339,
          "train_metric": 0.04592819884419441,
          "validation_metric": 0.00011674447159748524,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.608377078999183
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.16817152965813875,
          "train_loss": 0.17492550611495972,
          "validation_loss": 0.021499410271644592,
          "train_metric": 0.04675732180476189,
          "validation_metric": 0.00024680752540007234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6183599369996955
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.17068258672952652,
          "train_loss": 0.17429353296756744,
          "validation_loss": 0.022804226726293564,
          "train_metric": 0.04648873582482338,
          "validation_metric": 0.00022830237867310643,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6288761630003137
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.17053289618343115,
          "train_loss": 0.16867241263389587,
          "validation_loss": 0.015573480166494846,
          "train_metric": 0.04606436565518379,
          "validation_metric": 8.992197399493307e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6392924659994605
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.16691717796493322,
          "train_loss": 0.1722138673067093,
          "validation_loss": 0.01764606684446335,
          "train_metric": 0.04637167602777481,
          "validation_metric": 0.000123190984595567,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6497304610002175
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.1719184944522567,
          "train_loss": 0.17460142076015472,
          "validation_loss": 0.02333873324096203,
          "train_metric": 0.04610001668334007,
          "validation_metric": 0.0001457614271203056,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6599732380000205
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.17179162497632205,
          "train_loss": 0.16899217665195465,
          "validation_loss": 0.0185379758477211,
          "train_metric": 0.04572397843003273,
          "validation_metric": 0.00017502244736533612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6701141960002133
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.179447284899652,
          "train_loss": 0.17288078367710114,
          "validation_loss": 0.021340293809771538,
          "train_metric": 0.045997969806194305,
          "validation_metric": 0.00011682840704452246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6804483969990542
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.17098367679864168,
          "train_loss": 0.17820318043231964,
          "validation_loss": 0.025069531053304672,
          "train_metric": 0.04684451222419739,
          "validation_metric": 0.00026829008129425347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.690310034999129
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.17524893209338188,
          "train_loss": 0.16937997937202454,
          "validation_loss": 0.017036518082022667,
          "train_metric": 0.04594828188419342,
          "validation_metric": 0.00018711778102442622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.700312336999559
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.17306714272126555,
          "train_loss": 0.17276933789253235,
          "validation_loss": 0.01966182328760624,
          "train_metric": 0.04560736566781998,
          "validation_metric": 0.00020668725483119488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7101774219991057
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.16955652832984924,
          "train_loss": 0.16698673367500305,
          "validation_loss": 0.01505013182759285,
          "train_metric": 0.045942239463329315,
          "validation_metric": 9.035396942635998e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7198097790005704
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.1694710119627416,
          "train_loss": 0.17200517654418945,
          "validation_loss": 0.018079865723848343,
          "train_metric": 0.0462058000266552,
          "validation_metric": 0.00015447897021658719,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7301604019994556
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.17372805532068014,
          "train_loss": 0.1747300624847412,
          "validation_loss": 0.024653060361742973,
          "train_metric": 0.04569672420620918,
          "validation_metric": 0.00030701514333486557,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8053507360000367
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.1709246332757175,
          "train_loss": 0.1648535132408142,
          "validation_loss": 0.012605728581547737,
          "train_metric": 0.04581702873110771,
          "validation_metric": 0.00013326923362910748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8158181569997396
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.17084573674947023,
          "train_loss": 0.1788715273141861,
          "validation_loss": 0.028050720691680908,
          "train_metric": 0.04588440805673599,
          "validation_metric": 0.0001450411946279928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.825654346999727
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.1705125046428293,
          "train_loss": 0.16737130284309387,
          "validation_loss": 0.015468803234398365,
          "train_metric": 0.04573269188404083,
          "validation_metric": 0.00013246791786514223,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8353470239999297
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.16999630071222782,
          "train_loss": 0.17522650957107544,
          "validation_loss": 0.023733356967568398,
          "train_metric": 0.04602641239762306,
          "validation_metric": 0.00013494947052095085,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8452205449993926
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.16898890887387097,
          "train_loss": 0.16870594024658203,
          "validation_loss": 0.014870043843984604,
          "train_metric": 0.04574260488152504,
          "validation_metric": 0.00014353086589835584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8548506639999687
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.16967230383306742,
          "train_loss": 0.16745606064796448,
          "validation_loss": 0.015813885256648064,
          "train_metric": 0.04570147022604942,
          "validation_metric": 0.0001968606375157833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8647246079999604
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.168359768926166,
          "train_loss": 0.16657079756259918,
          "validation_loss": 0.013513174839317799,
          "train_metric": 0.04586460813879967,
          "validation_metric": 0.000145283222082071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.874477776000276
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.17107082484290004,
          "train_loss": 0.1723281592130661,
          "validation_loss": 0.01814320869743824,
          "train_metric": 0.04617389291524887,
          "validation_metric": 0.0001359807647531852,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8841478710000956
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.16986712394282222,
          "train_loss": 0.18721464276313782,
          "validation_loss": 0.03528687357902527,
          "train_metric": 0.045846130698919296,
          "validation_metric": 0.0005434934864751995,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.893980068000019
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.17115590488538146,
          "train_loss": 0.16860495507717133,
          "validation_loss": 0.014010758139193058,
          "train_metric": 0.045960959047079086,
          "validation_metric": 9.215993486577645e-05,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9035927200002334
        }
      ],
      "best": {
        "epoch": 193,
        "updates": 3088
      },
      "best_validation_metric": 8.992197399493307e-05,
      "config": {
        "label": "MAE + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mae",
        "outliers": true,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.009482719757270752
    },
    "772fcca85495_22": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.12911630840972066,
          "train_loss": 0.08767668902873993,
          "validation_loss": 0.046990957111120224,
          "train_metric": 0.07658258825540543,
          "validation_metric": 0.016934528946876526,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.01806475000012142
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.06999778258614242,
          "train_loss": 0.05867474153637886,
          "validation_loss": 0.028861545026302338,
          "train_metric": 0.06042882055044174,
          "validation_metric": 0.015319924801588058,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.035511426998709794
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.056610019179061055,
          "train_loss": 0.05326199159026146,
          "validation_loss": 0.024446509778499603,
          "train_metric": 0.05621645227074623,
          "validation_metric": 0.011523162946105003,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.053447398999196594
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.05250030546449125,
          "train_loss": 0.04952489212155342,
          "validation_loss": 0.020355287939310074,
          "train_metric": 0.05510217696428299,
          "validation_metric": 0.010421463288366795,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07104362999962177
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.048899700632318854,
          "train_loss": 0.047476425766944885,
          "validation_loss": 0.017766518518328667,
          "train_metric": 0.053195323795080185,
          "validation_metric": 0.008712165988981724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.08910192399889638
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.04611057264264673,
          "train_loss": 0.0449950136244297,
          "validation_loss": 0.015970399603247643,
          "train_metric": 0.05181830748915672,
          "validation_metric": 0.007313790265470743,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.10733068499939691
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.0448259609984234,
          "train_loss": 0.04454144835472107,
          "validation_loss": 0.015505759976804256,
          "train_metric": 0.05101657658815384,
          "validation_metric": 0.006611390504986048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.12515283499851648
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.044414358329959214,
          "train_loss": 0.04412376135587692,
          "validation_loss": 0.014681114815175533,
          "train_metric": 0.050914715975522995,
          "validation_metric": 0.0060462043620646,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.14324636199853558
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.04387735022464767,
          "train_loss": 0.04339509829878807,
          "validation_loss": 0.01432228647172451,
          "train_metric": 0.05075183883309364,
          "validation_metric": 0.005857511423528194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.16123154999877443
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.043728973250836134,
          "train_loss": 0.04313630610704422,
          "validation_loss": 0.014261090196669102,
          "train_metric": 0.05089719593524933,
          "validation_metric": 0.005559201817959547,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.17923190299916314
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.04294410959118977,
          "train_loss": 0.04207830876111984,
          "validation_loss": 0.012822195887565613,
          "train_metric": 0.050204087048769,
          "validation_metric": 0.0052221473306417465,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19646433900015836
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.04200495244003832,
          "train_loss": 0.041693877428770065,
          "validation_loss": 0.012471681460738182,
          "train_metric": 0.05037427693605423,
          "validation_metric": 0.004549980629235506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21351471199886873
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.04125866130925715,
          "train_loss": 0.04045838490128517,
          "validation_loss": 0.010852075181901455,
          "train_metric": 0.04980258271098137,
          "validation_metric": 0.004247887060046196,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.23035528099899238
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.04028061346616596,
          "train_loss": 0.03933114930987358,
          "validation_loss": 0.010311299934983253,
          "train_metric": 0.04897600784897804,
          "validation_metric": 0.004018560517579317,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24824557200008712
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.03941143170231953,
          "train_loss": 0.03908509761095047,
          "validation_loss": 0.008904479444026947,
          "train_metric": 0.0492522269487381,
          "validation_metric": 0.002925535896793008,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2655101949985692
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.038154285750351846,
          "train_loss": 0.03747904673218727,
          "validation_loss": 0.008371292613446712,
          "train_metric": 0.04797130078077316,
          "validation_metric": 0.0031997293699532747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.28162908800004516
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03746523498557508,
          "train_loss": 0.037121888250112534,
          "validation_loss": 0.007643993478268385,
          "train_metric": 0.04749250039458275,
          "validation_metric": 0.002333494136109948,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29828809999889927
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.03675688162911683,
          "train_loss": 0.03749716654419899,
          "validation_loss": 0.008958278223872185,
          "train_metric": 0.04698322340846062,
          "validation_metric": 0.003229757770895958,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3146835789993929
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.03627527928620111,
          "train_loss": 0.03560218960046768,
          "validation_loss": 0.006149096880108118,
          "train_metric": 0.047250621020793915,
          "validation_metric": 0.0020047842990607023,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.33158614300009503
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.03554004663601518,
          "train_loss": 0.03550124168395996,
          "validation_loss": 0.006265485659241676,
          "train_metric": 0.04733172804117203,
          "validation_metric": 0.0020994378719478846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34809903699897404
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.035878154973033816,
          "train_loss": 0.03623664006590843,
          "validation_loss": 0.006188511848449707,
          "train_metric": 0.048408765345811844,
          "validation_metric": 0.0016262533608824015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36557119999997667
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.03554828168125823,
          "train_loss": 0.03591229021549225,
          "validation_loss": 0.006776048336178064,
          "train_metric": 0.046779632568359375,
          "validation_metric": 0.0024156104773283005,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3824137569990853
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.03564702032599598,
          "train_loss": 0.03467928618192673,
          "validation_loss": 0.005225707311183214,
          "train_metric": 0.04716968163847923,
          "validation_metric": 0.0016991200391203165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3991272859984747
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.0351354424492456,
          "train_loss": 0.03444818779826164,
          "validation_loss": 0.005253870040178299,
          "train_metric": 0.04691446200013161,
          "validation_metric": 0.00158164300955832,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41674471399892354
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.035371451507671736,
          "train_loss": 0.03482343629002571,
          "validation_loss": 0.005418855231255293,
          "train_metric": 0.04702942445874214,
          "validation_metric": 0.001392899896018207,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5346622809993278
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.03444203967228532,
          "train_loss": 0.034063566476106644,
          "validation_loss": 0.004859456792473793,
          "train_metric": 0.04653055593371391,
          "validation_metric": 0.0013210978358983994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5529605880001327
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.03413718269439414,
          "train_loss": 0.0337958000600338,
          "validation_loss": 0.004456627182662487,
          "train_metric": 0.046687476336956024,
          "validation_metric": 0.0012853537918999791,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5705179849992419
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.03399913618341088,
          "train_loss": 0.03394312039017677,
          "validation_loss": 0.004468646831810474,
          "train_metric": 0.046917758882045746,
          "validation_metric": 0.0011450785677880049,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5879413629991177
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.03424958715913817,
          "train_loss": 0.03408296778798103,
          "validation_loss": 0.0044759660959243774,
          "train_metric": 0.04688164219260216,
          "validation_metric": 0.0014603828312829137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6041554859984899
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.034450050559826195,
          "train_loss": 0.036654282361269,
          "validation_loss": 0.006319101899862289,
          "train_metric": 0.048418231308460236,
          "validation_metric": 0.0014929476892575622,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6208964159995958
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.03470987276523374,
          "train_loss": 0.036719176918268204,
          "validation_loss": 0.007682063151150942,
          "train_metric": 0.046041250228881836,
          "validation_metric": 0.0024052890948951244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6372418079990894
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.03416135811130516,
          "train_loss": 0.03472663089632988,
          "validation_loss": 0.0045318566262722015,
          "train_metric": 0.048043545335531235,
          "validation_metric": 0.001059197704307735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6543256109998765
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.03476867044810206,
          "train_loss": 0.03356264531612396,
          "validation_loss": 0.004182079806923866,
          "train_metric": 0.0466458685696125,
          "validation_metric": 0.0010010229889303446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6712634559989965
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.033521216129884124,
          "train_loss": 0.03375759720802307,
          "validation_loss": 0.004704159218817949,
          "train_metric": 0.04568792134523392,
          "validation_metric": 0.0012184405932202935,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6874578729984933
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.033402438624762,
          "train_loss": 0.03374005854129791,
          "validation_loss": 0.0037070466205477715,
          "train_metric": 0.04747898876667023,
          "validation_metric": 0.0010431158589199185,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7033319539987133
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.03409045486478135,
          "train_loss": 0.03429950028657913,
          "validation_loss": 0.004932910203933716,
          "train_metric": 0.0461927130818367,
          "validation_metric": 0.0015676249749958515,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7194617059994926
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.033319865964585915,
          "train_loss": 0.033392295241355896,
          "validation_loss": 0.0037083206698298454,
          "train_metric": 0.0471649132668972,
          "validation_metric": 0.0008369127754122019,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.736506590999852
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.03313243872253224,
          "train_loss": 0.033474113792181015,
          "validation_loss": 0.004264386370778084,
          "train_metric": 0.04602501541376114,
          "validation_metric": 0.001221632119268179,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7528979679991608
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.03336117847356945,
          "train_loss": 0.033170826733112335,
          "validation_loss": 0.0034564717207103968,
          "train_metric": 0.04664596915245056,
          "validation_metric": 0.0008086083689704537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7698443439985567
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.0328718182281591,
          "train_loss": 0.032585568726062775,
          "validation_loss": 0.0030232425779104233,
          "train_metric": 0.046689461916685104,
          "validation_metric": 0.0007076227921061218,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7869032799990237
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.032558599254116416,
          "train_loss": 0.03253296762704849,
          "validation_loss": 0.0029996507801115513,
          "train_metric": 0.04654906690120697,
          "validation_metric": 0.000691343389917165,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8038263629987341
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.032708884798921645,
          "train_loss": 0.032642412930727005,
          "validation_loss": 0.003018439980223775,
          "train_metric": 0.04618598520755768,
          "validation_metric": 0.0007097255438566208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.820310110999344
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.032417875714600086,
          "train_loss": 0.03260992839932442,
          "validation_loss": 0.003062897827476263,
          "train_metric": 0.04667239636182785,
          "validation_metric": 0.000666824693325907,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8372116779992211
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.03302342363167554,
          "train_loss": 0.03280799090862274,
          "validation_loss": 0.003796074539422989,
          "train_metric": 0.04558521881699562,
          "validation_metric": 0.001023121876642108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8535727959988435
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.03281133825657889,
          "train_loss": 0.031933631747961044,
          "validation_loss": 0.0024150742683559656,
          "train_metric": 0.04643174633383751,
          "validation_metric": 0.0005690279649570584,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8703098149999278
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.032010317081585526,
          "train_loss": 0.032044894993305206,
          "validation_loss": 0.0027432686183601618,
          "train_metric": 0.04584489390254021,
          "validation_metric": 0.0006796579109504819,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8866698109995923
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.03207013296196237,
          "train_loss": 0.03227918595075607,
          "validation_loss": 0.002465946599841118,
          "train_metric": 0.04681882634758949,
          "validation_metric": 0.0005606836057268083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9036959949989978
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.032374700356740505,
          "train_loss": 0.03200826421380043,
          "validation_loss": 0.0027917721308767796,
          "train_metric": 0.0454188734292984,
          "validation_metric": 0.0006603447836823761,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9203171270000894
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.032321112928912044,
          "train_loss": 0.03222816437482834,
          "validation_loss": 0.002678859978914261,
          "train_metric": 0.04622502624988556,
          "validation_metric": 0.0007236434030346572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9361730759992497
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.032019175530876964,
          "train_loss": 0.03159394487738609,
          "validation_loss": 0.00226251850835979,
          "train_metric": 0.04578602313995361,
          "validation_metric": 0.0005296170711517334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9528687139991234
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.03174719511298463,
          "train_loss": 0.03214666247367859,
          "validation_loss": 0.0025420270394533873,
          "train_metric": 0.04685073718428612,
          "validation_metric": 0.0005355621688067913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.969061685998895
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.03245711357158143,
          "train_loss": 0.03175269067287445,
          "validation_loss": 0.00261586531996727,
          "train_metric": 0.04536685720086098,
          "validation_metric": 0.0006215464090928435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9856644430001325
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.03232198421028443,
          "train_loss": 0.03202081471681595,
          "validation_loss": 0.0023775489535182714,
          "train_metric": 0.046654582023620605,
          "validation_metric": 0.0004991220775991678,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0026259269998263
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.03220660501392558,
          "train_loss": 0.03182138502597809,
          "validation_loss": 0.002256380394101143,
          "train_metric": 0.04660506173968315,
          "validation_metric": 0.00047280604485422373,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0190522619996045
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.03168950595500064,
          "train_loss": 0.03140084818005562,
          "validation_loss": 0.0018495448166504502,
          "train_metric": 0.04615635424852371,
          "validation_metric": 0.0004408792301546782,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0352440789993125
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.03162399656139314,
          "train_loss": 0.03201337903738022,
          "validation_loss": 0.002437299583107233,
          "train_metric": 0.04654153063893318,
          "validation_metric": 0.0005094425869174302,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0507547329998488
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.03172153087507468,
          "train_loss": 0.03135430067777634,
          "validation_loss": 0.0019755864050239325,
          "train_metric": 0.04561593011021614,
          "validation_metric": 0.00048030095058493316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0661264619993744
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.0314573057112284,
          "train_loss": 0.03136066347360611,
          "validation_loss": 0.0017911514732986689,
          "train_metric": 0.04638756811618805,
          "validation_metric": 0.0003739756066352129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0825259670000378
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.031616854772437364,
          "train_loss": 0.031589437276124954,
          "validation_loss": 0.0022974619641900063,
          "train_metric": 0.04560419172048569,
          "validation_metric": 0.0005750443669967353,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0975751149999269
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.03188469578162767,
          "train_loss": 0.03135782480239868,
          "validation_loss": 0.0017406620318070054,
          "train_metric": 0.046435654163360596,
          "validation_metric": 0.00040063593769446015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1120383759989636
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.031498329946771264,
          "train_loss": 0.03134674206376076,
          "validation_loss": 0.001632989034987986,
          "train_metric": 0.04636938124895096,
          "validation_metric": 0.0003488290240056813,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2172218499999872
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.031548154045594856,
          "train_loss": 0.03109595738351345,
          "validation_loss": 0.0015333376359194517,
          "train_metric": 0.04628472775220871,
          "validation_metric": 0.0003278891381341964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.233145901998796
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.03120956913335249,
          "train_loss": 0.031109895557165146,
          "validation_loss": 0.0016126657137647271,
          "train_metric": 0.04582919552922249,
          "validation_metric": 0.00037471839459612966,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2476043209990166
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.031321622896939516,
          "train_loss": 0.03124009259045124,
          "validation_loss": 0.001766334054991603,
          "train_metric": 0.04595046490430832,
          "validation_metric": 0.0003628192935138941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2618680309988122
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.03133977758261608,
          "train_loss": 0.03105790540575981,
          "validation_loss": 0.001584451412782073,
          "train_metric": 0.045977603644132614,
          "validation_metric": 0.0003348740574438125,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2760203029993136
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.03191507305018604,
          "train_loss": 0.03294133394956589,
          "validation_loss": 0.0036385483108460903,
          "train_metric": 0.04487290233373642,
          "validation_metric": 0.0009018019773066044,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2905699299990374
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.032281403662636876,
          "train_loss": 0.03196560591459274,
          "validation_loss": 0.001710241544060409,
          "train_metric": 0.047069963067770004,
          "validation_metric": 0.00036135566188022494,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3050393809990055
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.03189369768369943,
          "train_loss": 0.031508274376392365,
          "validation_loss": 0.002062973566353321,
          "train_metric": 0.04547242447733879,
          "validation_metric": 0.00042726838728412986,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3199188049984514
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.031182031467324123,
          "train_loss": 0.031311631202697754,
          "validation_loss": 0.0018646932439878583,
          "train_metric": 0.046037521213293076,
          "validation_metric": 0.0004428868414834142,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3344122660000721
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.031171624432317913,
          "train_loss": 0.03096613846719265,
          "validation_loss": 0.0016272200737148523,
          "train_metric": 0.045642562210559845,
          "validation_metric": 0.0003380440757609904,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.34858551599973
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.031231783344992436,
          "train_loss": 0.031160429120063782,
          "validation_loss": 0.0017676603747531772,
          "train_metric": 0.04592045024037361,
          "validation_metric": 0.0004196439986117184,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.362760153999261
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.031483845305047,
          "train_loss": 0.030830973759293556,
          "validation_loss": 0.001344233867712319,
          "train_metric": 0.04564930126070976,
          "validation_metric": 0.00029600440757349133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3782527589992242
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.031059253786224872,
          "train_loss": 0.031728021800518036,
          "validation_loss": 0.0023765719961375,
          "train_metric": 0.045409344136714935,
          "validation_metric": 0.0005922954878769815,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3927361070000188
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.03128785773878917,
          "train_loss": 0.031472887843847275,
          "validation_loss": 0.0018188306130468845,
          "train_metric": 0.046527713537216187,
          "validation_metric": 0.00037649943260475993,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.407089164998979
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.031296666245907545,
          "train_loss": 0.031058913096785545,
          "validation_loss": 0.0018089448567479849,
          "train_metric": 0.04578207805752754,
          "validation_metric": 0.0003935070417355746,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4214733139997406
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.03156402462627739,
          "train_loss": 0.031857606023550034,
          "validation_loss": 0.001979722874239087,
          "train_metric": 0.04720716178417206,
          "validation_metric": 0.000411856482969597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4361142859997926
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.0317364641232416,
          "train_loss": 0.030903300270438194,
          "validation_loss": 0.0013582278043031693,
          "train_metric": 0.04598679393529892,
          "validation_metric": 0.00027580649475567043,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4507304219987418
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.031028644534671912,
          "train_loss": 0.030806822702288628,
          "validation_loss": 0.001433486002497375,
          "train_metric": 0.045998167246580124,
          "validation_metric": 0.0003238724893890321,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4648377069988783
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.030784380047407467,
          "train_loss": 0.03087467886507511,
          "validation_loss": 0.001273266738280654,
          "train_metric": 0.04604528471827507,
          "validation_metric": 0.0002797235792968422,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4789089799996873
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.031050071462232154,
          "train_loss": 0.03124985098838806,
          "validation_loss": 0.0019052366260439157,
          "train_metric": 0.04543932527303696,
          "validation_metric": 0.00046091570402495563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4928961139994499
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.03121379663934931,
          "train_loss": 0.03220715373754501,
          "validation_loss": 0.002540621906518936,
          "train_metric": 0.04684273526072502,
          "validation_metric": 0.0005156203988008201,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.50667124499887
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.031712641794001684,
          "train_loss": 0.03202613815665245,
          "validation_loss": 0.002630672650411725,
          "train_metric": 0.04554757848381996,
          "validation_metric": 0.0006786136073060334,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.520691170999271
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.03153349864805932,
          "train_loss": 0.03168874979019165,
          "validation_loss": 0.001994140911847353,
          "train_metric": 0.04682048410177231,
          "validation_metric": 0.00041761467582546175,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5347298229989974
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.031221312761772424,
          "train_loss": 0.031183263286948204,
          "validation_loss": 0.0019177908543497324,
          "train_metric": 0.04589509963989258,
          "validation_metric": 0.0004473020962905139,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5490241979987331
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.030961272830609232,
          "train_loss": 0.03057127259671688,
          "validation_loss": 0.0011545040179044008,
          "train_metric": 0.04557345062494278,
          "validation_metric": 0.00024229615519288927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.563517195998429
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.031061466652317904,
          "train_loss": 0.031126052141189575,
          "validation_loss": 0.0014470128808170557,
          "train_metric": 0.04631946235895157,
          "validation_metric": 0.0003340269613545388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5777318549990014
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.031446696375496686,
          "train_loss": 0.030953915789723396,
          "validation_loss": 0.0013959113275632262,
          "train_metric": 0.04600468650460243,
          "validation_metric": 0.00028167947311885655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.591687780999564
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.03101347212214023,
          "train_loss": 0.03121880628168583,
          "validation_loss": 0.0015791055047884583,
          "train_metric": 0.046336691826581955,
          "validation_metric": 0.0003725386632140726,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.606251199998951
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.03133848482684698,
          "train_loss": 0.032240692526102066,
          "validation_loss": 0.003061760915443301,
          "train_metric": 0.04505973681807518,
          "validation_metric": 0.0007500756182707846,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6203263449988299
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.03191809443524107,
          "train_loss": 0.031110430136322975,
          "validation_loss": 0.0016156788915395737,
          "train_metric": 0.04570464789867401,
          "validation_metric": 0.00033012477797456086,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.634279002999392
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.030747585456992965,
          "train_loss": 0.030566001310944557,
          "validation_loss": 0.0010625261347740889,
          "train_metric": 0.045875199139118195,
          "validation_metric": 0.00021857884712517262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6485869529988122
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.03070953999122139,
          "train_loss": 0.03106525167822838,
          "validation_loss": 0.0018166527152061462,
          "train_metric": 0.04531204700469971,
          "validation_metric": 0.0004229334881529212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6627555340000981
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.030889013141859323,
          "train_loss": 0.030700016766786575,
          "validation_loss": 0.001037444220855832,
          "train_metric": 0.04633129760622978,
          "validation_metric": 0.00021539811859838665,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6773782119998941
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.030878989142365754,
          "train_loss": 0.03139805793762207,
          "validation_loss": 0.0016399974701926112,
          "train_metric": 0.04652607813477516,
          "validation_metric": 0.00033376881037838757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6915217909991043
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.03185032191686332,
          "train_loss": 0.03230845928192139,
          "validation_loss": 0.0024323221296072006,
          "train_metric": 0.04613839462399483,
          "validation_metric": 0.0006206597317941487,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7054025209999963
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.03172259806888178,
          "train_loss": 0.03125763311982155,
          "validation_loss": 0.0018158359453082085,
          "train_metric": 0.045164357870817184,
          "validation_metric": 0.00038609301554970443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7195460099992488
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.030994303815532476,
          "train_loss": 0.030925726518034935,
          "validation_loss": 0.0013387727085500956,
          "train_metric": 0.046773575246334076,
          "validation_metric": 0.0002711528504732996,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8182566529994801
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.030797955056186765,
          "train_loss": 0.031197648495435715,
          "validation_loss": 0.0019884679932147264,
          "train_metric": 0.04527238383889198,
          "validation_metric": 0.000443670607637614,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8329311460001918
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.03108839712513145,
          "train_loss": 0.030601132661104202,
          "validation_loss": 0.0010816958965733647,
          "train_metric": 0.04610767215490341,
          "validation_metric": 0.00023126295127440244,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8470829199995933
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.030912449525203556,
          "train_loss": 0.031101563945412636,
          "validation_loss": 0.0016960161738097668,
          "train_metric": 0.04536309093236923,
          "validation_metric": 0.0003956007130909711,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8611353950000193
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.031340940069640055,
          "train_loss": 0.03072122484445572,
          "validation_loss": 0.0012001399882137775,
          "train_metric": 0.045928243547677994,
          "validation_metric": 0.0002417869254713878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8751634499985812
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.030884071718901396,
          "train_loss": 0.030646519735455513,
          "validation_loss": 0.0010960312793031335,
          "train_metric": 0.04611920565366745,
          "validation_metric": 0.00022096687462180853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8897265649993642
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.03080138290533796,
          "train_loss": 0.030563481152057648,
          "validation_loss": 0.0011525456793606281,
          "train_metric": 0.04565383493900299,
          "validation_metric": 0.00023278284061234444,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9037402349986223
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.0307717879768461,
          "train_loss": 0.03085188753902912,
          "validation_loss": 0.0015868539921939373,
          "train_metric": 0.045660775154829025,
          "validation_metric": 0.0003602925280574709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9178431429991178
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.030672253342345357,
          "train_loss": 0.030584055930376053,
          "validation_loss": 0.0010540328221395612,
          "train_metric": 0.045897580683231354,
          "validation_metric": 0.00023068737937137485,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.931765560999338
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.030523111985530704,
          "train_loss": 0.030505308881402016,
          "validation_loss": 0.000996289774775505,
          "train_metric": 0.04606429859995842,
          "validation_metric": 0.00021440416458062828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9470928559985623
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.030702410789672285,
          "train_loss": 0.03071942925453186,
          "validation_loss": 0.0013084713136777282,
          "train_metric": 0.045586783438920975,
          "validation_metric": 0.00029253834509290755,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9613051420001284
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.03202916914597154,
          "train_loss": 0.03125829994678497,
          "validation_loss": 0.0012120703468099236,
          "train_metric": 0.04694615676999092,
          "validation_metric": 0.0002441576507408172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9754441939985554
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.03135517648115638,
          "train_loss": 0.03168981894850731,
          "validation_loss": 0.0024230685085058212,
          "train_metric": 0.04486427456140518,
          "validation_metric": 0.0005557927070185542,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9897830549998616
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.030981061281636357,
          "train_loss": 0.030863523483276367,
          "validation_loss": 0.0012259144568815827,
          "train_metric": 0.04624129831790924,
          "validation_metric": 0.00027829204918816686,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.003550744999302
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.030812787998002023,
          "train_loss": 0.030570197850465775,
          "validation_loss": 0.0009417639230377972,
          "train_metric": 0.045823827385902405,
          "validation_metric": 0.00020350630802568048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.017854963998616
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.03106947037304053,
          "train_loss": 0.03130011633038521,
          "validation_loss": 0.0016224368009716272,
          "train_metric": 0.046139784157276154,
          "validation_metric": 0.00032989055034704506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0320524039998418
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.031058031192515045,
          "train_loss": 0.03165915235877037,
          "validation_loss": 0.0017402393277734518,
          "train_metric": 0.04705720767378807,
          "validation_metric": 0.0003644276293925941,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0460218179996446
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.03177221559599275,
          "train_loss": 0.030437029898166656,
          "validation_loss": 0.0009479834116064012,
          "train_metric": 0.0458315834403038,
          "validation_metric": 0.0001980071101570502,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0607501809990936
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.030648142215795815,
          "train_loss": 0.03031994216144085,
          "validation_loss": 0.000875522440765053,
          "train_metric": 0.04581058397889137,
          "validation_metric": 0.0001806849759304896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0757353949993558
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.030655098031274974,
          "train_loss": 0.030622536316514015,
          "validation_loss": 0.0009367361199110746,
          "train_metric": 0.046265412122011185,
          "validation_metric": 0.00018868922779802233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.087437852998846
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.03093713754788041,
          "train_loss": 0.030562613159418106,
          "validation_loss": 0.0011947276070713997,
          "train_metric": 0.045315228402614594,
          "validation_metric": 0.00025216268841177225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.097511227999348
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.030813558725640178,
          "train_loss": 0.030513908714056015,
          "validation_loss": 0.0009357270319014788,
          "train_metric": 0.0463888980448246,
          "validation_metric": 0.00018893224478233606,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.107504985999185
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.03077106410637498,
          "train_loss": 0.030705565586686134,
          "validation_loss": 0.0014250301755964756,
          "train_metric": 0.045293763279914856,
          "validation_metric": 0.0003089294768869877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.11746501000016
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.030656508846732322,
          "train_loss": 0.030758384615182877,
          "validation_loss": 0.0011990577913820744,
          "train_metric": 0.045865949243307114,
          "validation_metric": 0.00026477922801859677,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.127615662999233
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.030721656687092036,
          "train_loss": 0.030748046934604645,
          "validation_loss": 0.001150264753960073,
          "train_metric": 0.04646850749850273,
          "validation_metric": 0.00023204553872346878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1372663599995576
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.030582683415559586,
          "train_loss": 0.030780356377363205,
          "validation_loss": 0.0012715707998722792,
          "train_metric": 0.04572465643286705,
          "validation_metric": 0.0002925816224887967,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.147253961999013
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.030806719048996456,
          "train_loss": 0.030352648347616196,
          "validation_loss": 0.0008856079657562077,
          "train_metric": 0.04581755772233009,
          "validation_metric": 0.00017839194333646446,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1573887069989723
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.030691110936459154,
          "train_loss": 0.030622761696577072,
          "validation_loss": 0.0009821632411330938,
          "train_metric": 0.04634220525622368,
          "validation_metric": 0.00019783976313192397,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1673420299994177
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.0305730949039571,
          "train_loss": 0.030516380444169044,
          "validation_loss": 0.0010869009420275688,
          "train_metric": 0.0455610491335392,
          "validation_metric": 0.00024221307830885053,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.177433466998991
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.030737612807570258,
          "train_loss": 0.030560242012143135,
          "validation_loss": 0.0009744261624291539,
          "train_metric": 0.046110399067401886,
          "validation_metric": 0.0001962850510608405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.187195107999287
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.030512973164150026,
          "train_loss": 0.030314825475215912,
          "validation_loss": 0.0007649562321603298,
          "train_metric": 0.04599124193191528,
          "validation_metric": 0.00015732213796582073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.197692381998422
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.030430320999585092,
          "train_loss": 0.03100508451461792,
          "validation_loss": 0.0017217538552358747,
          "train_metric": 0.045318059623241425,
          "validation_metric": 0.0003971804981119931,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.208004620999418
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.030514949758071452,
          "train_loss": 0.03076939284801483,
          "validation_loss": 0.0011240403400734067,
          "train_metric": 0.046133462339639664,
          "validation_metric": 0.00022648475714959204,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.218155212000056
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.031061355948622804,
          "train_loss": 0.03082832507789135,
          "validation_loss": 0.0014712514821439981,
          "train_metric": 0.04625341296195984,
          "validation_metric": 0.00031997397309169173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2282790779991046
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.030752373568248004,
          "train_loss": 0.03045700676739216,
          "validation_loss": 0.0009642778895795345,
          "train_metric": 0.045823533087968826,
          "validation_metric": 0.0001945395051734522,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2379720919998363
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.03062520787352696,
          "train_loss": 0.030674519017338753,
          "validation_loss": 0.0010866918601095676,
          "train_metric": 0.04601608216762543,
          "validation_metric": 0.00021889523486606777,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2479244009991817
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.03053387306863442,
          "train_loss": 0.03092455491423607,
          "validation_loss": 0.0016306077595800161,
          "train_metric": 0.045432403683662415,
          "validation_metric": 0.00033991047530435026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3248820190001425
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.030793921905569732,
          "train_loss": 0.032759491354227066,
          "validation_loss": 0.0027637388557195663,
          "train_metric": 0.04749514162540436,
          "validation_metric": 0.0006358306272886693,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.334915550000005
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.03103040950372815,
          "train_loss": 0.030657010152935982,
          "validation_loss": 0.0013543353416025639,
          "train_metric": 0.04547852277755737,
          "validation_metric": 0.0003007844788953662,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.344691084999795
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.03185267170192674,
          "train_loss": 0.03257407620549202,
          "validation_loss": 0.0029651785735040903,
          "train_metric": 0.04662923514842987,
          "validation_metric": 0.0007487752591259778,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3543295479994413
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.03148300541215576,
          "train_loss": 0.03141123428940773,
          "validation_loss": 0.00198030611500144,
          "train_metric": 0.04540364071726799,
          "validation_metric": 0.0003988983226008713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.364044357998864
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.031080423719686223,
          "train_loss": 0.030508296564221382,
          "validation_loss": 0.0012135725701227784,
          "train_metric": 0.0458073653280735,
          "validation_metric": 0.000257765845162794,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3736473849985487
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.03092511532304343,
          "train_loss": 0.030758285894989967,
          "validation_loss": 0.0012060884619131684,
          "train_metric": 0.046013232320547104,
          "validation_metric": 0.00024294560716953129,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3831314479994035
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.030638474694569595,
          "train_loss": 0.030336277559399605,
          "validation_loss": 0.0008230609819293022,
          "train_metric": 0.04621594399213791,
          "validation_metric": 0.00016608282749075443,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.392735184999765
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.030597651726566255,
          "train_loss": 0.03046422079205513,
          "validation_loss": 0.0010608521988615394,
          "train_metric": 0.045498788356781006,
          "validation_metric": 0.0002149341453332454,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4022319199993944
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.0308080100803636,
          "train_loss": 0.03102393075823784,
          "validation_loss": 0.001306207966990769,
          "train_metric": 0.04659934714436531,
          "validation_metric": 0.00026695043197833,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.41194166799869
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.030636894749477506,
          "train_loss": 0.030841004103422165,
          "validation_loss": 0.0012426524190232158,
          "train_metric": 0.04573148116469383,
          "validation_metric": 0.000283789326203987,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.421377364998989
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.03094259399586008,
          "train_loss": 0.03071431815624237,
          "validation_loss": 0.0012232549488544464,
          "train_metric": 0.04547963663935661,
          "validation_metric": 0.0002464035351295024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4309587579991785
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.03102347464300692,
          "train_loss": 0.031224949285387993,
          "validation_loss": 0.0019187094876542687,
          "train_metric": 0.04590792581439018,
          "validation_metric": 0.00043798680417239666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4405519219999405
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.030539139756001532,
          "train_loss": 0.0303975697606802,
          "validation_loss": 0.0009955159621313214,
          "train_metric": 0.04578521102666855,
          "validation_metric": 0.00021636135352309793,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.450141466999412
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.030373855319339782,
          "train_loss": 0.030424926429986954,
          "validation_loss": 0.0009077169233933091,
          "train_metric": 0.04589959606528282,
          "validation_metric": 0.0001828438398661092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4598550390001037
        }
      ],
      "best": {
        "epoch": 127,
        "updates": 2032
      },
      "best_validation_metric": 0.00015732213796582073,
      "config": {
        "label": "Huber + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "huber",
        "outliers": true,
        "max_updates": null
      },
      "seed": 22,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.012542812203242968
    },
    "772fcca85495_33": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.1428014962002635,
          "train_loss": 0.08189748972654343,
          "validation_loss": 0.050383005291223526,
          "train_metric": 0.06807444989681244,
          "validation_metric": 0.020728617906570435,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.11801246800132503
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.06822966574691236,
          "train_loss": 0.0620587021112442,
          "validation_loss": 0.030782317742705345,
          "train_metric": 0.058830875903367996,
          "validation_metric": 0.01216837391257286,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.13635768700078188
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.055032467003911734,
          "train_loss": 0.050930529832839966,
          "validation_loss": 0.021589575335383415,
          "train_metric": 0.05575462803244591,
          "validation_metric": 0.009531122632324696,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.15353842100012116
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.04837614553980529,
          "train_loss": 0.04626211151480675,
          "validation_loss": 0.016337165609002113,
          "train_metric": 0.05358617380261421,
          "validation_metric": 0.007658348418772221,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19178199100133497
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.04471964918775484,
          "train_loss": 0.04425390809774399,
          "validation_loss": 0.013998115435242653,
          "train_metric": 0.05157621204853058,
          "validation_metric": 0.005606347229331732,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.22765736200017272
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.04373533581383526,
          "train_loss": 0.04325307160615921,
          "validation_loss": 0.01355108991265297,
          "train_metric": 0.05121893063187599,
          "validation_metric": 0.0055157202295959,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24438047700095922
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.04371360089862719,
          "train_loss": 0.04373786225914955,
          "validation_loss": 0.013347881846129894,
          "train_metric": 0.05179528519511223,
          "validation_metric": 0.004882802255451679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2611147190000338
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.042693184688687325,
          "train_loss": 0.041788164526224136,
          "validation_loss": 0.013101765885949135,
          "train_metric": 0.049633391201496124,
          "validation_metric": 0.005517751909792423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2772028270010196
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.040946113876998425,
          "train_loss": 0.04036783427000046,
          "validation_loss": 0.011438782326877117,
          "train_metric": 0.04874343425035477,
          "validation_metric": 0.004468848928809166,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29388894800104026
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.040501276263967156,
          "train_loss": 0.0400824248790741,
          "validation_loss": 0.010428894311189651,
          "train_metric": 0.049502190202474594,
          "validation_metric": 0.003801898565143347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3110799380010576
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.03918780735693872,
          "train_loss": 0.04035959765315056,
          "validation_loss": 0.010814020410180092,
          "train_metric": 0.049413468688726425,
          "validation_metric": 0.004518735688179731,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.327290659000937
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.038731069245841354,
          "train_loss": 0.03809644281864166,
          "validation_loss": 0.008597866632044315,
          "train_metric": 0.04862922057509422,
          "validation_metric": 0.0032276087440550327,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.34414633400047023
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.03786241670604795,
          "train_loss": 0.03694841265678406,
          "validation_loss": 0.007509253453463316,
          "train_metric": 0.048051197081804276,
          "validation_metric": 0.002576056867837906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36104666400024144
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.03750052209943533,
          "train_loss": 0.03745760768651962,
          "validation_loss": 0.007474484853446484,
          "train_metric": 0.04873830825090408,
          "validation_metric": 0.002150633605197072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37812391800071055
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.036930383066646755,
          "train_loss": 0.03650667145848274,
          "validation_loss": 0.00770757719874382,
          "train_metric": 0.047044191509485245,
          "validation_metric": 0.002521028509363532,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.39385380200110376
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.036111121298745275,
          "train_loss": 0.03563521429896355,
          "validation_loss": 0.006066084839403629,
          "train_metric": 0.047419577836990356,
          "validation_metric": 0.0018172612180933356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4101567289999366
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.03550480272679124,
          "train_loss": 0.035234399139881134,
          "validation_loss": 0.006223419681191444,
          "train_metric": 0.04666748270392418,
          "validation_metric": 0.0019954536110162735,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4259209780011588
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.035146358190104365,
          "train_loss": 0.0349900983273983,
          "validation_loss": 0.005557581316679716,
          "train_metric": 0.046951644122600555,
          "validation_metric": 0.0018466212786734104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44233168800019484
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.03498902090359479,
          "train_loss": 0.03467660769820213,
          "validation_loss": 0.0054047624580562115,
          "train_metric": 0.04664648696780205,
          "validation_metric": 0.0018030889332294464,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.458891759000835
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.035555200884118676,
          "train_loss": 0.03546370193362236,
          "validation_loss": 0.006090263370424509,
          "train_metric": 0.04616948962211609,
          "validation_metric": 0.001984664471819997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47470936799982155
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.03493857220746577,
          "train_loss": 0.03503114730119705,
          "validation_loss": 0.005300463642925024,
          "train_metric": 0.04744476452469826,
          "validation_metric": 0.0012891178485006094,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.49059897799998
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.03436033014440909,
          "train_loss": 0.03394469618797302,
          "validation_loss": 0.004516471177339554,
          "train_metric": 0.04651011899113655,
          "validation_metric": 0.0012218669289723039,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5075578830001177
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.03479123502620496,
          "train_loss": 0.03442396596074104,
          "validation_loss": 0.004720527678728104,
          "train_metric": 0.04731149971485138,
          "validation_metric": 0.0011707396479323506,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5243931470013194
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.03406581384479068,
          "train_loss": 0.033730197697877884,
          "validation_loss": 0.004258996807038784,
          "train_metric": 0.04675251618027687,
          "validation_metric": 0.0012488909997045994,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5411202169998433
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.03375655796844512,
          "train_loss": 0.03481942042708397,
          "validation_loss": 0.004926193971186876,
          "train_metric": 0.047578949481248856,
          "validation_metric": 0.0011352808214724064,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5576503259999299
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.03422363189747557,
          "train_loss": 0.03374017775058746,
          "validation_loss": 0.0042082988657057285,
          "train_metric": 0.046303439885377884,
          "validation_metric": 0.0012708562426269054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5740687230008916
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.033777374774217606,
          "train_loss": 0.03348848968744278,
          "validation_loss": 0.0041548823937773705,
          "train_metric": 0.04691184684634209,
          "validation_metric": 0.0010912809520959854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5905660130010801
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.034063213563058525,
          "train_loss": 0.033820077776908875,
          "validation_loss": 0.004114925861358643,
          "train_metric": 0.04726625978946686,
          "validation_metric": 0.0009655157919041812,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6072240290013724
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.033759839134290814,
          "train_loss": 0.03367706015706062,
          "validation_loss": 0.004127349704504013,
          "train_metric": 0.045937858521938324,
          "validation_metric": 0.0009834624361246824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.622770097999819
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.03380965784890577,
          "train_loss": 0.033930424600839615,
          "validation_loss": 0.004244695883244276,
          "train_metric": 0.046447087079286575,
          "validation_metric": 0.001058994559571147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6387741250000545
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.03391111263772473,
          "train_loss": 0.033377911895513535,
          "validation_loss": 0.0038330357056111097,
          "train_metric": 0.04574092850089073,
          "validation_metric": 0.001071553211659193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6543637209997542
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.033460460137575865,
          "train_loss": 0.033465638756752014,
          "validation_loss": 0.003828318091109395,
          "train_metric": 0.04593629017472267,
          "validation_metric": 0.0009575753938406706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6717205969998759
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.033219013770576566,
          "train_loss": 0.032909516245126724,
          "validation_loss": 0.0034298666287213564,
          "train_metric": 0.04683627560734749,
          "validation_metric": 0.0008334886515513062,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6889194539999153
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.0327588003128767,
          "train_loss": 0.03593100607395172,
          "validation_loss": 0.0056160818785429,
          "train_metric": 0.04896347597241402,
          "validation_metric": 0.0013066127430647612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7053844880010729
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.03483736148336902,
          "train_loss": 0.03306194022297859,
          "validation_loss": 0.003660640912130475,
          "train_metric": 0.04563331604003906,
          "validation_metric": 0.0010266602039337158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7210853739998129
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.033253241679631174,
          "train_loss": 0.032559823244810104,
          "validation_loss": 0.0032048688735812902,
          "train_metric": 0.046288102865219116,
          "validation_metric": 0.000850688957143575,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7370506080005725
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.03278916742419824,
          "train_loss": 0.03252429887652397,
          "validation_loss": 0.0030155053827911615,
          "train_metric": 0.04667215794324875,
          "validation_metric": 0.0007339575095102191,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7536301979998825
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.03296593215782195,
          "train_loss": 0.03249392658472061,
          "validation_loss": 0.0029754231218248606,
          "train_metric": 0.0466771274805069,
          "validation_metric": 0.0007271473295986652,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7704471160013782
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.0326227072509937,
          "train_loss": 0.03228198364377022,
          "validation_loss": 0.0028938346076756716,
          "train_metric": 0.04610252007842064,
          "validation_metric": 0.00076154904672876,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7861349860013433
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.03291224416898331,
          "train_loss": 0.03274695202708244,
          "validation_loss": 0.003351144725456834,
          "train_metric": 0.045470673590898514,
          "validation_metric": 0.0008924328722059727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8019697620002262
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.03296610445249826,
          "train_loss": 0.03253808617591858,
          "validation_loss": 0.003090714104473591,
          "train_metric": 0.045851342380046844,
          "validation_metric": 0.0008735070587135851,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8176618630004668
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.03262936136161443,
          "train_loss": 0.03228664770722389,
          "validation_loss": 0.0026632279623299837,
          "train_metric": 0.04654451832175255,
          "validation_metric": 0.0006320839165709913,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8342269690001558
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.03333227016264573,
          "train_loss": 0.03255922719836235,
          "validation_loss": 0.0034408739302307367,
          "train_metric": 0.045544497668743134,
          "validation_metric": 0.0009399988339282572,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8504535030006082
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.03253174317069352,
          "train_loss": 0.0323442742228508,
          "validation_loss": 0.0027968327049165964,
          "train_metric": 0.04669025540351868,
          "validation_metric": 0.0007326480117626488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8662620800005243
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.03281922794121783,
          "train_loss": 0.03206895291805267,
          "validation_loss": 0.002541799796745181,
          "train_metric": 0.04640009254217148,
          "validation_metric": 0.0005661846371367574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8827704040013487
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.03243218013085425,
          "train_loss": 0.03358844295144081,
          "validation_loss": 0.003548211185261607,
          "train_metric": 0.04713385924696922,
          "validation_metric": 0.0007745941984467208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8981626210006652
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.03322297957492992,
          "train_loss": 0.03707694262266159,
          "validation_loss": 0.007315121591091156,
          "train_metric": 0.046245694160461426,
          "validation_metric": 0.0022803558968007565,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9134573330011335
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.03430331946583465,
          "train_loss": 0.03211798146367073,
          "validation_loss": 0.0026901336386799812,
          "train_metric": 0.04622667655348778,
          "validation_metric": 0.0007404742646031082,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9293404600011854
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.032969741281704046,
          "train_loss": 0.03308772295713425,
          "validation_loss": 0.0028495038859546185,
          "train_metric": 0.04716848209500313,
          "validation_metric": 0.0006007468327879906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9447523950002505
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.03295442054513842,
          "train_loss": 0.03188150003552437,
          "validation_loss": 0.0024946462363004684,
          "train_metric": 0.04605082422494888,
          "validation_metric": 0.0006716126226820052,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9600610140005301
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.03240460861707106,
          "train_loss": 0.03257463127374649,
          "validation_loss": 0.0031658567022532225,
          "train_metric": 0.04586561769247055,
          "validation_metric": 0.0008920782711356878,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9749620370002958
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.0327305186947342,
          "train_loss": 0.03285291790962219,
          "validation_loss": 0.003039221279323101,
          "train_metric": 0.046767015010118484,
          "validation_metric": 0.0006375978118740022,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0822048940008244
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.032425907091237605,
          "train_loss": 0.03173479065299034,
          "validation_loss": 0.002411222318187356,
          "train_metric": 0.04573645815253258,
          "validation_metric": 0.0005804225220344961,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0969144530008634
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.03198847038584063,
          "train_loss": 0.031733185052871704,
          "validation_loss": 0.0022763877641409636,
          "train_metric": 0.04618040472269058,
          "validation_metric": 0.0005154553800821304,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.11174421800024
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.03266188409179449,
          "train_loss": 0.031990572810173035,
          "validation_loss": 0.002189145190641284,
          "train_metric": 0.04691360145807266,
          "validation_metric": 0.00047494645696133375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1270711500001198
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.03348537645069882,
          "train_loss": 0.03368443250656128,
          "validation_loss": 0.004639388062059879,
          "train_metric": 0.044854532927274704,
          "validation_metric": 0.0011904017301276326,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1414014140009385
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.032566992682404816,
          "train_loss": 0.03187689930200577,
          "validation_loss": 0.002167727332562208,
          "train_metric": 0.046890877187252045,
          "validation_metric": 0.000462186784716323,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.156276589999834
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.03184469781990629,
          "train_loss": 0.03292638063430786,
          "validation_loss": 0.00281457114033401,
          "train_metric": 0.04741595685482025,
          "validation_metric": 0.0005982432630844414,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1700657769997633
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.032094332156702876,
          "train_loss": 0.03159130737185478,
          "validation_loss": 0.0022030798718333244,
          "train_metric": 0.045638177543878555,
          "validation_metric": 0.0005310919368639588,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1842470170013257
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.031631570018362254,
          "train_loss": 0.03143545612692833,
          "validation_loss": 0.0019283103756606579,
          "train_metric": 0.04630361497402191,
          "validation_metric": 0.00040732265915721655,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1989256720007688
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.031612847320502624,
          "train_loss": 0.031279005110263824,
          "validation_loss": 0.0018292799359187484,
          "train_metric": 0.04598189890384674,
          "validation_metric": 0.000398215779569,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2136493460002384
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.03180725371930748,
          "train_loss": 0.031185859814286232,
          "validation_loss": 0.0018093348480761051,
          "train_metric": 0.04577663168311119,
          "validation_metric": 0.00040991167770698667,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2276004100003775
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.03182887885486707,
          "train_loss": 0.0322415716946125,
          "validation_loss": 0.0028054940048605204,
          "train_metric": 0.04545264691114426,
          "validation_metric": 0.0007288227207027376,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.24176538500069
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.0317946867653518,
          "train_loss": 0.03116675652563572,
          "validation_loss": 0.0016605125274509192,
          "train_metric": 0.04623950645327568,
          "validation_metric": 0.0003592930734157562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.256694472000163
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.03159998202318093,
          "train_loss": 0.031633634120225906,
          "validation_loss": 0.0017753585707396269,
          "train_metric": 0.046669382601976395,
          "validation_metric": 0.000378055585315451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2711686000002373
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.031806644808966666,
          "train_loss": 0.0311152134090662,
          "validation_loss": 0.0017401796067133546,
          "train_metric": 0.0456455834209919,
          "validation_metric": 0.00038515019696205854,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.28544084799978
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.031367689138278365,
          "train_loss": 0.031214900314807892,
          "validation_loss": 0.0017008621944114566,
          "train_metric": 0.0461522713303566,
          "validation_metric": 0.00035573719651438296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3001474180000514
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.03160209445923101,
          "train_loss": 0.0312529020011425,
          "validation_loss": 0.001587537000887096,
          "train_metric": 0.04643610864877701,
          "validation_metric": 0.0003325312281958759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3149685940006748
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.031475014868192375,
          "train_loss": 0.03150685131549835,
          "validation_loss": 0.002257121726870537,
          "train_metric": 0.04543900862336159,
          "validation_metric": 0.0004992816247977316,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3291265469997597
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.031181065714918077,
          "train_loss": 0.030948122963309288,
          "validation_loss": 0.0014562556752935052,
          "train_metric": 0.0460929311811924,
          "validation_metric": 0.00031410163501277566,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3437458890002745
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.03125756165536586,
          "train_loss": 0.03172691538929939,
          "validation_loss": 0.0020606964826583862,
          "train_metric": 0.04687158390879631,
          "validation_metric": 0.00042558510904200375,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3580066320009792
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.03164988779462874,
          "train_loss": 0.03100648894906044,
          "validation_loss": 0.001595845678821206,
          "train_metric": 0.04592273011803627,
          "validation_metric": 0.00035776878939941525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.372209754999858
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.03227133081963984,
          "train_loss": 0.031886324286460876,
          "validation_loss": 0.00269677652977407,
          "train_metric": 0.04510672017931938,
          "validation_metric": 0.0006334505742415786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3833580430000438
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.03298668167553842,
          "train_loss": 0.03230725973844528,
          "validation_loss": 0.0024747734423726797,
          "train_metric": 0.046573441475629807,
          "validation_metric": 0.0005061576375737786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3937275200005388
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.03175247014587512,
          "train_loss": 0.031197484582662582,
          "validation_loss": 0.0017845684196799994,
          "train_metric": 0.04589526355266571,
          "validation_metric": 0.00039755456964485347,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4034380419998342
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.03167965305692633,
          "train_loss": 0.03104061260819435,
          "validation_loss": 0.0018814700888469815,
          "train_metric": 0.04564366117119789,
          "validation_metric": 0.00042187818326056004,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4134825840010308
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.031203816062770784,
          "train_loss": 0.0314963199198246,
          "validation_loss": 0.001930614816956222,
          "train_metric": 0.046460483223199844,
          "validation_metric": 0.0003954051935579628,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4234702670000843
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.031835220608627424,
          "train_loss": 0.03160944953560829,
          "validation_loss": 0.002595019293949008,
          "train_metric": 0.045430634170770645,
          "validation_metric": 0.000636602402664721,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4334875929998816
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.03120106994174421,
          "train_loss": 0.031006542965769768,
          "validation_loss": 0.0017000753432512283,
          "train_metric": 0.04545034468173981,
          "validation_metric": 0.0003823485167231411,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.443418216000282
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.03132724342867732,
          "train_loss": 0.030746541917324066,
          "validation_loss": 0.0014100456610321999,
          "train_metric": 0.04583911597728729,
          "validation_metric": 0.0002970278437715024,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4534120349999284
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.031652782403398305,
          "train_loss": 0.03231584280729294,
          "validation_loss": 0.0020690944511443377,
          "train_metric": 0.04741077497601509,
          "validation_metric": 0.0004301238222979009,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.463443440001356
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.032487676369783,
          "train_loss": 0.03110012412071228,
          "validation_loss": 0.0015529177617281675,
          "train_metric": 0.04641200229525566,
          "validation_metric": 0.00031416930141858757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4734310830008326
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.031792404508451,
          "train_loss": 0.030826956033706665,
          "validation_loss": 0.0015395813388749957,
          "train_metric": 0.04550271853804588,
          "validation_metric": 0.00033148526563309133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4832471400004579
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.0315678883343935,
          "train_loss": 0.03203922137618065,
          "validation_loss": 0.0027131391689181328,
          "train_metric": 0.04559997841715813,
          "validation_metric": 0.0007050200365483761,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.493169121000392
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.03146226311218925,
          "train_loss": 0.031151598319411278,
          "validation_loss": 0.001959660556167364,
          "train_metric": 0.045968323945999146,
          "validation_metric": 0.000395912560634315,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5029300780006452
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.03180668130517006,
          "train_loss": 0.030886158347129822,
          "validation_loss": 0.0014304982032626867,
          "train_metric": 0.046070463955402374,
          "validation_metric": 0.0003249126602895558,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5129783840002347
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.031388113275170326,
          "train_loss": 0.031573206186294556,
          "validation_loss": 0.001837760559283197,
          "train_metric": 0.04601820558309555,
          "validation_metric": 0.00045755485189147294,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5228255180009
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.03272363707947079,
          "train_loss": 0.033309925347566605,
          "validation_loss": 0.003656513523310423,
          "train_metric": 0.04660727456212044,
          "validation_metric": 0.0007762781460769475,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5327099610003643
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.03175888433906948,
          "train_loss": 0.03099144995212555,
          "validation_loss": 0.0017894974444061518,
          "train_metric": 0.04594680666923523,
          "validation_metric": 0.0003840421559289098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5425957740008016
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.03103437353274785,
          "train_loss": 0.03113172948360443,
          "validation_loss": 0.00134428427554667,
          "train_metric": 0.0465414859354496,
          "validation_metric": 0.0002737738541327417,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5529235100002552
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.030851975083351135,
          "train_loss": 0.030829614028334618,
          "validation_loss": 0.0015139655442908406,
          "train_metric": 0.045542728155851364,
          "validation_metric": 0.00032810872653499246,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5631996120009717
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.031088276380614843,
          "train_loss": 0.032423894852399826,
          "validation_loss": 0.0033197251614183187,
          "train_metric": 0.04547065868973732,
          "validation_metric": 0.0008587859920226038,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5730072290007229
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.031131965050008148,
          "train_loss": 0.031192108988761902,
          "validation_loss": 0.0015738995280116796,
          "train_metric": 0.046094126999378204,
          "validation_metric": 0.00037279046955518425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5827814730000682
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.03085399535484612,
          "train_loss": 0.03259699046611786,
          "validation_loss": 0.0027889679186046124,
          "train_metric": 0.04705049842596054,
          "validation_metric": 0.0006155834998935461,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5924701350013493
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.03069496765965596,
          "train_loss": 0.03218904137611389,
          "validation_loss": 0.002378005301579833,
          "train_metric": 0.0457603819668293,
          "validation_metric": 0.0006145855877548456,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.602155403001234
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.031247161590727046,
          "train_loss": 0.03140828758478165,
          "validation_loss": 0.0019904079381376505,
          "train_metric": 0.045936357229948044,
          "validation_metric": 0.0004017332976218313,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6122236770006566
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.031491402092797216,
          "train_loss": 0.03141407668590546,
          "validation_loss": 0.0020126786548644304,
          "train_metric": 0.04566221311688423,
          "validation_metric": 0.0004896221216768026,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6219826779997675
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.03148472239263356,
          "train_loss": 0.030515197664499283,
          "validation_loss": 0.0011583085870370269,
          "train_metric": 0.04581701382994652,
          "validation_metric": 0.00024822907289490104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6324641760002123
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.031094598933123052,
          "train_loss": 0.030798478052020073,
          "validation_loss": 0.0018140728352591395,
          "train_metric": 0.04523681104183197,
          "validation_metric": 0.00038091125315986574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6425262530010514
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.031101700780709507,
          "train_loss": 0.031383588910102844,
          "validation_loss": 0.0014178843703120947,
          "train_metric": 0.04664138704538345,
          "validation_metric": 0.0002863992122001946,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6521879419997276
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.030949742591474205,
          "train_loss": 0.030689693987369537,
          "validation_loss": 0.0014874771004542708,
          "train_metric": 0.04541622847318649,
          "validation_metric": 0.00030181941110640764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6621187970013125
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.0313816198031418,
          "train_loss": 0.032508544623851776,
          "validation_loss": 0.0024694427847862244,
          "train_metric": 0.04770772531628609,
          "validation_metric": 0.0005201988969929516,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6720039379997615
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.03123186427546898,
          "train_loss": 0.030550645664334297,
          "validation_loss": 0.0013547340640798211,
          "train_metric": 0.045653920620679855,
          "validation_metric": 0.00027566467178985476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6820272360000672
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.030745935306185856,
          "train_loss": 0.030543562024831772,
          "validation_loss": 0.0011158627457916737,
          "train_metric": 0.04571232572197914,
          "validation_metric": 0.00023979274556040764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6922946730010153
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.031130320858210325,
          "train_loss": 0.03240647166967392,
          "validation_loss": 0.003151506185531616,
          "train_metric": 0.045334719121456146,
          "validation_metric": 0.0007818837184458971,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.702067481000995
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.031813755798793864,
          "train_loss": 0.030610771849751472,
          "validation_loss": 0.0010635401122272015,
          "train_metric": 0.04594593495130539,
          "validation_metric": 0.0002314714074600488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7124697550007113
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.03072772031009663,
          "train_loss": 0.031176164746284485,
          "validation_loss": 0.0018862535944208503,
          "train_metric": 0.04619297385215759,
          "validation_metric": 0.0003851841320283711,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7223193300014827
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.03083945030812174,
          "train_loss": 0.03067818656563759,
          "validation_loss": 0.0011359001509845257,
          "train_metric": 0.04612601175904274,
          "validation_metric": 0.00024738049251027405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7322718750001513
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.030666289676446468,
          "train_loss": 0.03033623658120632,
          "validation_loss": 0.0010397147852927446,
          "train_metric": 0.04562725871801376,
          "validation_metric": 0.00021375225333031267,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7426930290002929
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.031044082090375014,
          "train_loss": 0.030460916459560394,
          "validation_loss": 0.0009279473451897502,
          "train_metric": 0.046107735484838486,
          "validation_metric": 0.00018823861319106072,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7529473560007318
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.030761816327867564,
          "train_loss": 0.030427895486354828,
          "validation_loss": 0.001110633835196495,
          "train_metric": 0.04580586776137352,
          "validation_metric": 0.00022372441890183836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7629326500009483
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.030679683390189894,
          "train_loss": 0.03115617297589779,
          "validation_loss": 0.0014620579313486814,
          "train_metric": 0.0458313524723053,
          "validation_metric": 0.00032209878554567695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7727342520011007
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.031147726345807314,
          "train_loss": 0.031466152518987656,
          "validation_loss": 0.0027880368288606405,
          "train_metric": 0.04480317234992981,
          "validation_metric": 0.0005793062737211585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7826571210007387
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.030999308015452698,
          "train_loss": 0.03065311908721924,
          "validation_loss": 0.0009738695807754993,
          "train_metric": 0.04603516310453415,
          "validation_metric": 0.00020855863112956285,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7925717560010526
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.03144098893972114,
          "train_loss": 0.03280216082930565,
          "validation_loss": 0.0025191691238433123,
          "train_metric": 0.04746819660067558,
          "validation_metric": 0.000553326855879277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8023002560003079
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.030837241793051362,
          "train_loss": 0.030628452077507973,
          "validation_loss": 0.0014059914974495769,
          "train_metric": 0.045407965779304504,
          "validation_metric": 0.00031075815786607563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8123594330008927
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.03092650060352753,
          "train_loss": 0.03039480559527874,
          "validation_loss": 0.0009434777311980724,
          "train_metric": 0.04584764689207077,
          "validation_metric": 0.00019813769904430956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8220580400011386
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.030697042937390506,
          "train_loss": 0.03125278651714325,
          "validation_loss": 0.0016916993772611022,
          "train_metric": 0.045603614300489426,
          "validation_metric": 0.00036891052150167525,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.832080718000725
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.031044638948515058,
          "train_loss": 0.03122684173285961,
          "validation_loss": 0.002078142948448658,
          "train_metric": 0.04517097398638725,
          "validation_metric": 0.0004788570513483137,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8421378600014577
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.03140004264423624,
          "train_loss": 0.03112884610891342,
          "validation_loss": 0.00202403636649251,
          "train_metric": 0.045050933957099915,
          "validation_metric": 0.00045138783752918243,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8519680370009155
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.0313502661374514,
          "train_loss": 0.0324164517223835,
          "validation_loss": 0.0017945917788892984,
          "train_metric": 0.04786069318652153,
          "validation_metric": 0.0003654761821962893,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8619796300008602
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.031075151171535254,
          "train_loss": 0.030453959479928017,
          "validation_loss": 0.001108350232243538,
          "train_metric": 0.04580771178007126,
          "validation_metric": 0.00022926591918803751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9404536700003518
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.030478378816042095,
          "train_loss": 0.030876465141773224,
          "validation_loss": 0.0017533638747408986,
          "train_metric": 0.04525604471564293,
          "validation_metric": 0.000383678445359692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9505743740010075
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.030675002744828817,
          "train_loss": 0.030736170709133148,
          "validation_loss": 0.0014113340293988585,
          "train_metric": 0.045642685145139694,
          "validation_metric": 0.0003174164448864758,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9606392120003875
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.030804554640781134,
          "train_loss": 0.031015994027256966,
          "validation_loss": 0.0016289487248286605,
          "train_metric": 0.04536404460668564,
          "validation_metric": 0.0003616121248342097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9704065480000281
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.03170208289520815,
          "train_loss": 0.031395938247442245,
          "validation_loss": 0.0015486839693039656,
          "train_metric": 0.046659499406814575,
          "validation_metric": 0.0003177486069034785,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9804767910009105
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.03074141830438748,
          "train_loss": 0.03040551021695137,
          "validation_loss": 0.0008792566950432956,
          "train_metric": 0.04619486629962921,
          "validation_metric": 0.00018484410247765481,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.990810915000111
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.031060316076036543,
          "train_loss": 0.031311675906181335,
          "validation_loss": 0.002059133490547538,
          "train_metric": 0.04519331827759743,
          "validation_metric": 0.0004772381216753274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.000801919000878
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.031342678688815795,
          "train_loss": 0.031207142397761345,
          "validation_loss": 0.001021399861201644,
          "train_metric": 0.04703754931688309,
          "validation_metric": 0.00020574367954395711,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0108264650007186
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.03161652496783063,
          "train_loss": 0.030536919832229614,
          "validation_loss": 0.0010368117364123464,
          "train_metric": 0.04600323364138603,
          "validation_metric": 0.00021625400404445827,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.020549660999677
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.0304195003118366,
          "train_loss": 0.03043108992278576,
          "validation_loss": 0.0010985390981659293,
          "train_metric": 0.045680396258831024,
          "validation_metric": 0.0002212816762039438,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0306091970014677
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.030717083951458335,
          "train_loss": 0.03139191120862961,
          "validation_loss": 0.001656338688917458,
          "train_metric": 0.04651424661278725,
          "validation_metric": 0.00037935710861347616,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0404108060010913
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.030942175420932472,
          "train_loss": 0.03052329458296299,
          "validation_loss": 0.0014809167478233576,
          "train_metric": 0.045222774147987366,
          "validation_metric": 0.0003033172688446939,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0501095770014217
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.03046675183577463,
          "train_loss": 0.03021595999598503,
          "validation_loss": 0.0009837797842919827,
          "train_metric": 0.04553058370947838,
          "validation_metric": 0.00020054330525454134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0599461649999284
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.03039627526595723,
          "train_loss": 0.03038148581981659,
          "validation_loss": 0.0010226749582216144,
          "train_metric": 0.04597209021449089,
          "validation_metric": 0.00020600292191375047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.06972425999993
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.03068088524742052,
          "train_loss": 0.03048786148428917,
          "validation_loss": 0.0008224206976592541,
          "train_metric": 0.046479739248752594,
          "validation_metric": 0.0001657365501159802,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0802731950007
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.031249507213942707,
          "train_loss": 0.03116229549050331,
          "validation_loss": 0.0017801998183131218,
          "train_metric": 0.0456397645175457,
          "validation_metric": 0.00042300292989239097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0901123310013645
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.031377179606352,
          "train_loss": 0.030409131199121475,
          "validation_loss": 0.0012183457147330046,
          "train_metric": 0.04524818807840347,
          "validation_metric": 0.0002460608084220439,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.099802548000298
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.030818487983196974,
          "train_loss": 0.030912261456251144,
          "validation_loss": 0.0016557881608605385,
          "train_metric": 0.04591020569205284,
          "validation_metric": 0.00036945444298908114,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1095333480006957
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.03280367674597073,
          "train_loss": 0.03140148147940636,
          "validation_loss": 0.0013039453187957406,
          "train_metric": 0.04698985069990158,
          "validation_metric": 0.0002747987164184451,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1190816560010717
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.03125934023410082,
          "train_loss": 0.030839964747428894,
          "validation_loss": 0.0019197443034499884,
          "train_metric": 0.045181531459093094,
          "validation_metric": 0.00038669921923428774,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1291223210009775
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.03051063825841993,
          "train_loss": 0.03167155012488365,
          "validation_loss": 0.0020815376192331314,
          "train_metric": 0.04584897682070732,
          "validation_metric": 0.0005004414124414325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1388787320011033
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.031341209236416034,
          "train_loss": 0.030818922445178032,
          "validation_loss": 0.0010522333905100822,
          "train_metric": 0.04666908457875252,
          "validation_metric": 0.00021201501658651978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.148705568000878
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.030839772545732558,
          "train_loss": 0.030710002407431602,
          "validation_loss": 0.0014839889481663704,
          "train_metric": 0.045999255031347275,
          "validation_metric": 0.00029895143234170973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1587172569998074
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.031178671575617045,
          "train_loss": 0.030462993308901787,
          "validation_loss": 0.000857893901411444,
          "train_metric": 0.046298798173666,
          "validation_metric": 0.00017650207155384123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1685530150007253
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.030812666518613696,
          "train_loss": 0.030946282669901848,
          "validation_loss": 0.001276614610105753,
          "train_metric": 0.045689769089221954,
          "validation_metric": 0.00030006514862179756,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1786193780008034
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.031061617250088602,
          "train_loss": 0.03213118761777878,
          "validation_loss": 0.002644319087266922,
          "train_metric": 0.045685768127441406,
          "validation_metric": 0.0006681644008494914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1882587090003653
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.031008849218778778,
          "train_loss": 0.030514493584632874,
          "validation_loss": 0.001194693031720817,
          "train_metric": 0.046068787574768066,
          "validation_metric": 0.00024065021716523916,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1979979330008064
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.030605361331254244,
          "train_loss": 0.03058699704706669,
          "validation_loss": 0.0009588187094777822,
          "train_metric": 0.04619155451655388,
          "validation_metric": 0.00019313741358928382,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.207533933000377
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.03085917921271175,
          "train_loss": 0.030473828315734863,
          "validation_loss": 0.0010096777696162462,
          "train_metric": 0.04615941643714905,
          "validation_metric": 0.00020338583271950483,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2170375380010228
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.030686738915392198,
          "train_loss": 0.03034389577805996,
          "validation_loss": 0.0010379886953160167,
          "train_metric": 0.04556240886449814,
          "validation_metric": 0.00021943775936961174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2269402830006584
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.03065169556066394,
          "train_loss": 0.03021569550037384,
          "validation_loss": 0.0007342025637626648,
          "train_metric": 0.04591430723667145,
          "validation_metric": 0.00014840274525340647,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2371943289999763
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.03147498215548694,
          "train_loss": 0.031652290374040604,
          "validation_loss": 0.0026035872288048267,
          "train_metric": 0.04509064182639122,
          "validation_metric": 0.0006226192926988006,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2472300810004526
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.0306945949268993,
          "train_loss": 0.030262762680649757,
          "validation_loss": 0.0007334770052693784,
          "train_metric": 0.04631359875202179,
          "validation_metric": 0.00014774625014979392,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2574869230011245
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.030595465377700748,
          "train_loss": 0.03122166357934475,
          "validation_loss": 0.0016497119795531034,
          "train_metric": 0.04668010398745537,
          "validation_metric": 0.00034688133746385574,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2673366260005423
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.030637507254141383,
          "train_loss": 0.031226716935634613,
          "validation_loss": 0.0014446692075580359,
          "train_metric": 0.04717031866312027,
          "validation_metric": 0.00029703197651542723,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2772509120004543
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.030906791449524462,
          "train_loss": 0.030158795416355133,
          "validation_loss": 0.0008469221647828817,
          "train_metric": 0.045584168285131454,
          "validation_metric": 0.00017627194756641984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.286751491999894
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.03057454392546788,
          "train_loss": 0.03051614947617054,
          "validation_loss": 0.0012460009893402457,
          "train_metric": 0.04543805494904518,
          "validation_metric": 0.00026840573991648853,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2964015970010223
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.0317154333461076,
          "train_loss": 0.030912743881344795,
          "validation_loss": 0.001645619166083634,
          "train_metric": 0.04510989785194397,
          "validation_metric": 0.00036556905251927674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.306009511999946
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.031219525568303652,
          "train_loss": 0.030879737809300423,
          "validation_loss": 0.0009703147225081921,
          "train_metric": 0.04630342870950699,
          "validation_metric": 0.00021606743393931538,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3154936660012027
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.030452019593212754,
          "train_loss": 0.03029276244342327,
          "validation_loss": 0.0009605939267203212,
          "train_metric": 0.04591861367225647,
          "validation_metric": 0.00019355567928869277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.325019191001047
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.030475522973574698,
          "train_loss": 0.03068518452346325,
          "validation_loss": 0.0016928615514189005,
          "train_metric": 0.04487898573279381,
          "validation_metric": 0.0003483012260403484,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3344367110003077
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.030673837340145838,
          "train_loss": 0.030315062031149864,
          "validation_loss": 0.0008202099124900997,
          "train_metric": 0.04584439471364021,
          "validation_metric": 0.00017184630269184709,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.343897607001054
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.030651657842099667,
          "train_loss": 0.030082907527685165,
          "validation_loss": 0.0007064330857247114,
          "train_metric": 0.04589345306158066,
          "validation_metric": 0.00014229872613213956,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.353805865999675
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.030350185348652303,
          "train_loss": 0.030414095148444176,
          "validation_loss": 0.000795855070464313,
          "train_metric": 0.04605066403746605,
          "validation_metric": 0.00016497084288857877,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.363503815000513
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.03056957299122587,
          "train_loss": 0.031522419303655624,
          "validation_loss": 0.0021180384792387486,
          "train_metric": 0.04528548941016197,
          "validation_metric": 0.00047466051182709634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.373123299999861
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.030674134322907776,
          "train_loss": 0.030385034158825874,
          "validation_loss": 0.000859808293171227,
          "train_metric": 0.04610995575785637,
          "validation_metric": 0.00017460559320170432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3825894210003753
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.030372491528396495,
          "train_loss": 0.03181388974189758,
          "validation_loss": 0.0022362421732395887,
          "train_metric": 0.04556668549776077,
          "validation_metric": 0.0005498695536516607,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3921367050006666
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.03127258074528072,
          "train_loss": 0.030347654595971107,
          "validation_loss": 0.0010853551793843508,
          "train_metric": 0.04571007564663887,
          "validation_metric": 0.0002186259807785973,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4015637940010492
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.030599605292081833,
          "train_loss": 0.03082403726875782,
          "validation_loss": 0.0011376435868442059,
          "train_metric": 0.04660509154200554,
          "validation_metric": 0.00023119956313166767,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.411501267000858
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.030324303486850113,
          "train_loss": 0.03003951907157898,
          "validation_loss": 0.0006674150354228914,
          "train_metric": 0.04562545195221901,
          "validation_metric": 0.00013443920761346817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4213297389997024
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.030226084461901337,
          "train_loss": 0.030077515169978142,
          "validation_loss": 0.0006321418331936002,
          "train_metric": 0.04596119374036789,
          "validation_metric": 0.00012733401672448963,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4309818980000273
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.03041944350115955,
          "train_loss": 0.031262487173080444,
          "validation_loss": 0.0016703801229596138,
          "train_metric": 0.04573053494095802,
          "validation_metric": 0.000391609180951491,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.440203678001126
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.031020369824545924,
          "train_loss": 0.030375981703400612,
          "validation_loss": 0.001017413567751646,
          "train_metric": 0.04543473199009895,
          "validation_metric": 0.00021709472639486194,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4491586170006485
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.030828982999082655,
          "train_loss": 0.03012152388691902,
          "validation_loss": 0.0007420717738568783,
          "train_metric": 0.0457184873521328,
          "validation_metric": 0.00014990800991654396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4584352250003576
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.030275687990069855,
          "train_loss": 0.03020205721259117,
          "validation_loss": 0.0006890574004501104,
          "train_metric": 0.046007007360458374,
          "validation_metric": 0.00013879869948141277,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.467317197000739
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.03051343490369618,
          "train_loss": 0.03034152276813984,
          "validation_loss": 0.0007212945492938161,
          "train_metric": 0.04646962881088257,
          "validation_metric": 0.00014529228792525828,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.476820839001448
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.030582979990867898,
          "train_loss": 0.030202500522136688,
          "validation_loss": 0.0009357734234072268,
          "train_metric": 0.04554763436317444,
          "validation_metric": 0.0001886137033579871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.485962611001014
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.030302057159133255,
          "train_loss": 0.03147101029753685,
          "validation_loss": 0.0022520769853144884,
          "train_metric": 0.044984370470047,
          "validation_metric": 0.0005085230222903192,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4953130670000974
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.03094177984894486,
          "train_loss": 0.030028345063328743,
          "validation_loss": 0.0005360577488318086,
          "train_metric": 0.04600106179714203,
          "validation_metric": 0.00010797955474117771,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5049833550001495
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.030280905929430446,
          "train_loss": 0.030096856877207756,
          "validation_loss": 0.0006467922939918935,
          "train_metric": 0.04617442563176155,
          "validation_metric": 0.00013028510147705674,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5146501600011106
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.030451924132648855,
          "train_loss": 0.03173798695206642,
          "validation_loss": 0.0025589235592633486,
          "train_metric": 0.045029886066913605,
          "validation_metric": 0.0005960682756267488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.524338754999917
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.03146732447203249,
          "train_loss": 0.03180747479200363,
          "validation_loss": 0.002621890278533101,
          "train_metric": 0.04490526020526886,
          "validation_metric": 0.0005892033805139363,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5338745810004184
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.0333063147845678,
          "train_loss": 0.03454083576798439,
          "validation_loss": 0.004591136705130339,
          "train_metric": 0.04750962182879448,
          "validation_metric": 0.0013125406112521887,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.543595779001407
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.03159824572503567,
          "train_loss": 0.030348893254995346,
          "validation_loss": 0.001081043970771134,
          "train_metric": 0.045804791152477264,
          "validation_metric": 0.00021775758068542928,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5530797580013314
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.030905173443898093,
          "train_loss": 0.030920540913939476,
          "validation_loss": 0.0011120493290945888,
          "train_metric": 0.04656178876757622,
          "validation_metric": 0.00022539943165611476,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5626123160000134
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.030426588375121355,
          "train_loss": 0.03027356043457985,
          "validation_loss": 0.0007351193344220519,
          "train_metric": 0.04603365808725357,
          "validation_metric": 0.0001480770588386804,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5719879000007495
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.030394525500014424,
          "train_loss": 0.030098948627710342,
          "validation_loss": 0.0007935037137940526,
          "train_metric": 0.045678265392780304,
          "validation_metric": 0.000159837567480281,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.581383712000388
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.030266403045970947,
          "train_loss": 0.031150707975029945,
          "validation_loss": 0.001506541040726006,
          "train_metric": 0.045793842524290085,
          "validation_metric": 0.00034402793971821666,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.590713385001436
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.03148374567535939,
          "train_loss": 0.03363249450922012,
          "validation_loss": 0.00343307014554739,
          "train_metric": 0.04828121140599251,
          "validation_metric": 0.0008074448560364544,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.59969871300018
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.03292571898782626,
          "train_loss": 0.030726375058293343,
          "validation_loss": 0.001161098014563322,
          "train_metric": 0.045623041689395905,
          "validation_metric": 0.00023635293473489583,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.608901835001234
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.03091008064802736,
          "train_loss": 0.030210938304662704,
          "validation_loss": 0.0008182597230188549,
          "train_metric": 0.0455864742398262,
          "validation_metric": 0.00016486475942656398,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6178992470013327
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.031502636906225234,
          "train_loss": 0.030927438288927078,
          "validation_loss": 0.0016055676387622952,
          "train_metric": 0.046453122049570084,
          "validation_metric": 0.0003239314828533679,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6274198640003306
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.03094011827124632,
          "train_loss": 0.03043530322611332,
          "validation_loss": 0.0007297603879123926,
          "train_metric": 0.046066705137491226,
          "validation_metric": 0.0001494243333581835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.636358271000063
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.030587118370021926,
          "train_loss": 0.030155664309859276,
          "validation_loss": 0.000842111767269671,
          "train_metric": 0.04555242508649826,
          "validation_metric": 0.00017362752987537533,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6457936950009753
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.030468796147033572,
          "train_loss": 0.030445698648691177,
          "validation_loss": 0.0010705478489398956,
          "train_metric": 0.04574628546833992,
          "validation_metric": 0.00022870635439176112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.654735203999735
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.03162533463910222,
          "train_loss": 0.030789176002144814,
          "validation_loss": 0.0015320214442908764,
          "train_metric": 0.045569922775030136,
          "validation_metric": 0.00030859920661896467,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.66397367799982
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.031124748580623418,
          "train_loss": 0.030573006719350815,
          "validation_loss": 0.0008204340701922774,
          "train_metric": 0.046796806156635284,
          "validation_metric": 0.0001652622304391116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6730837380000594
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.030820857326034456,
          "train_loss": 0.03043028898537159,
          "validation_loss": 0.0011170345824211836,
          "train_metric": 0.04546646028757095,
          "validation_metric": 0.00023738743038848042,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.682309810001243
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.030775666644331068,
          "train_loss": 0.030677195638418198,
          "validation_loss": 0.001430276082828641,
          "train_metric": 0.045064061880111694,
          "validation_metric": 0.00030578405130654573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.691776954001398
        }
      ],
      "best": {
        "epoch": 180,
        "updates": 2880
      },
      "best_validation_metric": 0.00010797955474117771,
      "config": {
        "label": "Huber + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "huber",
        "outliers": true,
        "max_updates": null
      },
      "seed": 33,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010391321125880853
    },
    "fc71c16e3acd_11": {
      "history": [
        {
          "epoch": 1,
          "updates": 16,
          "samples_seen": 1024,
          "online_loss": 0.8472034744918346,
          "train_loss": 0.6334609985351562,
          "validation_loss": 0.43482571840286255,
          "train_metric": 0.08828626573085785,
          "validation_metric": 0.031051643192768097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.012996055998883094
        },
        {
          "epoch": 2,
          "updates": 32,
          "samples_seen": 2048,
          "online_loss": 0.4547589737921953,
          "train_loss": 0.37369009852409363,
          "validation_loss": 0.22880148887634277,
          "train_metric": 0.055875420570373535,
          "validation_metric": 0.01284992415457964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.02578975399956107
        },
        {
          "epoch": 3,
          "updates": 48,
          "samples_seen": 3072,
          "online_loss": 0.34130298160016537,
          "train_loss": 0.3181215822696686,
          "validation_loss": 0.1833535134792328,
          "train_metric": 0.05665387213230133,
          "validation_metric": 0.013869816437363625,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.03793682399918907
        },
        {
          "epoch": 4,
          "updates": 64,
          "samples_seen": 4096,
          "online_loss": 0.2969324439764023,
          "train_loss": 0.28525635600090027,
          "validation_loss": 0.13791468739509583,
          "train_metric": 0.05338357016444206,
          "validation_metric": 0.008761219680309296,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.05048960999920382
        },
        {
          "epoch": 5,
          "updates": 80,
          "samples_seen": 5120,
          "online_loss": 0.2853407198563218,
          "train_loss": 0.28231081366539,
          "validation_loss": 0.13692638278007507,
          "train_metric": 0.05251999944448471,
          "validation_metric": 0.008317964151501656,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.06282964499951049
        },
        {
          "epoch": 6,
          "updates": 96,
          "samples_seen": 6144,
          "online_loss": 0.2818584246560931,
          "train_loss": 0.2831636667251587,
          "validation_loss": 0.1374346911907196,
          "train_metric": 0.051638733595609665,
          "validation_metric": 0.007298105396330357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.07546215000002121
        },
        {
          "epoch": 7,
          "updates": 112,
          "samples_seen": 7168,
          "online_loss": 0.28553814720362425,
          "train_loss": 0.2914629876613617,
          "validation_loss": 0.14144381880760193,
          "train_metric": 0.051672860980033875,
          "validation_metric": 0.00606126943603158,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.0889406179994694
        },
        {
          "epoch": 8,
          "updates": 128,
          "samples_seen": 8192,
          "online_loss": 0.2865592949092388,
          "train_loss": 0.28008946776390076,
          "validation_loss": 0.1326798051595688,
          "train_metric": 0.05075759440660477,
          "validation_metric": 0.0062399073503911495,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.18282873699899937
        },
        {
          "epoch": 9,
          "updates": 144,
          "samples_seen": 9216,
          "online_loss": 0.27712132409214973,
          "train_loss": 0.2711692750453949,
          "validation_loss": 0.12322766333818436,
          "train_metric": 0.05129736661911011,
          "validation_metric": 0.0064302352257072926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.19496820499989553
        },
        {
          "epoch": 10,
          "updates": 160,
          "samples_seen": 10240,
          "online_loss": 0.27257517958059907,
          "train_loss": 0.26904401183128357,
          "validation_loss": 0.11994107067584991,
          "train_metric": 0.05106697231531143,
          "validation_metric": 0.005939509253948927,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2073826249998092
        },
        {
          "epoch": 11,
          "updates": 176,
          "samples_seen": 11264,
          "online_loss": 0.26880511827766895,
          "train_loss": 0.26504284143447876,
          "validation_loss": 0.1167159229516983,
          "train_metric": 0.05094597861170769,
          "validation_metric": 0.006094294134527445,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.21930481999879703
        },
        {
          "epoch": 12,
          "updates": 192,
          "samples_seen": 12288,
          "online_loss": 0.26191655453294516,
          "train_loss": 0.26963678002357483,
          "validation_loss": 0.12330737709999084,
          "train_metric": 0.05091177299618721,
          "validation_metric": 0.00734011409804225,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2312909629999922
        },
        {
          "epoch": 13,
          "updates": 208,
          "samples_seen": 13312,
          "online_loss": 0.26246218010783195,
          "train_loss": 0.262800008058548,
          "validation_loss": 0.11049545556306839,
          "train_metric": 0.05096280947327614,
          "validation_metric": 0.004786462523043156,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.24376475200006098
        },
        {
          "epoch": 14,
          "updates": 224,
          "samples_seen": 14336,
          "online_loss": 0.2584428498521447,
          "train_loss": 0.2541007995605469,
          "validation_loss": 0.10550546646118164,
          "train_metric": 0.04974636062979698,
          "validation_metric": 0.005570126231759787,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.25592772099844296
        },
        {
          "epoch": 15,
          "updates": 240,
          "samples_seen": 15360,
          "online_loss": 0.2530196597799659,
          "train_loss": 0.2531958222389221,
          "validation_loss": 0.09987027943134308,
          "train_metric": 0.05016251280903816,
          "validation_metric": 0.004309202544391155,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.26811233399894263
        },
        {
          "epoch": 16,
          "updates": 256,
          "samples_seen": 16384,
          "online_loss": 0.2495593847706914,
          "train_loss": 0.2493380606174469,
          "validation_loss": 0.10012219846248627,
          "train_metric": 0.04897024855017662,
          "validation_metric": 0.004474825691431761,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.2800634959985473
        },
        {
          "epoch": 17,
          "updates": 272,
          "samples_seen": 17408,
          "online_loss": 0.2512243716046214,
          "train_loss": 0.2568526566028595,
          "validation_loss": 0.10207486152648926,
          "train_metric": 0.05053595453500748,
          "validation_metric": 0.004211553372442722,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.29232279999996535
        },
        {
          "epoch": 18,
          "updates": 288,
          "samples_seen": 18432,
          "online_loss": 0.2479320429265499,
          "train_loss": 0.25000834465026855,
          "validation_loss": 0.104845330119133,
          "train_metric": 0.04878416657447815,
          "validation_metric": 0.005155506543815136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3043779209983768
        },
        {
          "epoch": 19,
          "updates": 304,
          "samples_seen": 19456,
          "online_loss": 0.24684269586578012,
          "train_loss": 0.2364780306816101,
          "validation_loss": 0.08693799376487732,
          "train_metric": 0.048987239599227905,
          "validation_metric": 0.004229792859405279,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3162453839995578
        },
        {
          "epoch": 20,
          "updates": 320,
          "samples_seen": 20480,
          "online_loss": 0.2433467493392527,
          "train_loss": 0.24024797976016998,
          "validation_loss": 0.08617004007101059,
          "train_metric": 0.04917970672249794,
          "validation_metric": 0.003082706592977047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3285749329988903
        },
        {
          "epoch": 21,
          "updates": 336,
          "samples_seen": 21504,
          "online_loss": 0.24184863944537938,
          "train_loss": 0.23633165657520294,
          "validation_loss": 0.08759104460477829,
          "train_metric": 0.04812173917889595,
          "validation_metric": 0.0031762351281940937,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3401326089988288
        },
        {
          "epoch": 22,
          "updates": 352,
          "samples_seen": 22528,
          "online_loss": 0.23378359246999025,
          "train_loss": 0.2306252419948578,
          "validation_loss": 0.08166950941085815,
          "train_metric": 0.04847177118062973,
          "validation_metric": 0.003424891969189048,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3519404669987125
        },
        {
          "epoch": 23,
          "updates": 368,
          "samples_seen": 23552,
          "online_loss": 0.23795864824205637,
          "train_loss": 0.23200398683547974,
          "validation_loss": 0.0811237320303917,
          "train_metric": 0.04818291962146759,
          "validation_metric": 0.00313880224712193,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.36384410399841727
        },
        {
          "epoch": 24,
          "updates": 384,
          "samples_seen": 24576,
          "online_loss": 0.23405286855995655,
          "train_loss": 0.2337406873703003,
          "validation_loss": 0.08165998756885529,
          "train_metric": 0.04887531325221062,
          "validation_metric": 0.0029560914263129234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.37576705399987986
        },
        {
          "epoch": 25,
          "updates": 400,
          "samples_seen": 25600,
          "online_loss": 0.2326295105740428,
          "train_loss": 0.2392570823431015,
          "validation_loss": 0.08719898760318756,
          "train_metric": 0.04775833711028099,
          "validation_metric": 0.0030461065471172333,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3878182709995599
        },
        {
          "epoch": 26,
          "updates": 416,
          "samples_seen": 26624,
          "online_loss": 0.23659307090565562,
          "train_loss": 0.2261638641357422,
          "validation_loss": 0.07546182721853256,
          "train_metric": 0.04825703799724579,
          "validation_metric": 0.0029827929101884365,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.3995839719991636
        },
        {
          "epoch": 27,
          "updates": 432,
          "samples_seen": 27648,
          "online_loss": 0.23197325319051743,
          "train_loss": 0.22714591026306152,
          "validation_loss": 0.07464869320392609,
          "train_metric": 0.04844098538160324,
          "validation_metric": 0.0027517015114426613,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.41177004399833095
        },
        {
          "epoch": 28,
          "updates": 448,
          "samples_seen": 28672,
          "online_loss": 0.23030814854428172,
          "train_loss": 0.22367779910564423,
          "validation_loss": 0.0748310461640358,
          "train_metric": 0.04794502258300781,
          "validation_metric": 0.0029725199565291405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4235357130000921
        },
        {
          "epoch": 29,
          "updates": 464,
          "samples_seen": 29696,
          "online_loss": 0.2280875612050295,
          "train_loss": 0.2205941379070282,
          "validation_loss": 0.07244547456502914,
          "train_metric": 0.04775969684123993,
          "validation_metric": 0.003052150597795844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.4353499789995112
        },
        {
          "epoch": 30,
          "updates": 480,
          "samples_seen": 30720,
          "online_loss": 0.22080569295212626,
          "train_loss": 0.23312188684940338,
          "validation_loss": 0.07999424636363983,
          "train_metric": 0.048556938767433167,
          "validation_metric": 0.00220252713188529,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.44797997499881603
        },
        {
          "epoch": 31,
          "updates": 496,
          "samples_seen": 31744,
          "online_loss": 0.2280974588356912,
          "train_loss": 0.22735106945037842,
          "validation_loss": 0.0793030709028244,
          "train_metric": 0.04723707586526871,
          "validation_metric": 0.0030811834149062634,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.45996394099893223
        },
        {
          "epoch": 32,
          "updates": 512,
          "samples_seen": 32768,
          "online_loss": 0.22786636091768742,
          "train_loss": 0.22102266550064087,
          "validation_loss": 0.06875717639923096,
          "train_metric": 0.04829474911093712,
          "validation_metric": 0.002316453028470278,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.47184463999838044
        },
        {
          "epoch": 33,
          "updates": 528,
          "samples_seen": 33792,
          "online_loss": 0.22688776068389416,
          "train_loss": 0.21438203752040863,
          "validation_loss": 0.06466645002365112,
          "train_metric": 0.04747706651687622,
          "validation_metric": 0.0025155930779874325,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.483290626998496
        },
        {
          "epoch": 34,
          "updates": 544,
          "samples_seen": 34816,
          "online_loss": 0.2216674811206758,
          "train_loss": 0.2195335179567337,
          "validation_loss": 0.0724039152264595,
          "train_metric": 0.04760873690247536,
          "validation_metric": 0.0030865271110087633,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.495001372999468
        },
        {
          "epoch": 35,
          "updates": 560,
          "samples_seen": 35840,
          "online_loss": 0.21665675332769752,
          "train_loss": 0.21446160972118378,
          "validation_loss": 0.060734089463949203,
          "train_metric": 0.04767289012670517,
          "validation_metric": 0.002298670820891857,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5068699899984495
        },
        {
          "epoch": 36,
          "updates": 576,
          "samples_seen": 36864,
          "online_loss": 0.21638595312833786,
          "train_loss": 0.21308386325836182,
          "validation_loss": 0.06241282448172569,
          "train_metric": 0.047579944133758545,
          "validation_metric": 0.002154245274141431,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5189202309993561
        },
        {
          "epoch": 37,
          "updates": 592,
          "samples_seen": 37888,
          "online_loss": 0.21540275868028402,
          "train_loss": 0.21743151545524597,
          "validation_loss": 0.06684280931949615,
          "train_metric": 0.04722532257437706,
          "validation_metric": 0.0019557580817490816,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5313440099998843
        },
        {
          "epoch": 38,
          "updates": 608,
          "samples_seen": 38912,
          "online_loss": 0.22147082816809416,
          "train_loss": 0.2167816311120987,
          "validation_loss": 0.06947781890630722,
          "train_metric": 0.04738156124949455,
          "validation_metric": 0.002520181704312563,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5431847599993489
        },
        {
          "epoch": 39,
          "updates": 624,
          "samples_seen": 39936,
          "online_loss": 0.2139184116385877,
          "train_loss": 0.210571750998497,
          "validation_loss": 0.06123445928096771,
          "train_metric": 0.047273341566324234,
          "validation_metric": 0.0018834895454347134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5553458779995708
        },
        {
          "epoch": 40,
          "updates": 640,
          "samples_seen": 40960,
          "online_loss": 0.2141619671601802,
          "train_loss": 0.21511176228523254,
          "validation_loss": 0.06762804836034775,
          "train_metric": 0.04715782031416893,
          "validation_metric": 0.0022802872117608786,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5665829219997249
        },
        {
          "epoch": 41,
          "updates": 656,
          "samples_seen": 41984,
          "online_loss": 0.2118777702562511,
          "train_loss": 0.20636384189128876,
          "validation_loss": 0.05811159685254097,
          "train_metric": 0.04724687710404396,
          "validation_metric": 0.0022009366657584906,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.5780821499993181
        },
        {
          "epoch": 42,
          "updates": 672,
          "samples_seen": 43008,
          "online_loss": 0.20811071805655956,
          "train_loss": 0.2119808793067932,
          "validation_loss": 0.061893168836832047,
          "train_metric": 0.047107335180044174,
          "validation_metric": 0.0018731859745457768,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.590220103998945
        },
        {
          "epoch": 43,
          "updates": 688,
          "samples_seen": 44032,
          "online_loss": 0.2117788982577622,
          "train_loss": 0.21380840241909027,
          "validation_loss": 0.06057283282279968,
          "train_metric": 0.04674592986702919,
          "validation_metric": 0.00211644871160388,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6022298999996565
        },
        {
          "epoch": 44,
          "updates": 704,
          "samples_seen": 45056,
          "online_loss": 0.2125899000093341,
          "train_loss": 0.20993030071258545,
          "validation_loss": 0.05681420490145683,
          "train_metric": 0.04714091494679451,
          "validation_metric": 0.001921318587847054,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6142919699996128
        },
        {
          "epoch": 45,
          "updates": 720,
          "samples_seen": 46080,
          "online_loss": 0.21322494326159358,
          "train_loss": 0.2068510204553604,
          "validation_loss": 0.05457423999905586,
          "train_metric": 0.04729732498526573,
          "validation_metric": 0.0016694929217919707,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6267623699986871
        },
        {
          "epoch": 46,
          "updates": 736,
          "samples_seen": 47104,
          "online_loss": 0.2073088944889605,
          "train_loss": 0.20549532771110535,
          "validation_loss": 0.05432694032788277,
          "train_metric": 0.046906933188438416,
          "validation_metric": 0.0016381099121645093,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6390760049998789
        },
        {
          "epoch": 47,
          "updates": 752,
          "samples_seen": 48128,
          "online_loss": 0.20774402422830462,
          "train_loss": 0.2121511697769165,
          "validation_loss": 0.060550201684236526,
          "train_metric": 0.046695053577423096,
          "validation_metric": 0.0013748517958447337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.6678836979990592
        },
        {
          "epoch": 48,
          "updates": 768,
          "samples_seen": 49152,
          "online_loss": 0.2076642168685794,
          "train_loss": 0.20584097504615784,
          "validation_loss": 0.05525115504860878,
          "train_metric": 0.04724966734647751,
          "validation_metric": 0.0012702749809250236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7158681329983665
        },
        {
          "epoch": 49,
          "updates": 784,
          "samples_seen": 50176,
          "online_loss": 0.20466096699237823,
          "train_loss": 0.20636683702468872,
          "validation_loss": 0.05336374789476395,
          "train_metric": 0.047378458082675934,
          "validation_metric": 0.001237576361745596,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.7588906639994093
        },
        {
          "epoch": 50,
          "updates": 800,
          "samples_seen": 51200,
          "online_loss": 0.20255353674292564,
          "train_loss": 0.20438210666179657,
          "validation_loss": 0.05300099775195122,
          "train_metric": 0.04737161099910736,
          "validation_metric": 0.0012189724948257208,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8105300389997865
        },
        {
          "epoch": 51,
          "updates": 816,
          "samples_seen": 52224,
          "online_loss": 0.21072225482203066,
          "train_loss": 0.20672613382339478,
          "validation_loss": 0.05608496814966202,
          "train_metric": 0.04741101711988449,
          "validation_metric": 0.0014296503504738212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.8554578899984335
        },
        {
          "epoch": 52,
          "updates": 832,
          "samples_seen": 53248,
          "online_loss": 0.20534564927220345,
          "train_loss": 0.20284093916416168,
          "validation_loss": 0.053668469190597534,
          "train_metric": 0.046813998371362686,
          "validation_metric": 0.001608812133781612,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9011750429999665
        },
        {
          "epoch": 53,
          "updates": 848,
          "samples_seen": 54272,
          "online_loss": 0.20225003641098738,
          "train_loss": 0.20705145597457886,
          "validation_loss": 0.05794133245944977,
          "train_metric": 0.046793993562459946,
          "validation_metric": 0.00193605269305408,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9494296339998982
        },
        {
          "epoch": 54,
          "updates": 864,
          "samples_seen": 55296,
          "online_loss": 0.2054118812084198,
          "train_loss": 0.2007124125957489,
          "validation_loss": 0.04818487912416458,
          "train_metric": 0.04683458432555199,
          "validation_metric": 0.0013743528397753835,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 0.9920029599998088
        },
        {
          "epoch": 55,
          "updates": 880,
          "samples_seen": 56320,
          "online_loss": 0.20170691516250372,
          "train_loss": 0.19870200753211975,
          "validation_loss": 0.050496555864810944,
          "train_metric": 0.046595752239227295,
          "validation_metric": 0.0016040427144616842,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0343608019993553
        },
        {
          "epoch": 56,
          "updates": 896,
          "samples_seen": 57344,
          "online_loss": 0.20177775947377086,
          "train_loss": 0.19650103151798248,
          "validation_loss": 0.04691728949546814,
          "train_metric": 0.04672941192984581,
          "validation_metric": 0.001500184298492968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.0772543359998963
        },
        {
          "epoch": 57,
          "updates": 912,
          "samples_seen": 58368,
          "online_loss": 0.20726678473874927,
          "train_loss": 0.20690205693244934,
          "validation_loss": 0.05352145805954933,
          "train_metric": 0.04642737656831741,
          "validation_metric": 0.0011526867747306824,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1252791099996102
        },
        {
          "epoch": 58,
          "updates": 928,
          "samples_seen": 59392,
          "online_loss": 0.2055054847151041,
          "train_loss": 0.19964499771595,
          "validation_loss": 0.0477788932621479,
          "train_metric": 0.04655740410089493,
          "validation_metric": 0.0010470098350197077,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.1738012229998276
        },
        {
          "epoch": 59,
          "updates": 944,
          "samples_seen": 60416,
          "online_loss": 0.20181710552424192,
          "train_loss": 0.20081280171871185,
          "validation_loss": 0.0512363463640213,
          "train_metric": 0.04661919176578522,
          "validation_metric": 0.0014872618485242128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2178758409991133
        },
        {
          "epoch": 60,
          "updates": 960,
          "samples_seen": 61440,
          "online_loss": 0.20031364518217742,
          "train_loss": 0.19650386273860931,
          "validation_loss": 0.045928340405225754,
          "train_metric": 0.04643994942307472,
          "validation_metric": 0.0011656396090984344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2581601969995972
        },
        {
          "epoch": 61,
          "updates": 976,
          "samples_seen": 62464,
          "online_loss": 0.1996535004582256,
          "train_loss": 0.20247763395309448,
          "validation_loss": 0.054922349750995636,
          "train_metric": 0.04625309631228447,
          "validation_metric": 0.001161169377155602,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.2872856519989
        },
        {
          "epoch": 62,
          "updates": 992,
          "samples_seen": 63488,
          "online_loss": 0.20236677560023963,
          "train_loss": 0.1957172304391861,
          "validation_loss": 0.043594926595687866,
          "train_metric": 0.046820417046546936,
          "validation_metric": 0.0008668869850225747,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3058113849983783
        },
        {
          "epoch": 63,
          "updates": 1008,
          "samples_seen": 64512,
          "online_loss": 0.19883925979956985,
          "train_loss": 0.19540059566497803,
          "validation_loss": 0.04437372833490372,
          "train_metric": 0.04636301100254059,
          "validation_metric": 0.0011027251603081822,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.322905252000055
        },
        {
          "epoch": 64,
          "updates": 1024,
          "samples_seen": 65536,
          "online_loss": 0.199501795694232,
          "train_loss": 0.2033630907535553,
          "validation_loss": 0.05042165517807007,
          "train_metric": 0.04711848869919777,
          "validation_metric": 0.0009230801952071488,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3400130879999779
        },
        {
          "epoch": 65,
          "updates": 1040,
          "samples_seen": 66560,
          "online_loss": 0.19768179510720074,
          "train_loss": 0.2005590796470642,
          "validation_loss": 0.05099305883049965,
          "train_metric": 0.046451568603515625,
          "validation_metric": 0.0009265170083381236,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.35722763099875
        },
        {
          "epoch": 66,
          "updates": 1056,
          "samples_seen": 67584,
          "online_loss": 0.19698956236243248,
          "train_loss": 0.20046047866344452,
          "validation_loss": 0.04666176065802574,
          "train_metric": 0.045972928404808044,
          "validation_metric": 0.001178522128611803,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3782991469997796
        },
        {
          "epoch": 67,
          "updates": 1072,
          "samples_seen": 68608,
          "online_loss": 0.19692633533850312,
          "train_loss": 0.19496040046215057,
          "validation_loss": 0.044584400951862335,
          "train_metric": 0.04616134986281395,
          "validation_metric": 0.000854743062518537,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.3965084169994952
        },
        {
          "epoch": 68,
          "updates": 1088,
          "samples_seen": 69632,
          "online_loss": 0.19451817031949759,
          "train_loss": 0.1956029087305069,
          "validation_loss": 0.044957321137189865,
          "train_metric": 0.04654087498784065,
          "validation_metric": 0.001269470201805234,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.414501149998614
        },
        {
          "epoch": 69,
          "updates": 1104,
          "samples_seen": 70656,
          "online_loss": 0.20069015643093735,
          "train_loss": 0.2027534395456314,
          "validation_loss": 0.051834944635629654,
          "train_metric": 0.047014690935611725,
          "validation_metric": 0.0008152832160703838,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4325186789992586
        },
        {
          "epoch": 70,
          "updates": 1120,
          "samples_seen": 71680,
          "online_loss": 0.19792360486462712,
          "train_loss": 0.19730286300182343,
          "validation_loss": 0.046701498329639435,
          "train_metric": 0.04666723683476448,
          "validation_metric": 0.0008481592522002757,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.44956691599873
        },
        {
          "epoch": 71,
          "updates": 1136,
          "samples_seen": 72704,
          "online_loss": 0.19396262103691697,
          "train_loss": 0.19111032783985138,
          "validation_loss": 0.04180428758263588,
          "train_metric": 0.04612411558628082,
          "validation_metric": 0.000934368057642132,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4697981199988135
        },
        {
          "epoch": 72,
          "updates": 1152,
          "samples_seen": 73728,
          "online_loss": 0.19237481732852757,
          "train_loss": 0.19498933851718903,
          "validation_loss": 0.04475674405694008,
          "train_metric": 0.046819694340229034,
          "validation_metric": 0.0008244456257671118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.4935504009990836
        },
        {
          "epoch": 73,
          "updates": 1168,
          "samples_seen": 74752,
          "online_loss": 0.1918631405569613,
          "train_loss": 0.1900683045387268,
          "validation_loss": 0.03930334746837616,
          "train_metric": 0.04657519608736038,
          "validation_metric": 0.0007139115477912128,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5110228220000863
        },
        {
          "epoch": 74,
          "updates": 1184,
          "samples_seen": 75776,
          "online_loss": 0.19029873958788812,
          "train_loss": 0.18938687443733215,
          "validation_loss": 0.037180352956056595,
          "train_metric": 0.046172257512807846,
          "validation_metric": 0.0009737745858728886,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.52743037899927
        },
        {
          "epoch": 75,
          "updates": 1200,
          "samples_seen": 76800,
          "online_loss": 0.1908117043785751,
          "train_loss": 0.19503635168075562,
          "validation_loss": 0.04080701246857643,
          "train_metric": 0.04650472477078438,
          "validation_metric": 0.0010979275684803724,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5437030369994318
        },
        {
          "epoch": 76,
          "updates": 1216,
          "samples_seen": 77824,
          "online_loss": 0.193122434662655,
          "train_loss": 0.19426219165325165,
          "validation_loss": 0.04706550016999245,
          "train_metric": 0.04621501266956329,
          "validation_metric": 0.0010055886814370751,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5610463009998057
        },
        {
          "epoch": 77,
          "updates": 1232,
          "samples_seen": 78848,
          "online_loss": 0.18947952846065164,
          "train_loss": 0.19311228394508362,
          "validation_loss": 0.04051375016570091,
          "train_metric": 0.04606160521507263,
          "validation_metric": 0.0010196496732532978,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.5784475459986425
        },
        {
          "epoch": 78,
          "updates": 1248,
          "samples_seen": 79872,
          "online_loss": 0.19203729089349508,
          "train_loss": 0.19186630845069885,
          "validation_loss": 0.04109564051032066,
          "train_metric": 0.0465460941195488,
          "validation_metric": 0.0008051975746639073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.6863447010000527
        },
        {
          "epoch": 79,
          "updates": 1264,
          "samples_seen": 80896,
          "online_loss": 0.19234563712961972,
          "train_loss": 0.19566164910793304,
          "validation_loss": 0.04604481905698776,
          "train_metric": 0.04642017185688019,
          "validation_metric": 0.0011532400967553258,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.702941268998984
        },
        {
          "epoch": 80,
          "updates": 1280,
          "samples_seen": 81920,
          "online_loss": 0.19282050780020654,
          "train_loss": 0.1877191811800003,
          "validation_loss": 0.037170104682445526,
          "train_metric": 0.04630400612950325,
          "validation_metric": 0.0007218808750621974,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7191504449983768
        },
        {
          "epoch": 81,
          "updates": 1296,
          "samples_seen": 82944,
          "online_loss": 0.1953882467932999,
          "train_loss": 0.2001970112323761,
          "validation_loss": 0.04888308420777321,
          "train_metric": 0.046330977231264114,
          "validation_metric": 0.000625430082436651,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7360924360000354
        },
        {
          "epoch": 82,
          "updates": 1312,
          "samples_seen": 83968,
          "online_loss": 0.19597084168344736,
          "train_loss": 0.18908695876598358,
          "validation_loss": 0.038753993809223175,
          "train_metric": 0.04587369039654732,
          "validation_metric": 0.0006666954141110182,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7526435459985805
        },
        {
          "epoch": 83,
          "updates": 1328,
          "samples_seen": 84992,
          "online_loss": 0.18803060264326632,
          "train_loss": 0.18591593205928802,
          "validation_loss": 0.03523325175046921,
          "train_metric": 0.046299856156110764,
          "validation_metric": 0.0008029733435250819,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7684240319995297
        },
        {
          "epoch": 84,
          "updates": 1344,
          "samples_seen": 86016,
          "online_loss": 0.1884416462853551,
          "train_loss": 0.18994368612766266,
          "validation_loss": 0.03706507757306099,
          "train_metric": 0.04598996043205261,
          "validation_metric": 0.0009443620801903307,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.7845216459991207
        },
        {
          "epoch": 85,
          "updates": 1360,
          "samples_seen": 87040,
          "online_loss": 0.1884297039359808,
          "train_loss": 0.19009603559970856,
          "validation_loss": 0.038372963666915894,
          "train_metric": 0.046830371022224426,
          "validation_metric": 0.0006216329638846219,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.801383762000114
        },
        {
          "epoch": 86,
          "updates": 1376,
          "samples_seen": 88064,
          "online_loss": 0.19165403093211353,
          "train_loss": 0.19263914227485657,
          "validation_loss": 0.03952022269368172,
          "train_metric": 0.046157971024513245,
          "validation_metric": 0.0006854879902675748,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8176304139997228
        },
        {
          "epoch": 87,
          "updates": 1392,
          "samples_seen": 89088,
          "online_loss": 0.19102405989542603,
          "train_loss": 0.19337590038776398,
          "validation_loss": 0.04637022688984871,
          "train_metric": 0.045947056263685226,
          "validation_metric": 0.0007696454413235188,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8335800320001
        },
        {
          "epoch": 88,
          "updates": 1408,
          "samples_seen": 90112,
          "online_loss": 0.18728494178503752,
          "train_loss": 0.18276222050189972,
          "validation_loss": 0.034082699567079544,
          "train_metric": 0.04610259458422661,
          "validation_metric": 0.0006266506388783455,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8495579169994016
        },
        {
          "epoch": 89,
          "updates": 1424,
          "samples_seen": 91136,
          "online_loss": 0.18530678702518344,
          "train_loss": 0.18739695847034454,
          "validation_loss": 0.03766152262687683,
          "train_metric": 0.04623670130968094,
          "validation_metric": 0.0007315646507777274,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8656613589992048
        },
        {
          "epoch": 90,
          "updates": 1440,
          "samples_seen": 92160,
          "online_loss": 0.1862168211955577,
          "train_loss": 0.19014345109462738,
          "validation_loss": 0.04079370200634003,
          "train_metric": 0.046020884066820145,
          "validation_metric": 0.0007514212047681212,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8815082539986179
        },
        {
          "epoch": 91,
          "updates": 1456,
          "samples_seen": 93184,
          "online_loss": 0.1906776875257492,
          "train_loss": 0.1859208643436432,
          "validation_loss": 0.03190135955810547,
          "train_metric": 0.04607954993844032,
          "validation_metric": 0.0005700862966477871,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.8983152499986318
        },
        {
          "epoch": 92,
          "updates": 1472,
          "samples_seen": 94208,
          "online_loss": 0.18581579672172666,
          "train_loss": 0.18785609304904938,
          "validation_loss": 0.03805813193321228,
          "train_metric": 0.04682057350873947,
          "validation_metric": 0.0004944157553836703,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9152463519985758
        },
        {
          "epoch": 93,
          "updates": 1488,
          "samples_seen": 95232,
          "online_loss": 0.18556431378237903,
          "train_loss": 0.18332460522651672,
          "validation_loss": 0.03227845951914787,
          "train_metric": 0.046435315161943436,
          "validation_metric": 0.00047247865586541593,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.932457084998532
        },
        {
          "epoch": 94,
          "updates": 1504,
          "samples_seen": 96256,
          "online_loss": 0.1854980350472033,
          "train_loss": 0.18781928718090057,
          "validation_loss": 0.03576493635773659,
          "train_metric": 0.046700652688741684,
          "validation_metric": 0.00048329192213714123,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.948279876000015
        },
        {
          "epoch": 95,
          "updates": 1520,
          "samples_seen": 97280,
          "online_loss": 0.18660559179261327,
          "train_loss": 0.1895899772644043,
          "validation_loss": 0.041617441922426224,
          "train_metric": 0.04569278657436371,
          "validation_metric": 0.000545930175576359,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.96420815700003
        },
        {
          "epoch": 96,
          "updates": 1536,
          "samples_seen": 98304,
          "online_loss": 0.19090361101552844,
          "train_loss": 0.1853620558977127,
          "validation_loss": 0.03243931755423546,
          "train_metric": 0.04595545306801796,
          "validation_metric": 0.0004917210899293423,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9808630609986722
        },
        {
          "epoch": 97,
          "updates": 1552,
          "samples_seen": 99328,
          "online_loss": 0.18878985522314906,
          "train_loss": 0.18930958211421967,
          "validation_loss": 0.03487376868724823,
          "train_metric": 0.046196047216653824,
          "validation_metric": 0.0005908129969611764,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 1.9968483239990746
        },
        {
          "epoch": 98,
          "updates": 1568,
          "samples_seen": 100352,
          "online_loss": 0.19324494013562799,
          "train_loss": 0.192399799823761,
          "validation_loss": 0.04274846613407135,
          "train_metric": 0.04622120037674904,
          "validation_metric": 0.0006974508287385106,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0126298639988818
        },
        {
          "epoch": 99,
          "updates": 1584,
          "samples_seen": 101376,
          "online_loss": 0.1884185508824885,
          "train_loss": 0.18322941660881042,
          "validation_loss": 0.031021788716316223,
          "train_metric": 0.046436820179224014,
          "validation_metric": 0.00041757122380658984,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0289905319987156
        },
        {
          "epoch": 100,
          "updates": 1600,
          "samples_seen": 102400,
          "online_loss": 0.18339042318984866,
          "train_loss": 0.18090133368968964,
          "validation_loss": 0.03214104473590851,
          "train_metric": 0.04603223875164986,
          "validation_metric": 0.00047864284715615213,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.044679273998554
        },
        {
          "epoch": 101,
          "updates": 1616,
          "samples_seen": 103424,
          "online_loss": 0.18304682127200067,
          "train_loss": 0.18498501181602478,
          "validation_loss": 0.03767405077815056,
          "train_metric": 0.04580410569906235,
          "validation_metric": 0.0004897099570371211,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0600627019994135
        },
        {
          "epoch": 102,
          "updates": 1632,
          "samples_seen": 104448,
          "online_loss": 0.19281436735764146,
          "train_loss": 0.19236429035663605,
          "validation_loss": 0.04253395274281502,
          "train_metric": 0.04610734060406685,
          "validation_metric": 0.0004911445430479944,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.075885671998549
        },
        {
          "epoch": 103,
          "updates": 1648,
          "samples_seen": 105472,
          "online_loss": 0.1912304216530174,
          "train_loss": 0.18484064936637878,
          "validation_loss": 0.03240245580673218,
          "train_metric": 0.046601440757513046,
          "validation_metric": 0.00041662328294478357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.0921596769985626
        },
        {
          "epoch": 104,
          "updates": 1664,
          "samples_seen": 106496,
          "online_loss": 0.19045235426165164,
          "train_loss": 0.19116193056106567,
          "validation_loss": 0.04145804047584534,
          "train_metric": 0.04596510902047157,
          "validation_metric": 0.000621449900791049,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.107065425998371
        },
        {
          "epoch": 105,
          "updates": 1680,
          "samples_seen": 107520,
          "online_loss": 0.19043279881589115,
          "train_loss": 0.19109943509101868,
          "validation_loss": 0.0413692407310009,
          "train_metric": 0.0457201786339283,
          "validation_metric": 0.0004925524117425084,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1216664439998567
        },
        {
          "epoch": 106,
          "updates": 1696,
          "samples_seen": 108544,
          "online_loss": 0.191932815592736,
          "train_loss": 0.1931667923927307,
          "validation_loss": 0.04487920179963112,
          "train_metric": 0.04556669667363167,
          "validation_metric": 0.000678084441460669,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.136422722998759
        },
        {
          "epoch": 107,
          "updates": 1712,
          "samples_seen": 109568,
          "online_loss": 0.1899055684916675,
          "train_loss": 0.18506629765033722,
          "validation_loss": 0.03193402290344238,
          "train_metric": 0.046318475157022476,
          "validation_metric": 0.0003755744837690145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1522541439990164
        },
        {
          "epoch": 108,
          "updates": 1728,
          "samples_seen": 110592,
          "online_loss": 0.18557773507200181,
          "train_loss": 0.18421748280525208,
          "validation_loss": 0.030185982584953308,
          "train_metric": 0.04634192958474159,
          "validation_metric": 0.00040833550156094134,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1676666280000063
        },
        {
          "epoch": 109,
          "updates": 1744,
          "samples_seen": 111616,
          "online_loss": 0.18339949566870928,
          "train_loss": 0.1802888810634613,
          "validation_loss": 0.028424914926290512,
          "train_metric": 0.046473436057567596,
          "validation_metric": 0.0003681015223264694,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.183043335999173
        },
        {
          "epoch": 110,
          "updates": 1760,
          "samples_seen": 112640,
          "online_loss": 0.1861727898940444,
          "train_loss": 0.1862727552652359,
          "validation_loss": 0.03161996975541115,
          "train_metric": 0.046081334352493286,
          "validation_metric": 0.00041151634650304914,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.1976867279990984
        },
        {
          "epoch": 111,
          "updates": 1776,
          "samples_seen": 113664,
          "online_loss": 0.19139358354732394,
          "train_loss": 0.1837170273065567,
          "validation_loss": 0.036010801792144775,
          "train_metric": 0.046030689030885696,
          "validation_metric": 0.00041242342558689415,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2120553409986314
        },
        {
          "epoch": 112,
          "updates": 1792,
          "samples_seen": 114688,
          "online_loss": 0.18684841319918633,
          "train_loss": 0.20013824105262756,
          "validation_loss": 0.04579143226146698,
          "train_metric": 0.04703010246157646,
          "validation_metric": 0.0006211191648617387,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.226581974999135
        },
        {
          "epoch": 113,
          "updates": 1808,
          "samples_seen": 115712,
          "online_loss": 0.19207738572731614,
          "train_loss": 0.19199970364570618,
          "validation_loss": 0.043280038982629776,
          "train_metric": 0.04602595418691635,
          "validation_metric": 0.0005949246697127819,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2411933629991836
        },
        {
          "epoch": 114,
          "updates": 1824,
          "samples_seen": 116736,
          "online_loss": 0.18845410039648414,
          "train_loss": 0.19100074470043182,
          "validation_loss": 0.039938785135746,
          "train_metric": 0.04678554832935333,
          "validation_metric": 0.0004830146790482104,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2555837189993326
        },
        {
          "epoch": 115,
          "updates": 1840,
          "samples_seen": 117760,
          "online_loss": 0.18445199076086283,
          "train_loss": 0.18014942109584808,
          "validation_loss": 0.028296135365962982,
          "train_metric": 0.046292051672935486,
          "validation_metric": 0.0004121385863982141,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2698162329998013
        },
        {
          "epoch": 116,
          "updates": 1856,
          "samples_seen": 118784,
          "online_loss": 0.18249386502429843,
          "train_loss": 0.18152441084384918,
          "validation_loss": 0.032454900443553925,
          "train_metric": 0.04588548094034195,
          "validation_metric": 0.0004220905539114028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2841875629983406
        },
        {
          "epoch": 117,
          "updates": 1872,
          "samples_seen": 119808,
          "online_loss": 0.17937904433347285,
          "train_loss": 0.1791139394044876,
          "validation_loss": 0.026317179203033447,
          "train_metric": 0.04608255997300148,
          "validation_metric": 0.00044647674076259136,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.2987908049999533
        },
        {
          "epoch": 118,
          "updates": 1888,
          "samples_seen": 120832,
          "online_loss": 0.18007237871643156,
          "train_loss": 0.1790529489517212,
          "validation_loss": 0.028926201164722443,
          "train_metric": 0.04622650891542435,
          "validation_metric": 0.00033531890949234366,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3139919309996912
        },
        {
          "epoch": 119,
          "updates": 1904,
          "samples_seen": 121856,
          "online_loss": 0.18448241357691586,
          "train_loss": 0.18232880532741547,
          "validation_loss": 0.031175704672932625,
          "train_metric": 0.046218812465667725,
          "validation_metric": 0.0003919742302969098,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.328738820999206
        },
        {
          "epoch": 120,
          "updates": 1920,
          "samples_seen": 122880,
          "online_loss": 0.18686964199878275,
          "train_loss": 0.19318972527980804,
          "validation_loss": 0.04312494024634361,
          "train_metric": 0.047534726560115814,
          "validation_metric": 0.0007744049653410912,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3431228089993965
        },
        {
          "epoch": 121,
          "updates": 1936,
          "samples_seen": 123904,
          "online_loss": 0.187184804584831,
          "train_loss": 0.18020497262477875,
          "validation_loss": 0.03085777536034584,
          "train_metric": 0.046087462455034256,
          "validation_metric": 0.0003397597174625844,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.357472568999583
        },
        {
          "epoch": 122,
          "updates": 1952,
          "samples_seen": 124928,
          "online_loss": 0.18248726148158312,
          "train_loss": 0.18221525847911835,
          "validation_loss": 0.034777455031871796,
          "train_metric": 0.04585389047861099,
          "validation_metric": 0.0004735231923405081,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.3723124699990876
        },
        {
          "epoch": 123,
          "updates": 1968,
          "samples_seen": 125952,
          "online_loss": 0.1798487352207303,
          "train_loss": 0.18141774833202362,
          "validation_loss": 0.030564680695533752,
          "train_metric": 0.0463879257440567,
          "validation_metric": 0.00034985283855348825,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.386954699999478
        },
        {
          "epoch": 124,
          "updates": 1984,
          "samples_seen": 126976,
          "online_loss": 0.18375125201418996,
          "train_loss": 0.18553733825683594,
          "validation_loss": 0.038469549268484116,
          "train_metric": 0.04618044197559357,
          "validation_metric": 0.00043612642912194133,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.401311381998312
        },
        {
          "epoch": 125,
          "updates": 2000,
          "samples_seen": 128000,
          "online_loss": 0.1820525696966797,
          "train_loss": 0.18992102146148682,
          "validation_loss": 0.03709185868501663,
          "train_metric": 0.04666827619075775,
          "validation_metric": 0.0005637868889607489,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.415736269998888
        },
        {
          "epoch": 126,
          "updates": 2016,
          "samples_seen": 129024,
          "online_loss": 0.18811462773010135,
          "train_loss": 0.18236124515533447,
          "validation_loss": 0.03428847715258598,
          "train_metric": 0.04632088541984558,
          "validation_metric": 0.0003549543325789273,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4301781429985567
        },
        {
          "epoch": 127,
          "updates": 2032,
          "samples_seen": 130048,
          "online_loss": 0.18810447468422353,
          "train_loss": 0.18049347400665283,
          "validation_loss": 0.027615100145339966,
          "train_metric": 0.04594146087765694,
          "validation_metric": 0.00029947227449156344,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.444892413999696
        },
        {
          "epoch": 128,
          "updates": 2048,
          "samples_seen": 131072,
          "online_loss": 0.18240067991428077,
          "train_loss": 0.17906096577644348,
          "validation_loss": 0.025513460859656334,
          "train_metric": 0.04601209983229637,
          "validation_metric": 0.0003215684846509248,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.458935927999846
        },
        {
          "epoch": 129,
          "updates": 2064,
          "samples_seen": 132096,
          "online_loss": 0.179812157060951,
          "train_loss": 0.17727568745613098,
          "validation_loss": 0.030530445277690887,
          "train_metric": 0.04588871821761131,
          "validation_metric": 0.0003314850036986172,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.4731810169996606
        },
        {
          "epoch": 130,
          "updates": 2080,
          "samples_seen": 133120,
          "online_loss": 0.17831406882032752,
          "train_loss": 0.18118003010749817,
          "validation_loss": 0.030152248218655586,
          "train_metric": 0.046568017452955246,
          "validation_metric": 0.00040726116276346147,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.487470691999988
        },
        {
          "epoch": 131,
          "updates": 2096,
          "samples_seen": 134144,
          "online_loss": 0.1797321131452918,
          "train_loss": 0.17804332077503204,
          "validation_loss": 0.023624999448657036,
          "train_metric": 0.0459638312458992,
          "validation_metric": 0.0002950861817225814,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.502054105998468
        },
        {
          "epoch": 132,
          "updates": 2112,
          "samples_seen": 135168,
          "online_loss": 0.1803583784494549,
          "train_loss": 0.17929010093212128,
          "validation_loss": 0.026178840547800064,
          "train_metric": 0.04636799916625023,
          "validation_metric": 0.00028171256417408586,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.516727598998841
        },
        {
          "epoch": 133,
          "updates": 2128,
          "samples_seen": 136192,
          "online_loss": 0.18151052668690681,
          "train_loss": 0.17823563516139984,
          "validation_loss": 0.028285562992095947,
          "train_metric": 0.046101637184619904,
          "validation_metric": 0.0003160958585795015,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.5309218689999398
        },
        {
          "epoch": 134,
          "updates": 2144,
          "samples_seen": 137216,
          "online_loss": 0.17948301741853356,
          "train_loss": 0.17698074877262115,
          "validation_loss": 0.024567589163780212,
          "train_metric": 0.04606013372540474,
          "validation_metric": 0.0002883430861402303,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.545120369999495
        },
        {
          "epoch": 135,
          "updates": 2160,
          "samples_seen": 138240,
          "online_loss": 0.17974429856985807,
          "train_loss": 0.17884957790374756,
          "validation_loss": 0.02599189430475235,
          "train_metric": 0.046079352498054504,
          "validation_metric": 0.00027998030418530107,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.559811147999426
        },
        {
          "epoch": 136,
          "updates": 2176,
          "samples_seen": 139264,
          "online_loss": 0.17919202998746186,
          "train_loss": 0.18373627960681915,
          "validation_loss": 0.028167758136987686,
          "train_metric": 0.046014454215765,
          "validation_metric": 0.0003503325569909066,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.574084545000005
        },
        {
          "epoch": 137,
          "updates": 2192,
          "samples_seen": 140288,
          "online_loss": 0.17857887595891953,
          "train_loss": 0.17869794368743896,
          "validation_loss": 0.02610158547759056,
          "train_metric": 0.0462796613574028,
          "validation_metric": 0.0003007825289387256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.588309977998506
        },
        {
          "epoch": 138,
          "updates": 2208,
          "samples_seen": 141312,
          "online_loss": 0.183066142257303,
          "train_loss": 0.18035288155078888,
          "validation_loss": 0.029794564470648766,
          "train_metric": 0.046383507549762726,
          "validation_metric": 0.0002637857978697866,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6032105359990965
        },
        {
          "epoch": 139,
          "updates": 2224,
          "samples_seen": 142336,
          "online_loss": 0.1769752367399633,
          "train_loss": 0.1754777580499649,
          "validation_loss": 0.026970669627189636,
          "train_metric": 0.0461050383746624,
          "validation_metric": 0.0003095809952355921,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.617684865999763
        },
        {
          "epoch": 140,
          "updates": 2240,
          "samples_seen": 143360,
          "online_loss": 0.1787069453857839,
          "train_loss": 0.1783042848110199,
          "validation_loss": 0.025417320430278778,
          "train_metric": 0.04564667120575905,
          "validation_metric": 0.0002925049338955432,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6315744609983085
        },
        {
          "epoch": 141,
          "updates": 2256,
          "samples_seen": 144384,
          "online_loss": 0.17985138250514865,
          "train_loss": 0.1784105747938156,
          "validation_loss": 0.025572387501597404,
          "train_metric": 0.046208880841732025,
          "validation_metric": 0.0002754836168605834,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6451958379984717
        },
        {
          "epoch": 142,
          "updates": 2272,
          "samples_seen": 145408,
          "online_loss": 0.17766721174120903,
          "train_loss": 0.17806540429592133,
          "validation_loss": 0.025973014533519745,
          "train_metric": 0.04564143717288971,
          "validation_metric": 0.0002675143477972597,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.659221329999127
        },
        {
          "epoch": 143,
          "updates": 2288,
          "samples_seen": 146432,
          "online_loss": 0.17835698300041258,
          "train_loss": 0.17874155938625336,
          "validation_loss": 0.028099676594138145,
          "train_metric": 0.046468306332826614,
          "validation_metric": 0.00024838102399371564,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.674028120998628
        },
        {
          "epoch": 144,
          "updates": 2304,
          "samples_seen": 147456,
          "online_loss": 0.1793772648088634,
          "train_loss": 0.17862652242183685,
          "validation_loss": 0.03112942911684513,
          "train_metric": 0.045914050191640854,
          "validation_metric": 0.0003141509077977389,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.6882850609999878
        },
        {
          "epoch": 145,
          "updates": 2320,
          "samples_seen": 148480,
          "online_loss": 0.17699681303929538,
          "train_loss": 0.17607222497463226,
          "validation_loss": 0.029062537476420403,
          "train_metric": 0.04585917294025421,
          "validation_metric": 0.0002954802184831351,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.702301371999056
        },
        {
          "epoch": 146,
          "updates": 2336,
          "samples_seen": 149504,
          "online_loss": 0.18150516506284475,
          "train_loss": 0.19076140224933624,
          "validation_loss": 0.04047885909676552,
          "train_metric": 0.04583856090903282,
          "validation_metric": 0.0006121405167505145,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7163584509999055
        },
        {
          "epoch": 147,
          "updates": 2352,
          "samples_seen": 150528,
          "online_loss": 0.18398145120590925,
          "train_loss": 0.17768576741218567,
          "validation_loss": 0.02414981834590435,
          "train_metric": 0.04621720686554909,
          "validation_metric": 0.00024331263557542115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.731256721999671
        },
        {
          "epoch": 148,
          "updates": 2368,
          "samples_seen": 151552,
          "online_loss": 0.1853088573552668,
          "train_loss": 0.18125267326831818,
          "validation_loss": 0.026119407266378403,
          "train_metric": 0.04643414914608002,
          "validation_metric": 0.0002750153362285346,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.7455583729988575
        },
        {
          "epoch": 149,
          "updates": 2384,
          "samples_seen": 152576,
          "online_loss": 0.18047995120286942,
          "train_loss": 0.17630012333393097,
          "validation_loss": 0.025683030486106873,
          "train_metric": 0.04612242802977562,
          "validation_metric": 0.0002435211936244741,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.760009222998633
        },
        {
          "epoch": 150,
          "updates": 2400,
          "samples_seen": 153600,
          "online_loss": 0.1779226716607809,
          "train_loss": 0.18870773911476135,
          "validation_loss": 0.03396005928516388,
          "train_metric": 0.045603785663843155,
          "validation_metric": 0.0005552324000746012,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.77443970099921
        },
        {
          "epoch": 151,
          "updates": 2416,
          "samples_seen": 154624,
          "online_loss": 0.18885222496464849,
          "train_loss": 0.18253611028194427,
          "validation_loss": 0.030721411108970642,
          "train_metric": 0.04553120210766792,
          "validation_metric": 0.0003184720117133111,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.788825920999443
        },
        {
          "epoch": 152,
          "updates": 2432,
          "samples_seen": 155648,
          "online_loss": 0.18149437196552753,
          "train_loss": 0.1774706095457077,
          "validation_loss": 0.028368806466460228,
          "train_metric": 0.04598063975572586,
          "validation_metric": 0.0002462603442836553,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.803040347998831
        },
        {
          "epoch": 153,
          "updates": 2448,
          "samples_seen": 156672,
          "online_loss": 0.17765640513971448,
          "train_loss": 0.1757843792438507,
          "validation_loss": 0.02432292141020298,
          "train_metric": 0.046211373060941696,
          "validation_metric": 0.00021269246644806117,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.817709611999817
        },
        {
          "epoch": 154,
          "updates": 2464,
          "samples_seen": 157696,
          "online_loss": 0.17673377855680883,
          "train_loss": 0.1760338842868805,
          "validation_loss": 0.026491990312933922,
          "train_metric": 0.04577714204788208,
          "validation_metric": 0.00029157823882997036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.832294946998445
        },
        {
          "epoch": 155,
          "updates": 2480,
          "samples_seen": 158720,
          "online_loss": 0.179470244795084,
          "train_loss": 0.18244752287864685,
          "validation_loss": 0.03126392886042595,
          "train_metric": 0.04546336084604263,
          "validation_metric": 0.00034512937418185174,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8463152869990154
        },
        {
          "epoch": 156,
          "updates": 2496,
          "samples_seen": 159744,
          "online_loss": 0.1839639120735228,
          "train_loss": 0.1815776377916336,
          "validation_loss": 0.0305103100836277,
          "train_metric": 0.045938316732645035,
          "validation_metric": 0.00032191857462748885,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8602961149990733
        },
        {
          "epoch": 157,
          "updates": 2512,
          "samples_seen": 160768,
          "online_loss": 0.1780435461550951,
          "train_loss": 0.18120647966861725,
          "validation_loss": 0.030739441514015198,
          "train_metric": 0.045940037816762924,
          "validation_metric": 0.0002334222081117332,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8744031339992944
        },
        {
          "epoch": 158,
          "updates": 2528,
          "samples_seen": 161792,
          "online_loss": 0.17893271264620125,
          "train_loss": 0.17465704679489136,
          "validation_loss": 0.027227936312556267,
          "train_metric": 0.0459660142660141,
          "validation_metric": 0.00023036195489112288,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.8888450609993015
        },
        {
          "epoch": 159,
          "updates": 2544,
          "samples_seen": 162816,
          "online_loss": 0.17662165220826864,
          "train_loss": 0.1845584213733673,
          "validation_loss": 0.030559614300727844,
          "train_metric": 0.04601689800620079,
          "validation_metric": 0.00043567109969444573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.902959743998508
        },
        {
          "epoch": 160,
          "updates": 2560,
          "samples_seen": 163840,
          "online_loss": 0.18687924835830927,
          "train_loss": 0.18048448860645294,
          "validation_loss": 0.029791828244924545,
          "train_metric": 0.04554128274321556,
          "validation_metric": 0.00031464864150620997,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9170528419999755
        },
        {
          "epoch": 161,
          "updates": 2576,
          "samples_seen": 164864,
          "online_loss": 0.1797365085221827,
          "train_loss": 0.1766897439956665,
          "validation_loss": 0.026734862476587296,
          "train_metric": 0.04585203155875206,
          "validation_metric": 0.00022739080304745585,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9315053509999416
        },
        {
          "epoch": 162,
          "updates": 2592,
          "samples_seen": 165888,
          "online_loss": 0.1776600987650454,
          "train_loss": 0.1771225929260254,
          "validation_loss": 0.023856204003095627,
          "train_metric": 0.046243637800216675,
          "validation_metric": 0.00028869908419437706,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.946131152999442
        },
        {
          "epoch": 163,
          "updates": 2608,
          "samples_seen": 166912,
          "online_loss": 0.1785817602649331,
          "train_loss": 0.1725664585828781,
          "validation_loss": 0.01868414506316185,
          "train_metric": 0.04598623514175415,
          "validation_metric": 0.0002207590005127713,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9602288579990272
        },
        {
          "epoch": 164,
          "updates": 2624,
          "samples_seen": 167936,
          "online_loss": 0.17391994828358293,
          "train_loss": 0.17530961334705353,
          "validation_loss": 0.024505099281668663,
          "train_metric": 0.04574986547231674,
          "validation_metric": 0.0002984365273732692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.97425779599871
        },
        {
          "epoch": 165,
          "updates": 2640,
          "samples_seen": 168960,
          "online_loss": 0.17609357833862305,
          "train_loss": 0.18049798905849457,
          "validation_loss": 0.029709164053201675,
          "train_metric": 0.046057309955358505,
          "validation_metric": 0.0002815928601194173,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 2.9884396309989825
        },
        {
          "epoch": 166,
          "updates": 2656,
          "samples_seen": 169984,
          "online_loss": 0.17686439165845513,
          "train_loss": 0.17384223639965057,
          "validation_loss": 0.024861106649041176,
          "train_metric": 0.04642808064818382,
          "validation_metric": 0.0002968921617139131,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0020980659992347
        },
        {
          "epoch": 167,
          "updates": 2672,
          "samples_seen": 171008,
          "online_loss": 0.17497884621843696,
          "train_loss": 0.17282752692699432,
          "validation_loss": 0.022148316726088524,
          "train_metric": 0.04585175961256027,
          "validation_metric": 0.0002017872320720926,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0164841480000177
        },
        {
          "epoch": 168,
          "updates": 2688,
          "samples_seen": 172032,
          "online_loss": 0.17348900996148586,
          "train_loss": 0.17221994698047638,
          "validation_loss": 0.02201193943619728,
          "train_metric": 0.04580102488398552,
          "validation_metric": 0.00019141485972795635,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.031255773999874
        },
        {
          "epoch": 169,
          "updates": 2704,
          "samples_seen": 173056,
          "online_loss": 0.1779123554006219,
          "train_loss": 0.1874139904975891,
          "validation_loss": 0.03247880935668945,
          "train_metric": 0.04623802751302719,
          "validation_metric": 0.00023835088359192014,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0453889399996115
        },
        {
          "epoch": 170,
          "updates": 2720,
          "samples_seen": 174080,
          "online_loss": 0.18121055816300213,
          "train_loss": 0.17820651829242706,
          "validation_loss": 0.02630014903843403,
          "train_metric": 0.04638349264860153,
          "validation_metric": 0.0002363758976571262,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0593164849997265
        },
        {
          "epoch": 171,
          "updates": 2736,
          "samples_seen": 175104,
          "online_loss": 0.18058136233594269,
          "train_loss": 0.17592647671699524,
          "validation_loss": 0.02819466032087803,
          "train_metric": 0.04620810225605965,
          "validation_metric": 0.0002178264840040356,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0734263979993557
        },
        {
          "epoch": 172,
          "updates": 2752,
          "samples_seen": 176128,
          "online_loss": 0.17945047514513135,
          "train_loss": 0.18446417152881622,
          "validation_loss": 0.03469570353627205,
          "train_metric": 0.04618711769580841,
          "validation_metric": 0.00033360879751853645,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0872829529998853
        },
        {
          "epoch": 173,
          "updates": 2768,
          "samples_seen": 177152,
          "online_loss": 0.17853210913017392,
          "train_loss": 0.18023720383644104,
          "validation_loss": 0.028090018779039383,
          "train_metric": 0.045904044061899185,
          "validation_metric": 0.0002698488242458552,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.0995975689984334
        },
        {
          "epoch": 174,
          "updates": 2784,
          "samples_seen": 178176,
          "online_loss": 0.1796227260492742,
          "train_loss": 0.17652027308940887,
          "validation_loss": 0.024826807901263237,
          "train_metric": 0.04634265601634979,
          "validation_metric": 0.00019963926752097905,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1099346989994956
        },
        {
          "epoch": 175,
          "updates": 2800,
          "samples_seen": 179200,
          "online_loss": 0.17594955884851515,
          "train_loss": 0.17149269580841064,
          "validation_loss": 0.02008371613919735,
          "train_metric": 0.045957162976264954,
          "validation_metric": 0.0001787161163520068,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.120338972999889
        },
        {
          "epoch": 176,
          "updates": 2816,
          "samples_seen": 180224,
          "online_loss": 0.1753619876690209,
          "train_loss": 0.18268714845180511,
          "validation_loss": 0.031290456652641296,
          "train_metric": 0.04705827310681343,
          "validation_metric": 0.0003969463868997991,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.1308582389992807
        },
        {
          "epoch": 177,
          "updates": 2832,
          "samples_seen": 181248,
          "online_loss": 0.18193532445002347,
          "train_loss": 0.19242097437381744,
          "validation_loss": 0.038402341306209564,
          "train_metric": 0.04553534463047981,
          "validation_metric": 0.0006237547495402396,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.140912907998427
        },
        {
          "epoch": 178,
          "updates": 2848,
          "samples_seen": 182272,
          "online_loss": 0.18199956696480513,
          "train_loss": 0.18187738955020905,
          "validation_loss": 0.030537348240613937,
          "train_metric": 0.045773811638355255,
          "validation_metric": 0.0004196099180262536,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.150669921998997
        },
        {
          "epoch": 179,
          "updates": 2864,
          "samples_seen": 183296,
          "online_loss": 0.1753149414435029,
          "train_loss": 0.17351986467838287,
          "validation_loss": 0.01986399106681347,
          "train_metric": 0.04581955820322037,
          "validation_metric": 0.0002464699500706047,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.160600475999672
        },
        {
          "epoch": 180,
          "updates": 2880,
          "samples_seen": 184320,
          "online_loss": 0.17388802487403154,
          "train_loss": 0.17210188508033752,
          "validation_loss": 0.02018100582063198,
          "train_metric": 0.046152371913194656,
          "validation_metric": 0.00017033772019203752,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.170866202999605
        },
        {
          "epoch": 181,
          "updates": 2896,
          "samples_seen": 185344,
          "online_loss": 0.17546344513539225,
          "train_loss": 0.1732049435377121,
          "validation_loss": 0.021542631089687347,
          "train_metric": 0.04629391059279442,
          "validation_metric": 0.00019731183419935405,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.181368730000031
        },
        {
          "epoch": 182,
          "updates": 2912,
          "samples_seen": 186368,
          "online_loss": 0.176187374163419,
          "train_loss": 0.177447110414505,
          "validation_loss": 0.027788739651441574,
          "train_metric": 0.04639844596385956,
          "validation_metric": 0.00020782359933946282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2590162659998896
        },
        {
          "epoch": 183,
          "updates": 2928,
          "samples_seen": 187392,
          "online_loss": 0.18039167346432805,
          "train_loss": 0.18945373594760895,
          "validation_loss": 0.040022313594818115,
          "train_metric": 0.04639066010713577,
          "validation_metric": 0.00030996001441963017,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.269027931999517
        },
        {
          "epoch": 184,
          "updates": 2944,
          "samples_seen": 188416,
          "online_loss": 0.17923489725217223,
          "train_loss": 0.1802673637866974,
          "validation_loss": 0.030468225479125977,
          "train_metric": 0.04603498429059982,
          "validation_metric": 0.0002901763073168695,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2795024269998976
        },
        {
          "epoch": 185,
          "updates": 2960,
          "samples_seen": 189440,
          "online_loss": 0.18020300334319472,
          "train_loss": 0.174956277012825,
          "validation_loss": 0.025829939171671867,
          "train_metric": 0.045804206281900406,
          "validation_metric": 0.00018140992324333638,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2894403979989875
        },
        {
          "epoch": 186,
          "updates": 2976,
          "samples_seen": 190464,
          "online_loss": 0.17449883744120598,
          "train_loss": 0.17161040008068085,
          "validation_loss": 0.018304014578461647,
          "train_metric": 0.0459936261177063,
          "validation_metric": 0.00015911583614069968,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.2998553509987687
        },
        {
          "epoch": 187,
          "updates": 2992,
          "samples_seen": 191488,
          "online_loss": 0.180181335657835,
          "train_loss": 0.17695483565330505,
          "validation_loss": 0.02488071843981743,
          "train_metric": 0.04563707113265991,
          "validation_metric": 0.00020544402650557458,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3101082439989113
        },
        {
          "epoch": 188,
          "updates": 3008,
          "samples_seen": 192512,
          "online_loss": 0.18243959569372237,
          "train_loss": 0.17448239028453827,
          "validation_loss": 0.020173639059066772,
          "train_metric": 0.04605598375201225,
          "validation_metric": 0.0002126361068803817,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3199039089995495
        },
        {
          "epoch": 189,
          "updates": 3024,
          "samples_seen": 193536,
          "online_loss": 0.1778879347257316,
          "train_loss": 0.1722775399684906,
          "validation_loss": 0.02332901395857334,
          "train_metric": 0.04612840339541435,
          "validation_metric": 0.000170545419678092,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3300125440000556
        },
        {
          "epoch": 190,
          "updates": 3040,
          "samples_seen": 194560,
          "online_loss": 0.1733678092714399,
          "train_loss": 0.17535190284252167,
          "validation_loss": 0.02410568855702877,
          "train_metric": 0.04656774178147316,
          "validation_metric": 0.0003270595334470272,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3396154940000997
        },
        {
          "epoch": 191,
          "updates": 3056,
          "samples_seen": 195584,
          "online_loss": 0.17865268187597394,
          "train_loss": 0.17680230736732483,
          "validation_loss": 0.024295300245285034,
          "train_metric": 0.045863814651966095,
          "validation_metric": 0.00020377099281176925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.34935807599868
        },
        {
          "epoch": 192,
          "updates": 3072,
          "samples_seen": 196608,
          "online_loss": 0.1781102237291634,
          "train_loss": 0.17303989827632904,
          "validation_loss": 0.0211764108389616,
          "train_metric": 0.04559393599629402,
          "validation_metric": 0.00021355324133764952,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3591493909989367
        },
        {
          "epoch": 193,
          "updates": 3088,
          "samples_seen": 197632,
          "online_loss": 0.1748636793345213,
          "train_loss": 0.17171576619148254,
          "validation_loss": 0.018392829224467278,
          "train_metric": 0.04586183279752731,
          "validation_metric": 0.0001871707063401118,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3685837519988127
        },
        {
          "epoch": 194,
          "updates": 3104,
          "samples_seen": 198656,
          "online_loss": 0.17519878526218235,
          "train_loss": 0.18114571273326874,
          "validation_loss": 0.03207824006676674,
          "train_metric": 0.046319786459207535,
          "validation_metric": 0.00035904685501009226,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.3788168399987626
        },
        {
          "epoch": 195,
          "updates": 3120,
          "samples_seen": 199680,
          "online_loss": 0.18163301376625896,
          "train_loss": 0.17754164338111877,
          "validation_loss": 0.025334211066365242,
          "train_metric": 0.04647672548890114,
          "validation_metric": 0.00017845221736934036,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.388326252999832
        },
        {
          "epoch": 196,
          "updates": 3136,
          "samples_seen": 200704,
          "online_loss": 0.17806775588542223,
          "train_loss": 0.1804700642824173,
          "validation_loss": 0.03239764645695686,
          "train_metric": 0.04602831229567528,
          "validation_metric": 0.00027293956372886896,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.397942648998651
        },
        {
          "epoch": 197,
          "updates": 3152,
          "samples_seen": 201728,
          "online_loss": 0.17692670971155167,
          "train_loss": 0.18561206758022308,
          "validation_loss": 0.03162449598312378,
          "train_metric": 0.04696434736251831,
          "validation_metric": 0.00048059955588541925,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4078312040001038
        },
        {
          "epoch": 198,
          "updates": 3168,
          "samples_seen": 202752,
          "online_loss": 0.1724033672362566,
          "train_loss": 0.17060281336307526,
          "validation_loss": 0.019199123606085777,
          "train_metric": 0.045751724392175674,
          "validation_metric": 0.00018197325698565692,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4173973580000165
        },
        {
          "epoch": 199,
          "updates": 3184,
          "samples_seen": 203776,
          "online_loss": 0.17338484129868448,
          "train_loss": 0.17032228410243988,
          "validation_loss": 0.018463343381881714,
          "train_metric": 0.046132445335388184,
          "validation_metric": 0.00015612343850079924,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4281849799990596
        },
        {
          "epoch": 200,
          "updates": 3200,
          "samples_seen": 204800,
          "online_loss": 0.17418215004727244,
          "train_loss": 0.1708221435546875,
          "validation_loss": 0.021249042823910713,
          "train_metric": 0.04617665708065033,
          "validation_metric": 0.00017018830112647265,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.437870258998373
        },
        {
          "epoch": 201,
          "updates": 3216,
          "samples_seen": 205824,
          "online_loss": 0.17284948378801346,
          "train_loss": 0.17310400307178497,
          "validation_loss": 0.023112380877137184,
          "train_metric": 0.04607830196619034,
          "validation_metric": 0.00020795634191017598,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4478519069998583
        },
        {
          "epoch": 202,
          "updates": 3232,
          "samples_seen": 206848,
          "online_loss": 0.17479289369657636,
          "train_loss": 0.17147228121757507,
          "validation_loss": 0.01679772324860096,
          "train_metric": 0.04601721838116646,
          "validation_metric": 0.00015167621313594282,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4583778339983837
        },
        {
          "epoch": 203,
          "updates": 3248,
          "samples_seen": 207872,
          "online_loss": 0.17454399121925235,
          "train_loss": 0.1764916330575943,
          "validation_loss": 0.026082277297973633,
          "train_metric": 0.04653388261795044,
          "validation_metric": 0.00018232769798487425,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.468094513998949
        },
        {
          "epoch": 204,
          "updates": 3264,
          "samples_seen": 208896,
          "online_loss": 0.17655382491648197,
          "train_loss": 0.1719675064086914,
          "validation_loss": 0.017716314643621445,
          "train_metric": 0.0457230880856514,
          "validation_metric": 0.00016329565551131964,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.477793317999385
        },
        {
          "epoch": 205,
          "updates": 3280,
          "samples_seen": 209920,
          "online_loss": 0.17256408976390958,
          "train_loss": 0.16815219819545746,
          "validation_loss": 0.017586324363946915,
          "train_metric": 0.045874252915382385,
          "validation_metric": 0.0001420334301656112,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4874681949986552
        },
        {
          "epoch": 206,
          "updates": 3296,
          "samples_seen": 210944,
          "online_loss": 0.1726267347112298,
          "train_loss": 0.16993799805641174,
          "validation_loss": 0.016271311789751053,
          "train_metric": 0.04573330283164978,
          "validation_metric": 0.00020185465109534562,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.4971313209989603
        },
        {
          "epoch": 207,
          "updates": 3312,
          "samples_seen": 211968,
          "online_loss": 0.17124091472942382,
          "train_loss": 0.169364333152771,
          "validation_loss": 0.018326599150896072,
          "train_metric": 0.0458328016102314,
          "validation_metric": 0.00013794958067592233,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5070361669986596
        },
        {
          "epoch": 208,
          "updates": 3328,
          "samples_seen": 212992,
          "online_loss": 0.17234040144830942,
          "train_loss": 0.1737666130065918,
          "validation_loss": 0.024014156311750412,
          "train_metric": 0.04570610821247101,
          "validation_metric": 0.000243265152676031,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5164560670000355
        },
        {
          "epoch": 209,
          "updates": 3344,
          "samples_seen": 214016,
          "online_loss": 0.1735172012122348,
          "train_loss": 0.18049167096614838,
          "validation_loss": 0.029357263818383217,
          "train_metric": 0.046467214822769165,
          "validation_metric": 0.00021332975302357227,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.526307785999961
        },
        {
          "epoch": 210,
          "updates": 3360,
          "samples_seen": 215040,
          "online_loss": 0.17845820216462016,
          "train_loss": 0.17731182277202606,
          "validation_loss": 0.025608062744140625,
          "train_metric": 0.04623030871152878,
          "validation_metric": 0.00020550766203086823,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5355842299995857
        },
        {
          "epoch": 211,
          "updates": 3376,
          "samples_seen": 216064,
          "online_loss": 0.17524578794836998,
          "train_loss": 0.1721387803554535,
          "validation_loss": 0.01785741187632084,
          "train_metric": 0.0458095483481884,
          "validation_metric": 0.0001628962781978771,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5452426349984307
        },
        {
          "epoch": 212,
          "updates": 3392,
          "samples_seen": 217088,
          "online_loss": 0.1703739138902165,
          "train_loss": 0.17488767206668854,
          "validation_loss": 0.021176403388381004,
          "train_metric": 0.04657793045043945,
          "validation_metric": 0.00023000054352451116,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.554508981998879
        },
        {
          "epoch": 213,
          "updates": 3408,
          "samples_seen": 218112,
          "online_loss": 0.18472300609573722,
          "train_loss": 0.19259056448936462,
          "validation_loss": 0.04128631576895714,
          "train_metric": 0.04641922190785408,
          "validation_metric": 0.00043955972068943083,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5643896469991887
        },
        {
          "epoch": 214,
          "updates": 3424,
          "samples_seen": 219136,
          "online_loss": 0.18417813233099878,
          "train_loss": 0.1756649762392044,
          "validation_loss": 0.02286606654524803,
          "train_metric": 0.04566895589232445,
          "validation_metric": 0.0001426443923264742,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.574373295999976
        },
        {
          "epoch": 215,
          "updates": 3440,
          "samples_seen": 220160,
          "online_loss": 0.17350526835070923,
          "train_loss": 0.17153221368789673,
          "validation_loss": 0.020399445667862892,
          "train_metric": 0.046306952834129333,
          "validation_metric": 0.00015316983626689762,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5840424509988225
        },
        {
          "epoch": 216,
          "updates": 3456,
          "samples_seen": 221184,
          "online_loss": 0.17454995639855042,
          "train_loss": 0.17185591161251068,
          "validation_loss": 0.018433142453432083,
          "train_metric": 0.04600067064166069,
          "validation_metric": 0.00014197103155311197,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.5939928889984003
        },
        {
          "epoch": 217,
          "updates": 3472,
          "samples_seen": 222208,
          "online_loss": 0.17365039372816682,
          "train_loss": 0.17368541657924652,
          "validation_loss": 0.022200990468263626,
          "train_metric": 0.045968420803546906,
          "validation_metric": 0.00014199016732163727,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6034858279999753
        },
        {
          "epoch": 218,
          "updates": 3488,
          "samples_seen": 223232,
          "online_loss": 0.1725686735007912,
          "train_loss": 0.17039312422275543,
          "validation_loss": 0.01905985176563263,
          "train_metric": 0.04624983295798302,
          "validation_metric": 0.00012270896695554256,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6137636019993806
        },
        {
          "epoch": 219,
          "updates": 3504,
          "samples_seen": 224256,
          "online_loss": 0.17024110769852996,
          "train_loss": 0.17053088545799255,
          "validation_loss": 0.01882140152156353,
          "train_metric": 0.04566873982548714,
          "validation_metric": 0.00018586580699775368,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.62363071999971
        },
        {
          "epoch": 220,
          "updates": 3520,
          "samples_seen": 225280,
          "online_loss": 0.17345195822417736,
          "train_loss": 0.16886571049690247,
          "validation_loss": 0.01705053448677063,
          "train_metric": 0.046054620295763016,
          "validation_metric": 0.00014339112385641783,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.633046517999901
        },
        {
          "epoch": 221,
          "updates": 3536,
          "samples_seen": 226304,
          "online_loss": 0.1745311104459688,
          "train_loss": 0.16875052452087402,
          "validation_loss": 0.017828620970249176,
          "train_metric": 0.046080395579338074,
          "validation_metric": 0.00013549206778407097,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6427726980000443
        },
        {
          "epoch": 222,
          "updates": 3552,
          "samples_seen": 227328,
          "online_loss": 0.1700847635511309,
          "train_loss": 0.17397767305374146,
          "validation_loss": 0.0196264386177063,
          "train_metric": 0.04651849344372749,
          "validation_metric": 0.00017757942259777337,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.65203060299973
        },
        {
          "epoch": 223,
          "updates": 3568,
          "samples_seen": 228352,
          "online_loss": 0.17334021779242903,
          "train_loss": 0.17607375979423523,
          "validation_loss": 0.02305087074637413,
          "train_metric": 0.04631150886416435,
          "validation_metric": 0.00021634227596223354,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6615853829989646
        },
        {
          "epoch": 224,
          "updates": 3584,
          "samples_seen": 229376,
          "online_loss": 0.17344366619363427,
          "train_loss": 0.17110730707645416,
          "validation_loss": 0.01843510940670967,
          "train_metric": 0.04585719481110573,
          "validation_metric": 0.0001540951052447781,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.670826452998881
        },
        {
          "epoch": 225,
          "updates": 3600,
          "samples_seen": 230400,
          "online_loss": 0.1733408742584288,
          "train_loss": 0.1696503907442093,
          "validation_loss": 0.015278110280632973,
          "train_metric": 0.04591439291834831,
          "validation_metric": 0.000158315830049105,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.6807365539989405
        },
        {
          "epoch": 226,
          "updates": 3616,
          "samples_seen": 231424,
          "online_loss": 0.176285607740283,
          "train_loss": 0.1729109287261963,
          "validation_loss": 0.018508877605199814,
          "train_metric": 0.04585086926817894,
          "validation_metric": 0.00013022498751524836,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.690186988998903
        },
        {
          "epoch": 227,
          "updates": 3632,
          "samples_seen": 232448,
          "online_loss": 0.17354943556711078,
          "train_loss": 0.17276133596897125,
          "validation_loss": 0.021533308550715446,
          "train_metric": 0.04626169800758362,
          "validation_metric": 0.00014633142563980073,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.699464341998464
        },
        {
          "epoch": 228,
          "updates": 3648,
          "samples_seen": 233472,
          "online_loss": 0.1738593813497573,
          "train_loss": 0.17827051877975464,
          "validation_loss": 0.02456817403435707,
          "train_metric": 0.046101029962301254,
          "validation_metric": 0.00036791845923289657,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7089013519998844
        },
        {
          "epoch": 229,
          "updates": 3664,
          "samples_seen": 234496,
          "online_loss": 0.1711067823925987,
          "train_loss": 0.17094115912914276,
          "validation_loss": 0.017219267785549164,
          "train_metric": 0.04590823873877525,
          "validation_metric": 0.00011880337115144357,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7184687349999876
        },
        {
          "epoch": 230,
          "updates": 3680,
          "samples_seen": 235520,
          "online_loss": 0.17387889930978417,
          "train_loss": 0.17216169834136963,
          "validation_loss": 0.01970960572361946,
          "train_metric": 0.045890115201473236,
          "validation_metric": 0.0001136346545536071,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.728883115998542
        },
        {
          "epoch": 231,
          "updates": 3696,
          "samples_seen": 236544,
          "online_loss": 0.17301891837269068,
          "train_loss": 0.17248724400997162,
          "validation_loss": 0.021121308207511902,
          "train_metric": 0.04645959660410881,
          "validation_metric": 0.00014723393542226404,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7381500149986095
        },
        {
          "epoch": 232,
          "updates": 3712,
          "samples_seen": 237568,
          "online_loss": 0.1697788720484823,
          "train_loss": 0.16785570979118347,
          "validation_loss": 0.014882884919643402,
          "train_metric": 0.04611717909574509,
          "validation_metric": 0.00010087719419971108,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7485206269993796
        },
        {
          "epoch": 233,
          "updates": 3728,
          "samples_seen": 238592,
          "online_loss": 0.17058148700743914,
          "train_loss": 0.1688472330570221,
          "validation_loss": 0.016851292923092842,
          "train_metric": 0.04598472639918327,
          "validation_metric": 0.0001477668556617573,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7580550929997116
        },
        {
          "epoch": 234,
          "updates": 3744,
          "samples_seen": 239616,
          "online_loss": 0.17375536682084203,
          "train_loss": 0.17208057641983032,
          "validation_loss": 0.017634497955441475,
          "train_metric": 0.04618449509143829,
          "validation_metric": 0.0001188433452625759,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.767180547998578
        },
        {
          "epoch": 235,
          "updates": 3760,
          "samples_seen": 240640,
          "online_loss": 0.176851951982826,
          "train_loss": 0.17293384671211243,
          "validation_loss": 0.022597238421440125,
          "train_metric": 0.04626302048563957,
          "validation_metric": 0.00014114832447376102,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.776610881999659
        },
        {
          "epoch": 236,
          "updates": 3776,
          "samples_seen": 241664,
          "online_loss": 0.1773425699211657,
          "train_loss": 0.17468328773975372,
          "validation_loss": 0.02226800099015236,
          "train_metric": 0.04582643508911133,
          "validation_metric": 0.00014349784760270268,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7857629539994377
        },
        {
          "epoch": 237,
          "updates": 3792,
          "samples_seen": 242688,
          "online_loss": 0.17382978508248925,
          "train_loss": 0.1678812950849533,
          "validation_loss": 0.015334900468587875,
          "train_metric": 0.04579618573188782,
          "validation_metric": 0.0001630243204999715,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.7951790269999037
        },
        {
          "epoch": 238,
          "updates": 3808,
          "samples_seen": 243712,
          "online_loss": 0.174702936434187,
          "train_loss": 0.1740553230047226,
          "validation_loss": 0.019398147240281105,
          "train_metric": 0.04639820009469986,
          "validation_metric": 0.00016079180932138115,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.804237447999185
        },
        {
          "epoch": 239,
          "updates": 3824,
          "samples_seen": 244736,
          "online_loss": 0.17024068330647424,
          "train_loss": 0.16939552128314972,
          "validation_loss": 0.017520621418952942,
          "train_metric": 0.046229951083660126,
          "validation_metric": 0.00012140381295466796,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.813468839998677
        },
        {
          "epoch": 240,
          "updates": 3840,
          "samples_seen": 245760,
          "online_loss": 0.16873406595550478,
          "train_loss": 0.16718655824661255,
          "validation_loss": 0.01258570235222578,
          "train_metric": 0.04589468613266945,
          "validation_metric": 0.00010539007053012028,
          "lr_first": 0.003,
          "lr_last": 0.003,
          "seconds": 3.82256445999883
        }
      ],
      "best": {
        "epoch": 232,
        "updates": 3712
      },
      "best_validation_metric": 0.00010087719419971108,
      "config": {
        "label": "MAE + outliers",
        "batch": 64,
        "lr": 0.003,
        "schedule": "constant",
        "normalization": "none",
        "residual": false,
        "concatenate": false,
        "dropout": 0.0,
        "activation": "silu",
        "loss": "mae",
        "outliers": true,
        "max_updates": null
      },
      "seed": 11,
      "parameters": 2241,
      "train_samples": 1024,
      "validation_rmse_m_s": 0.010043763945837789
    }
  },
  "epochs": 240,
  "seeds": [
    11,
    22,
    33
  ],
  "workers": 16,
  "available_cpus": 16,
  "threads": 1,
  "torch": "2.14.1+cpu",
  "seconds": 23.480980145001013,
  "selected_id": "14985df6b90f",
  "selected_config": {
    "label": "LayerNorm + silu + plateau LR",
    "batch": 64,
    "lr": 0.003,
    "schedule": "plateau",
    "normalization": "layer",
    "residual": false,
    "concatenate": false,
    "dropout": 0.0,
    "activation": "silu",
    "loss": "mse",
    "outliers": false,
    "max_updates": null
  },
  "test": [
    {
      "seed": 11,
      "rmse_m_s": 0.005738990490365495
    },
    {
      "seed": 22,
      "rmse_m_s": 0.004062206619929928
    },
    {
      "seed": 33,
      "rmse_m_s": 0.006834206351271969
    }
  ],
  "source_sha256": {
    "training_tools.py": "73f9c91df0bdc2d0681a2025d3d554d11d289e1fbfed038047dec4277c3e6236",
    "training_strategies_experiments.py": "d45307e44b19857ad48314ab8933ed928537b17c540020af6681d6dcc062c9bd"
  }
}